Quick Links

• Where we started

• Goals and Objectives

  • Needs and wants

  • Power station metrics

  • Strategies and options

• Finding a power station

  • Anker Solix

  • Ecoflow

  • Bluetti

  • A point about standalone batteries

  • Picking a power station

• Now, theorycrafting

• About ROI...

• My current setup

• Milestone

Where we started

When I rebuilt my network cart a few years ago, I had originally planned for having two UPS units. I had hoped to deploy a maintenance-bypass ($$) or automatic-transfer switch ($$$) so I could actually do staggered testing of the units without having to first power everything down. Powering down everything beforehand is a necessity - should the UPS fail the test, it would actually drop everything it was meant to power up.

A dual-UPS setup (specifically with an ATS) has the additional benefit of [nearly] doubling the runtime. According to the manufacturer, with my network cart pulling 300W, I should expect somewhere in the range of 20-25 minutes however, real-world outages are slightly less optimistic — definitely less than 20 minutes. Considering the fact that it it takes a non-insignificant (in reality, a substantial) amount of time for my NAS to actually power down once the determination has been made, I realistically plan for having a measly 7 minutes before things get exciting.

Even hypothetically doubling that runtime doesn’t exactly give me a ton of breathing room.

Currently, I do have an EGO Power+ Nexus Portable Power Station and four 5Ah batteries: a cumulative 1120Wh of power. This can (and has!) easily kept the network cart running for over two hours.

So, when the power goes out for more than a blip, I need to do a mental evaluation of whether this will end up being an extended outage or not. Once the determination has been made to transition over to this separate power station, the process is:

1. Grab the EGO power station from the garage and bring them to the office

2. Grab four batteries, bring them to the office and plug it in

3. Power up the power station

4. Move the network cart over to the EGO power station

Luckily I have been home for every major power outage, so I just handle this — it’s definitely doable within 7 minutes, but there certainly isn’t a lot of wiggle room.

Last fall, I got around to deploying a small proof of concept for using solar/batteries to peak-shave a work laptop. Even with the reduced solar generation over the winter months, for such a small draw (mostly averaging 30W), the power station has easily been able to peak- (and mid-) shave the laptop for my entire work day without any intervention. It’s been great.

Thinking about how easily the EGO power station handles the load, I started thinking about deploying a dedicated power station for the office…

At a high level, I wanted some kind of battery system that I could use to carry some (or all) of the office electrical load for some (to be determined) amount of time. I wanted to address both backup-power and peak-shaving. There was a bit of wiggle-room in the final requirements to account for cost, complexity, scalability, ease of use etc.

There are three broad solution categories:

1. An off the shelf, ready-to-go solution. You pay a premium for the convenience but you have to fit your needs inside the manufacturer’s product stack (and there is the risk of being locked in an ecosystem)

2. A fully DIY solution where you mix and match inverters, batteries, charge controllers etc. In the long-run, this ends up being a much more cost effective solution, but getting to the first (and possibly each) checkpoint costs more (absolutely) and brings with it a lot of complexity

3. A professionally installed, home-level system. This is something we’d like to consider down the line but definitely not at this time and definitely outside of the scope of what I want to do

Although the DIY route is definitely appealing and I think I will eventually get there, this isn’t my only project — having a power station is just a checkpoint on a much more complicated journey. For the time being, I need the simplicity of ‘open the box and go’.

[Back to Top]

Goals and Objectives

Needs and wants

Whatever system (or systems) I landed on, I absolutely wanted

• To have a “substantial” amount of runtime to the network cart. While I didn’t initially have a specific runtime I wanted to hit, after seeing that my EGO power station could run the network cart for 2 hours, I set that as my starting target (as this would be largely influenced by cost)

• To be able to peak-shave as much as a I can (and to do it robustly, without minimal fiddling on automations)

Dependant on pricing, system complexity etc., I wanted to extend both of the previous goals to the computers in my office

Looking at the on-peak windows for the above pie charts:

• Winter: I need to be able to carry the office for 4 hours in the morning (0700-1100). There’s a 6-hour window mid-day where I can either solar-charge (winter solar will be a challenge) or AC-charge (at a lower rate). My experience in the winter will largely be determined by the battery capacity.

• Summer: here, I need to be able to carry the office for a continuous 6-hours (1100-1700) but summer solar generation is more promising; my experience in the summer will largely be determined by solar input.

[Back to Top]

Power station metrics

1. Inverter size. This is measured in watts and the metric sets the limit on what you can do. For example, something like a hair dryer might need 1500W to operate — if you have a 200W inverter, you’ll never be able to run it.

2. Battery capacity. Now that you can run whatever it is you’re trying to run, this determines how long you can run it for. Measured in watt-hours, and continuing our example,
   

   • If your battery has a capacity of 500Wh, then you could run your 1500W hair dryer for 1/3rd of an hour (20-minutes)

   • If your battery capacity was 1500Wh, then you could run it for 1 hour

   • If your battery capacity was 3000Wh, then you could run it for 2 hours

3. Solar input capacity. Measured in watts, this expresses how fast you can recharge your battery capacity (even while actively draining it). By focusing on solar input, we are looking at how fast (or how much) we can recharge the battery for free. Indirectly, this value affects how long you can run things for. Assuming we have a 500Wh battery and continuing our example:

   • If we have 500W of solar-input, then our 1500W hair dryer effectively only pulls 1000W, extending our runtime from 1/3 hour to 1/2 hour

   • If we have 1000W of solar-input, then our hair dryer effectively only pulls 500W, extending our runtime from 1/3 hour to 1 hour

   • If we have 1500W of solar-input, then our hair dryer effectively pulls 0W, meaning we can run indefinitely (while the sun is shining of course)

As I alluded to previously, there is a bit of a ‘soft-limit’ on battery size determined by:

(1440W - Continuous Draw) x 12-hours (of off-peak)

or

(1440W - Continuous Draw) x 18-hours (of off-peak + mid-peak)

With zero draw, this gives us a (silly) hypothetical peak of 17.3-25.9kWh for battery capacity - any larger and we wouldn’t (really) be able to recharge it within a 24-hour period on a typical 15A residential circuit. Note that for safety purposes, even though a 15A circuit can pull 1800W, we shouldn’t be doing so continuously.

More importantly though — my draw is never going to be zero. But can come back to this later when we go down the theory-crafting rabbit hole.

[Back to Top]

Strategies and options

Option 1

I would get a smaller power station exclusively for the network cart; I can integrate it into the cart directly and this would be the end of the project. This gives me a self-contained starting point that fulfills my core requirement of giving the cart extended runtime and peak-shaving.
For the main computers in the office, I could follow up with a more substantial setup (either DIY or a larger off-the-shelf solution). The downside with this option is that, down the road, there will be two separate systems to manage and I would want to try and find a way to charge both systems with a single solar array.

Option 2

Here, I get substantially larger power station — with enough oomph to drive everything I need. In addition to the higher initial cost, I would need to accept that initially, my runtime may be sub-optimal — since I’d be loading the everything in the office on battery power.

I would need to limit myself to models that support expansion batteries (either officially or DIY) and I would have to prioritize power-stations that more usable solar input (as there is more draw to try and offset).

For both strategies, I wanted to tie this into Home Assistant - with or without official support (but obviously, platforms with official integration support would be prioritized). Ultimately, despite the jump in price, it made a bit more sense to go with Option 2 — this way I only have one system to worry about maintaining, integrating and automating.

[Back to Top]

Finding a power station

I definitely wanted a ‘open the box and go’ solution (and recognized that I would be paying a premium for that). One advantage of going this route is having access to operating modes out of the box, without having to further do custom automation work. Specifically, I was looking to do some level of basic time-of-use/peak-shaving natively with whatever app the power station came with.

Anker Solix

Since I already have an Anker Solix C300, I naturally started my search with Anker products. When I initially looked into this, there was an unofficial integration to get Solix power stations into Home Assistant. Since then, Anker now has an official integration. At this time, there’s only support for a single product, so the official route is something to keep an eye on.

• [$] C1000 Gen 2

  This is the smallest, most cost-effective power station of the bunch. It technically has enough oomph to handle the entire office but only just.

  • Inverter: 2000W. My nominal load for the office is 1100W so this can carry everything

  • Battery capacity: 1024Wh. Without considering solar, I have less than 1-hour of battery-based runtime — not enough to meaningfully do any peak-shaving. Since the Gen 2 model doesn’t support expansion batteries, I’m at a dead end here.

  • Solar Input: 600W. Even with maximum solar input, with my 1100W load, I’m draining my battery at a rate of 500W. This puts my hypothetical best case runtime at roughly 2 hours.

  • Price point: ~$600

  Even if I were to use a standalone battery to act as faux-server, the built-in battery will continue to drain — and once the built-in battery drains down to empty, the inverter will shut off.

• [$$] C2000 Gen 2

  Bigger in every way compared to the C1000G2, this unit realistically represents the smallest offering I would consider.

  • Inverter: 2400W. With a [slightly] bigger inverter, my hope is that, because my nominal load represents a lower utilization (of the inverter), there’s less noise in the office (fans, inverter whine etc.). Unless the noise is extreme, this is just a nice perk.

  • Battery Capacity: 2048Wh. Doubling the battery capacity is only half the equation: this unit support expansion batteries as well. Out of the box this gives me just under 2-hours of runtime

  • Solar input: 800W. Although it’s a small bump in the solar rating, it does have a meaningful impact on my runtime. With a nominal 1100W load, the (peak) 800W solar input makes the effective drain only 300W. This gives me a runtime of roughly 6 hours.

  • Price point: ~$1000

  To actually hit 800W is a different challenge: I would probably do this via standalone battery (at additional cost) acting as faux-solar. But at least it’s possible. Like the C1000G2 though, with a 1100W load and only 800W solar input, I will eventually run the battery down and then the inverter cuts off.

• [$$$] F3000

  This is the biggest/most-expensive power station I wanted to consider. This does ‘all of the above’ and more.

  • Inverter: 3600W. For the office, I definitely don’t need the extra oomph here, but my hope is that the fans would be minimal

  • Battery capacity: 3072Wh. Out of the box I’m looking at roughly 3-hours of run time (for the entire office, no less). With the tons of options for battery expansion this power station has a lot of points for scalability (if you have the money to spend)

  • Solar input: 2400W. Here we finally start getting into some seriously big numbers. Here I can (theoretically) run the office entirely on solar.

  • Price point: ~$2000

  The F3000 brings a bit more to the table too (at additional cost).

  • Multiple units can be paired together for 240V power

  • There’s more support for directly tying into your breaker-box to act as an automatic peak-shaving transfer-switch

I was really torn between between the C2000G2 and the F3000. For what I wanted to do, the C2000G2 was a near-perfect fit — the only limiting factor was the solar input. The F3000 is clearly a much more capable unit - but to unlock that extra solar, I was paying for a lot of [awesome] features I wasn’t going to use; also, some of the awesome features (like 240V/home-transfer switch) required buying an additional F3000 to capitalize on…

[Back to Top]

Ecoflow

Ecoflow was one of the two brands I wanted to look into. They seem to be positioned as a premium, tier-1 power station option. The downside? Their entire 3rd Gen product stack is a disaster. I did some spot-checks for some models but there was no way I was going to deep-dive into their product stack (and there was no way I was going to spend a ton of money without doing a deep dive). Just for this post, I did a very quick blast through the three attributes I care about and loosely grouped things (without considering price) by solar input:

• Minimal solar

  • Delta 3 1000 Air: 500W, 960Wh, 500W solar

  • Delta 3 Classic: 1800W, 1024Wh, 500W solar

  • Delta 3 1500: 1800W, 1536Wh (expandable), 500W solar

  • Delta 3 Max: 2400W, 2048Wh, 500W solar

• Medium Solar

  • Delta 3 Plus: 1800W, 1024Wh (expandable), 1000W solar

  • Delta 3 Ultra: 3600W, 3072Wh (expandable), 800W solar

  • Delta 3 Max Plus: 3000W, 2048Wh (expandable), 1000W solar

• Plenty

  • Delta 3 Ultra Plus: 3600W, 3072Wh (expandable), 1600W solar

  • Delta Pro 3: 4000W, 4096Wh (expandable), 2600W solar

• Crazy

  • Delta Pro Ultra: 7200W, 6144Wh (expandable), 5600W solar

  • Delta Pro Ultra X: 12000W, 12288Wh (expandable), 10000W solar

As someone who was struggled to make sense of QNAP’s NAS offerings, I find the entire 3rd Gen EcoFlow product stack to be triggering.

From an automation perspective, there’s an unofficial Home Assistant integration.

[Back to Top]

Bluetti

After the disappointment that was the Ecoflow product stack, I was mostly resigned to going back to debating the C2000G2 vs F3000 but for completeness sake I decided to look at Bluetti. Until this point, I hadn’t considered Bluetti as my impression at the time was solidly Tier-2 vibes: “everything works but the polish isn’t there”. After looking at their product stack, I was refreshingly surprised…

One really compelling point in Bluetti’s favor is official support for Home Assistant. To make my life easier on the automation front, I specifically wanted to focus on models with Wi-Fi, which rules out the more affordable AC-line of power stations and mostly restrict myself to Elite (or more upmarket) models.

• [$] Elite 100 V2

  On paper, this appears to be similar to the C1000 Gen 2 but with a much more substantial 1000W solar input

  • Inverter: 1800W. While this is the smallest inverter, it can carry all my needs. The only concern is how hard the fans might run

  • Battery Capacity: 1024Wh. By itself, the battery capacity only affords me ~1hr so my only hope is too leverage the substantial solar to increase this.

  • Solar Input: 1000W. This is a massive amount of solar for such a tiny unit. At peak solar input, my runtime increases to ~10 hours.

  • Price Point: <$600

  Looking at just these core metrics, this is an easy win over the C1000G2. My only concern whether or not my inverter utilization% will cause any fans to run and how loud they might be.

  Since there is no way to expand the battery here, the most practical strategy would be to use a standalone battery and hook it up as a 1000W faux-solar. Doing things this way would give me nearly 10 hours theoretical run time for the whole office. Very compelling.

• [$$] Elite 200 V2

  Double the battery, and a bit bigger of an inverter, on paper this is better ’value’ than the 100V2. Compared to the C2000G2, has a 200W solar input advantage but loses the ability to expand the battery.

  • Inverter: 2600W

  • Battery Capacity: 2074Wh

  • Solar Input: 1000W

  • Price Point: $1000

  Again, using a standalone battery as a 1000W faux-solar setup, this can give my office a theoretical peak run time of 20 hours. Despite this being an absolutely bonkers runtime, this doesn’t seem like a compelling option for me (as 10-hours from the 100V2 is already plenty sufficient).

• [$$$] Elite 300

  Despite the existence of the 200V2, this power station feels like it would be a better option.

  • Inverter: 2400W

  • Battery Capacity: 3014Wh

  • Solar Input: 1200W

  • Price Point: $1700

  With this power station, I have the option (using a standalone battery configured as faux-solar) to run my office ‘indefinitely’ from solar. The asking price for this is a lot though.

• [$$$$] Apex 300

  W solar input

  • Inverter: 3840W

  • Battery Capacity: 2765Wh

  • Solar Input: 2400W / 6400W

  • Price Point: $1900

  Now this is an interesting option. The inverter on this power station is a tank and can do a ton more than I need at this time (but all within the wish-list of what I might want down the road).

  The super attractive point is the dual high-wattage solar charge controllers for a very respectable 2400W solar (out of the box). With a unique option to add an additional 4000W of solar input via an accessory, a power station like this could grow with me over time.

  Compared to the Elite 300, this is clearly the better strategic option to go for.

[Back to Top]

A point about standalone batteries

So far I’ve considered the idea of using a standalone-battery as faux-solar: the idea here is to connect the battery to the solar-input port on the power station. Under the hood, this is DC-DC power transfer — the power station distinguish (or care) that the power is coming from a standalone battery versus a solar panel. When doing this, the power station happily charges its internal battery using the ‘solar’, subject to solar-input limits. This is why I’ve focused so much on the solar-input of the power stations.

Something to note though: if the power station is net-draining (i.e., Load > Solar Input), then eventually the battery on the power station will hit zero — when this happens, the inverter on the power station will shut off, regardless of how much power is remaining on the standalone-battery. Since the standalone battery is providing ‘solar’, the power station will continue to charge and eventually it will turn back on. Assuming your load is consistently bigger than the solar input, you will enter a loop where the power station runs down, shuts off, charges back up, turns on, runs down, turns off…

Using the C1000 Gen 2 just as an example, it has

• 1024Wh battery

• 600W solar-input

Given my nominal 1100W draw, this means that I would have a constant drain of 500W on the battery which would yield a theoretical runtime of just over 2-hours.

Without any way to expand the battery, this is a hard limit no matter how big our standalone battery is.

There is one more thing we could do: double-dip the standalone battery with a separate inverter. The C1000 Gen 2 has a limit of 1200W AC charging, so a small 2000W inverter is plenty for our experiment. With 600W faux-solar and 1200W AC charging (or really, with just the AC charging alone), we can run our loads entirely off battery power (of course, until the standalone battery runs out).

Between the solar and inverter, we would only be pulling 55A — well within the 100-200A operational limit of most standalone batteries. There is one minor downside: the power station has no way to directly query the standalone battery (remaining life, etc.). While many standalone batteries do have their own [Bluetooth] app, this becomes an additional complication to consider from an automation perspective.

While double-dipping technically works, it’s really inefficient: we would be taking DC power, converting it to AC via the inverter, just to have the power station convert it back to DC as part of the charging process.

Another thing to consider is the potential for noise: both the inverter and the dedicated charger (for the standalone battery) have fans that will likely be running in addition to the fans on the power station. For my first go at this, I wanted to keep things a bit simpler, so I’m not considering double-dipping.

[Back to Top]

Picking a power station

Putting all of the models under consideration in a spreadsheet, we get can narrow things down a bit:

• The Elite 100V2 beats out the C1000G2 based on the substantial jump in solar input

• Between the Elite 200V2 and the C2000G2, it’s much closer. Entirely by itself, I think the Elite 200V2 takes the win here with the not-insignificant advantage in solar-input however, the option to get an expansion battery on the C2000G2 should not be ruled out (despite OEM expansion batteries being silly expensive).

• In the context of the available options, the Elite 300 is a bit of an odd duck and only really makes sense if you were considering the C2000G2 with the expansion battery (~2200CAD). If we scale the price of C2000G2+Expansion down to match the battery capacity of the Elite 300, it is roughly 1620CAD. In that context, the Elite 300’s substantial advantage in solar input makes it a compelling alternative.

• The F3000 and Apex 300 are fairly evenly matched: the F3000 has more battery capacity out of the box vs the Apex 300 being able to do 240V natively. Having the option to get an additional 4000W of solar input into the Apex 300 is really unique.

My final short list was:

1. Apex 300 — the most expensive option of the three, but it’s a ‘do everything’ option

2. Elite 100V2 — going with this option means that I would definitely have to get a standalone battery (the sooner the better)

3. C2000G2 — this unit is ‘medium’ at everything; I think I would value the potential for an official ($$) expansion battery versus not having the option at all

[Back to Top]

Now, theorycrafting

Now that I have a power station in mind, we can start to do the fun theory-crafting. I went down a bit of a rabbit hole to try and model power-usage in a spreadsheet. My scenario is a bit unique in that I’m modelling power-draw that is relatively stable - my computer is always doing something.

View fullsize

View fullsize

View fullsize

View fullsize

For each scenario, I tackle them in the same way:

1. See how much of an on-peak period we can offset using the battery

2. If there’s remaining battery capacity, try and offset mid-peak where applicable. Otherwise, allow the battery to recharge during mid-peak

3. During off-peak periods, switch the power station to AC charging

The two variables at play are: the amount of starting energy available (i.e., whether we have an expansion battery or not) and the amount of solar generation available. The following pictures provide a bit of context in the thinking behind each strategy

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

When we start going down the strategy rabbit-hole, there are few details we need to remember:

1. Inverters consume power. Even if there’s nothing plugged in, just the fact that the inverter is enabled and standing by consumes power; how much power is ‘wasted’ depends on the model. In our spreadsheet-model, we can account for this a bit by reducing the ‘efficiency’ number for our power station.

2. Solar generation can be very fickle. In all but the most intensely sunny days, solar production will not be consistent - a single cloud rolling by, branches swaying in the wind and casting shadows or even bird poop can impact solar generation. We can tackle this two ways: either over-panelling ($$) or reducing the numbers around solar input in the spreadsheet.

3. We don’t (and probably shouldn’t) run the power station down to absolutely empty. For my testing, I set my operating mode to bottom out at 15%; this means that if the battery drops below 15%, the power station will cut-over to grid-pass through to keep the battery from going lower. This ensures that there is always a minimum amount of power available in case there is an actual power outage. We can tackle this by either using additional expansion batteries or by reducing the efficiency metric in the spreadsheet even further to give us some buffer.

All this to say that the spreadsheet is very much a dynamic, work in progress that also may be very uniquely dialed into my specific use case.

[Back to Top]

About ROI...

After all the scenario work, we can pull all the aggregate data together and try and draw some conclusions. We can see from the summary below that theoretically, the best options are ‘power station with lots of solar input’ and the distinctly second-best options are “power station with an expansion battery but also lots of solar input”. For my electrical needs, it seems that two expansion batteries are definitely not the way to go.

Here, we composite all of the daily-operating costs of my office across all of the different scenarios, accounting for both winter- and summer- billing schedules.

This spreadsheet is a bit of a work in progress and I’ll need to do some fine-tuning over time both to tailor the baseline settings (assumptions), change strategies and/or correct formula errors. Despite the raw numbers suggesting otherwise, my gut feeling is that going for an expansion battery is the better option for three reasons:

1. The expansion battery gives me additional runtime outside of any of this cost-optimization stuff

2. It’s possible that the factors I’ve built into my theory-crafting around the sun are unrealistic/optimistic

3. Even if my assumptions around sunlight are reasonable, any additional capacity allows me to offset load without the additional variability of depending on the sun

That being said, price (and availability) obviously plays a massive part in this equation.

[Back to Top]

My current setup

So I went with the Bluetti Apex 300 and currently, I’ve only got 100W of solar feeding it. I’ve defined the following in my configuration of the power station:

• Custom Mode [SoC 15-100%, with time-of-use and schedule defined]

  • The 15% means that no matter what, if the power station goes below 15%, then AC pass-through will be enabled. This won’t charge the battery, it just keeps the battery from draining any lower

  • The 100% is the point where, if the battery level goes higher than this, it stops the grid input with the idea being that the grid will top up the battery to the specified percentage and then solar power will fill up the rest of the way: if you have some decent solar available, you might set this to 80% (or lower!). In my case, I’ve got such an insignificant amount of solar input, it doesn’t really make a difference. I’ll revisit this later on.

  • The schedule allows me to define two types of time blocks: [a] charge the battery and [b] discharge/use the battery. Times not explicitly falling in that window will use AC-pass through (where the grid powers the load but doesn’t charge the battery). Note that even when the battery is in discharge/use mode, it still have the 15% safety net

• Backup Mode

  • I use this mode as a “make it charge now”

For my locale, in the winter-months, I have two on-peak time windows (0700-1100 and 1700-1900) and as such, I need to find a way to get the battery topped up before the second on-peak window. The behavior (of ‘Custom Mode’) is only to provide pass-through power during mid-peak periods.

Ideally, I would like to charge the power station during the mid-peak period but specifically avoiding charging more than necessary to handle that second on-peak window. This way, I can leave the majority of the charging to be done during the off-peak period. One way to work around the limitation of the app is to define some of that mid-peak period as off-peak. While this would work, it’s not a clean way of doing it. The second problem I have with this app is that it doesn’t support seasonal TOU scheduling so in the long run, this means I’ll need to handle the automations myself.

To get started, I know that if I have 75% or more remaining, I can definitely handle that second on-peak period regardless of what loads might be on the power station. To do what I want, I setup a few automations in Home Assistant:

• 1100h - If the battery is less than 65%, then I switch to Backup Mode allowing the power station to charge. I’m always running on empty (treading 15%) at 1100 so this always happens.

• 1330h and every half-hour after - If the battery is at least 75%, then I switch back to Custom Mode. This effectively stops charging the battery and just does AC pass through

• 1700h - Since the power station is already in Custom Mode and 1700h is defined as on-peak, it now uses the battery to power everything

[Back to Top]

Milestone

For the time being, the Apex 300 has been great - with the quick Home Assistant automations, the day to day operations are entirely hands-off.

Based on my theory-crafting, I knew that, even in the most optimistic scenarios, the power station doesn’t have enough juice to carry the office through the initial [winter] 4-hour on-peak period. Empirically, I’ve found that I can carry my office for 2.5-4 hours (most of the time, it lasts to 3.5 hours) — and this is running down to 15% reserve. The deviation comes from tweaking user-behaviour: I can explicitly chose to scale down intensive operations during the morning peak window.

That being said, with the summer ToU scheduling coming up, I know that I absolutely can’t offset the entire 6-hour continuous block.

I’d like to get a DIY expansion battery ($$) but I’m actually waiting for pricing and availability to go down that road. As the theory-crafting has already identified, a single 5kWh battery would allow me to offset the entire work day.

A much cheaper option would be to get started with one (or two!) 500W-class solar panels. My office is already ‘plumbed’ to run a panel in — it’s just a matter of identifying a specific panel to get and then the harder task of finding a vendor to order it from. That will be a the focus of another post ;)

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

I’ve had my Home Assistant instance spun up for a few weeks now and I’m starting to play around with sensors — the critical building block that determines what you can react to so this was a super easy opportunity to see if I can automate my way out of this.

Quick Links

• Where we started

• Goals and Objectives

• Rough Design

• Step 1 - Simplifying the socket

• Step 2 - Running the lights

• Step 3 - Motion triggers

• Milestone

• Appendix

Where we started

The lighting in our upstairs linen closet is not the greatest: with a single, awkwardly placed pull-string light socket, visibility was always an exercise in best effort. With a motion adapter socket and an incredible 250W (equivalent) bulb, I was able to brute-force my way to usable lighting. When triggered, the motion adapter lights up for 10-minutes and while I haven’t had any issues, this always bothered me for two reasons:

1. Most of the time, we’re in and our of the linen closest in under a minute and

2. Based on how we use the space, the very warm bulb often sits right next to flimsy plastic

View fullsize

View fullsize

View fullsize

To address the heat issue, I initially tried out a couple options:

1. I tried a bulb with a built in motion trigger: I found that these are often capped at 100W (equivalent) - a noticeable drop in light output but more crucially: the awkward placement of the light socket makes triggering the motion unreliable and inconsistent

2. I tried a basic smart bulb with a decoupled motion trigger. Functionally, this worked well - the trigger worked great but I had to deal with a massive drop in light output.

[Back to Top]

Goals and Objectives

I wanted a bit of a Goldilocks solution:

• I wanted to reduce the physical footprint of the bulb:

  • To avoid having a hot bulb resting against plastic and

  • To reduce the chances of totally smacking the bulb off the wall because of how the socket is oriented

• I didn’t want to compromise on lighting and in fact, wanted to find a way to get better light coverage

• I wanted to find a way to improve the motion trigger — if I can avoid having the lights needlessly on for 10-minutes

• It had to pass ‘the Wife Test’

[Back to Top]

Rough Design

Since we have the ubiquitous metal-grate shelving in our linen closet, I thought to hang light bar(s) on the underside of the grate and hope that the light coverage was sufficient (possibly by just throwing more light bars at the problem). The problem with this is that the towels and blankets etc block the light from evenly covering the space (who would have thought?). We can brute force this with six lights (two per layer) — which is too many damn lights (even for me).

We have some ducting behind the walls on in the closet - so it’s a surface I can hang lights off with magnets…

[Back to Top]

Step 1 - Simplifying the socket

The magic here comes in three parts: a socket to power adapter, a rotating power adapter and a smart switch (you can use whichever ecosystem you happen to be in). The compact nature of these parts allows us to completely remove the ‘bump risk’ that I had before.

[Back to Top]

Step 2 - Running the lights

I opted to use (just!) two 2-ft light bars on the far wall. I would hang them using magnets recessed into a plywood mounting plate. To route the wires out of the way, I ran them down along the wall and then up and across the ceiling. For my closet, I used a 10ft thin extension cord to give me the extra length but depending on your setup, a T5/T8 extension might work just as well/better.

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 3 - Motion triggers

I used a Zigbee motion sensor to kick off the lights: using the tail end of a cable tie, I can gently adjust the specific angle it points at, however I’ve found that realistically, it doesn’t really matter for this particular application.

View fullsize

View fullsize

Behind the scenes, I use Home Assistant with Zigbee2MQTT

alias: Motion Lights - Linen Closet ON
description: Recreate motion trigger for linen closet
triggers:
  - trigger: state
    entity_id:
      - binary_sensor.motion_sensor_linen_occupancy
    to:
      - "on"
conditions: []
actions:
  - action: switch.turn_on
    metadata: {}
    target:
      entity_id: switch.zigbee_13_linen
    data: {}
mode: single

alias: Motion Light - Linen Closet OFF
description: Recreate motion trigger for linen closet (off)
triggers:
  - trigger: state
    entity_id:
      - binary_sensor.motion_sensor_linen_occupancy
    to:
      - "off"
    for:
      hours: 0
      minutes: 0
      seconds: 30
conditions: []
actions:
  - action: switch.turn_off
    metadata: {}
    target:
      entity_id: switch.zigbee_13_linen
    data: {}
mode: single

And that’s it!

[Back to Top]

Milestone

This was a super easy project to knock out and I’ve now implemented a better version of a motion bulb for our linen closet! I met/exceeded all my original goals:

• I can control the placement and targeting of the motion trigger to reduce accidentally turning it on

• The cooldown window is much more realistic so the lights aren’t on unnecessarily

• The light coverage is much more even and usable

• I’ve reduced the ‘accidental bump’ factor

• Removed the hot-bulb-resting-on-plastic factor

When the weather warms up, I’ll consider painting the plywood to better make it blend away into the wall but this project is all done!

[Back to Top]

Appendix - Parts List

• 2ft light bar, 5000K

• 130-lb magnets

• Zigbee motion sensor

• Connecting to the socket

  • Socket to power adapter

  • Rotating power adapter (to tidy up the wiring)

  • Zigbee smart plug

  • 10ft thin extension cord

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Where we started

• Goals and Objectives

• Theorycrafting

  • What is peak-shaving?

  • Some easy examples

  • What about a bigger load?

  • A realistic example

• Step 1 - Using the C300 to offset on-peak

• Step 2 - Bypassing the inverter

• Step 3- Getting solar into the office

• Step 4 - Thinking ahead

• Step 5 - ROI

• Step 6 - Window shopping

• Step 7 -  Fine tuning

• Milestone

• Appendix - Parts list

Where we started

A few months ago, I was wandering through a liquidation store when we came across the Ryobi 18V Hybrid Cooler. Growing up, we had a 12V car-cooler that came in handy when we went to far-away grocery stores and meat markets. While some 12V coolers come with an AC wall option (and I guess I can always buy a dedicated 12V adapter), the Ryobi cooler gives me the option to totally disconnect, relying on batteries entirely.

Looking at this from a different perspective, the cooler comes with it’s own built-in UPS. With a regular price of $700CAD (!!), there was never a chance in hell I would have considered this but the liquidation price of $300? This puts it in the same ballpark as many 12V coolers. Early in the [camping] season, a pair of 9ah batteries (~162Wh each) was more than enough to last the weekend but as the weather really heated up, I needed to find a better way to power the cooler without packing a whole bunch of batteries…

As luck would have it, I stumbled upon the Anker Solix C300 shortly after. From a camping perspective, this provides me with 288Wh of power: between the Solix, the two batteries installed in the cooler and a couple of extras, I’ve got 936Wh of power. Enough to reliably go off-grid during a weekend camping trip regardless of the temperature. But enough about camping — how I can I really use this?

You know, for computers.

View fullsize

View fullsize

[Back to Top]

Goals and Objectives

It’s no surprise that, the computers represent a significant portion of my electrical bill. I’ve been toying with the idea of using power-stations to try and offset some of the load when electricity is the most expensive throughout the day. Obviously, getting a beefy enough power-station that can comfortably drive my computers is not a cheap undertaking, so I wanted to use this small, camping-oriented power-station as a sandbox test-bed.

At a glance, these are the relevant specs:

• 288Wh battery capacity

• LiFePO4 batteries rated for 3000-cycles (meaning, after 3000 charge-discharge cycles, it should retain 80% original capacity - 230Wh or so)

• 300W inverter with 600W max draw

• Rechargeable via: AC (100W min, 330W max), USB-PD (140W max) or solar (11-28V, 100W max)

• Idle power consumption of 11-12W

Since I already have the C300, I wanted to a way to leverage it’s battery.

[Back to Top]

Theorycrafting

What is peak shaving?

Before diving too deep into things, here’s a sample time-of-use cost of electricity chart.

So there’s this idea of peak shaving: if we can find a way to reduce electrical draw from red and orange periods — we shave the cost of electricity during the peak periods of the day.

[Back to Top]

Some easy examples

It can be easier to visualize and explore how we might leverage this with an example. For a simple example, let’s suppose I have a 100W load that needs to run 24x7; let’s make some assumptions to make things a bit simpler:

• That there are no efficiency losses from the inverter

• There’s no idle-power consumption

• The power-station has a max recharge rate of 300W

• We are working with an 1800W circuit or so (a standard 15A outlet)

• Once the battery is depleted, we will immediately switch to grid-power and start recharging the battery - I don’t know of any system that allows you to switch to grid-power without recharging the battery. What this actually means is that our load (currently 100W) and our recharging (currently 300W) must fit within this 1800W allowance.

  • You can easily see how we can run into issues as our load and charging rates go up

  • Some power-stations allow you to adjust the recharge rate so you can optimize things - more on this in a bit

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

What about a bigger load?

Things get a but more complicated as we start to introduce bigger loads: since the circuit’s capacity (1800W) is fixed, this means that we will need to reduce the charging rate so that we’re tripping the breakers.

In the case of an 1100W load, this means that we would have to limit our charging rate to a maximum 700W or so. Since our charging rate is less than the load, there is an upper limit to how much we can recoup during off-peak hours. This ends up being a rough ceiling on the battery size that makes sense.

For this specific scenario, we have a soft-limit of roughly 8400Wh on the battery since that is the most we can replenish during off-peak hours. Adding more battery capacity doesn’t change this and just kicks the can down the road so to speak. At some point we will run out of battery and when that time comes, we will only be able to put 8400Wh back into the battery - the only way to stretch the runtime is to either increase the rate of charge or reduce the load.

So far I’ve left this off, but there are two efficiency losses to consider:

• Inverter efficiency losses: if you’re using the inverter, do note that the is power used just for the inverter to function. This could be 10-watts, or 100-watts — this all depends on the inverter

• Battery degradation: that 100Wh battery we bought years ago might only have 85Wh of usable charge left in it now

[Back to Top]

A realistic example

To bring things back down to reality a bit, consider a much more realistic example. This might be my computer, with three power bands:

• 2hours @ High: 850W (i.e., gaming)

• 11 hours @ Medium: 500W (monitors on, computer still churning at 100%)

• 11 hours @ Baseline: 250W (monitors off, computer churning at 100%)

When we plot this out we can begin to see where we might try and offset costs.

This is all very exciting, but I’m not ready to tackle a real computer yet — I need to find a baby-steps version of this. By coincidence, my work laptop is a perfect candidate to test this all out:

• Not a significant power draw

• Fairly predictable [work] schedule

• If I somehow run out of battery, it’s not the end of the world — the laptop has a built in battery

[Back to Top]

Step 1 - Using the C300 to offset on-peak

Since the C300 was intended to be used in a camping capacity and totally not in a home power-offsetting capacity, a lot of the fancier time-of-use settings aren’t available. Especially for the way I intended use it, I had to make the following setting changes:

• Disable ‘Smart AC Output Mode’ — this forces the inverter to run continuously, even when the laptop isn’t running. I could leave this setting enabled but then I would have to power on the inverter (a single button press) every time I started my work day. That’s one more hurdle I don’t really want to deal with although another way to look at things: this is an option I can keep in my back pocket if I need to squeeze more runtime out of this setup

• Disable ‘Device Timeout’ — this also keeps the C300 from going to sleep in general

To work around the limitation of not having the built-in ability to control power scheduling, I a pair of power-monitoring smart-outlets. This would allow me to control (and measure) power coming into the C300 as well as measure the power going to the laptop.

By placing a power monitoring outlet upstream of the C300, I can monitor the idle power consumption when I know everything is off. The C300 has between 11-12W of idle consumption.

The loads are:

• My work laptop uses roughly 25W

• The C300 itself has roughly 11-12W idle power consumption

• Each of the smart outlets consumes 0.5W (idle) or 1.0W (active)

The ideal goal is to shut-off power recharging the C300 at 0700h (when off-peak rates end) and restore power after 1900h (when off-peak rates resumes). Would the 288Wh capacity be sufficient to carry everything for 12-hours?

Not quite, but there is an important nuance to consider: that I only needed to offset the laptop’s power draw during my work day and more specifically, I only needed to offset that power draw if it happened to fall during mid- and on-peak hours. What this means is that if I can shift some of my working hours into the off-peak period, that consumption would be out of scope of any optimization.

After testing for a few weeks, I found that I was able to offset the laptop’s power draw almost all of the time. To plan for this a bit better, I scheduled Switch A to turn on at 1000h for 10 minutes. With a 330W charging speed, this 10-minute window allows me to recharge roughly 55Wh (19%). This quick top-off ensures that I can make it all the way to off-peak hours no matter what my work day looks like, and this top-off is done at mid-tier rates.

[Back to Top]

Step 2 - Bypassing the inverter

So far I’ve been powering the laptop using a fairly standard AC laptop power brick but using the AC outlets a power station is inefficient because the inverters themselves require power to operate. In the case of my C300, it’s 11-12W — enough to drain the entire power station in a day with just idle consumption alone! If we can find a way to power the laptop using DC power, we could leave the AC outlets turned off, boosting overall efficiency. There are three ways to accomplish this:

1. Using the USB-C outlet

2. Via the 12V car socket directly using a 12V to USB-C charger

3. Via the 12V car socket indirectly using your own inverter; in my case, I happened to have an old Ridgid 100W inverter kicking around. By using your own inverter, you’re gambling that your inverter has less idle power consumption than the one built into the power-station

Just out of curiosity, I did an informal test:

• The laptop would be powered via the USB hub

• The laptop would not be at 100% power (since there can be some weird on-off cycling when the battery is that full) - I targeted 90% charge

• The laptop would be powered up, all applications closed and it would idle at the desktop, one VNC session would be open. I also did another test where I left the laptop running a stress-test for the duration of the test

• I would watch the app for a 5-minute period and track the reported power output values (it updates every few seconds)

• In between tests, I would top up the C300

• The upstream smart outlet would be powered off for each test

I performed each test (roughly x100 data-points) three times (x3), for each power source (x3), at idle and full load (x2). A ton of data. We already ‘know’ that powering the laptop via USB will be the most efficient, but it is kinda cool to see the numbers come in and validate this understanding.

It was also cool to validate that my standalone inverter was less efficient than the built-in one. The only test I wasn’t able to do was to power the laptop using a 12V socket to USB-C (since I don’t own an adapter like that). Since it’s all DC to DC power, I can’t imagine it being substantially more efficient than connecting directly via USB.

View fullsize

As a result of this test, I discovered two neat things:

1. Having a network card connected to the hub results in a 0-2W draw. If I was desperate to shave this off, I could resort to wireless (never) or consider something like a physical network kill switch (questionable reliability based on the reviews).

2. When powering the laptop via USB: if the laptop battery is not at 100% and we hibernate/power down, the power-station will show a 7-8W draw (this amount includes the 0-2W from the network connection). Once the battery is charged to 100% (either by manually charging it before shutting down, or the trickle charge happens to ding 100%), the draw disappears an the power station returns to ‘just’ the 0-2W.
   
   This oddity does not appear when using the AC ports (but may still be the case) — the power-station is not able to reliably report AC loads under 20W. This may just be an oddity of my specific laptop, or could be a quirk of Thunderbolt, or USB-PD, I’m not entirely sure.

Sadly, I won’t be able to power my actual computer using USB so when I eventually apply this to my computer, I’ll have to go with the inverter route.

[Back to Top]

Step 3 - Getting solar into the office

The next logical step in the experiment was to leverage solar charging to top up the battery during the day. I picked up a Renogy 100W panel and got this plumbed into the basement. I did this by mounting a junction box to the exterior of the house and fishing my MC4 cables through the wall and into the box. By doing it this way, I keep all the connections in one place: if I have to move/service anything, I can handle physical connection/disconnects without having to run back and forth into the house.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Looking online, I saw a lot of guides showing people attaching MC4 connections directly to the sides of the junction box. I specifically opted to run my connections through the front/lid of the box because:

1. This was my first time doing this, I didn’t want to gamble on screwing up crimping MC4

2. Replacing a lid is so much easier than unbolting and replacing a box

3. Servicing/upgrading connections through the lid is also much easier: I can remove the lid, take it over to my workbench and work on it rather than hunching over by the junction box

I specifically went with an RV-style gland with cables because I wasn’t sure how much depth I would need for the MC4 connections (if I were to pass them through the front) versus how deep the junction box would be. The extra length of cabling from the gland is a bit unwieldy but it does give me a bit of operating flexibility when servicing.

I went with a bigger junction box for the possibility to pass two sets of cables through the front but the extra space in the box was super helpful when coiling up the extra length of MC4.

I used a 15-ft MC4 cable inside the house be able to go from the junction box over near by C300 (in hindsight, I probably should have went with 20-ft inside the house as well). I used a spring-locked grommet to keep things tidy. The C300 requires an XT60 connector for solar input, so I used an adapter cable to run the last mile.

I’ve only had this installation for a few weeks so I only have limited (summer) experience but overall, it’s been great.

Obviously, my summertime experience represents the most ideal conditions and as the seasons change, it will be more challenging to keep the battery topped off. That being said, I did my initial testing with no solar recharging, so even if I get a small amount of solar over the winter months, I should come out ahead.

Worst case, I re-enable the schedule to add top-off charging during mid-peak.

View fullsize

[Back to Top]

Step 4 - Thinking ahead

I briefly considered over-panelling: running two 100W solar panels (in parallel) to feed into the 100W-max solar input of my C300. By over-panelling I can

• Expand my charging-window to cover more hours of the day

• Increase the charging output during cloudy periods

With parallel connections, voltage is constant and the current is added up; doing this allows me to keep the incoming voltage within the (low) voltage tolerance range of the C300 and I would be relying on the built in charge controller to clip the current, limiting itself to the 100W it can handle.

I ultimately chose to pass on this idea — for now. I want to get a bit more experience with my setup the way it is, before making substantial changes to it and also because it’s very likely that I’ll upgrade to a bigger power-station in the medium-term.

A bigger solution would allow for a much higher solar input and I’d be better served by transitioning to larger (i.e., 400W) panels.

To preemptively handle cloudy days (and winter), I opted to schedule Switch A to power-on for 5-minutes at 1705h (mid-peak); this would allow me to feed 27.5Wh (9.5%) back into the power station. On sunny/ideal days, these 5-minutes wouldn’t matter (since the battery would already be near 100%), but the quick-top off on a cloudy-day would allow me to stretch the battery to off-peak time.

[Back to Top]

Step 5 - ROI

Inevitably, when thinking about solar, it’s just a matter of time before the idea of ROI (return on investment) creeps up for most people because getting this stuff (at least at scale), is not cheap. Everyone’s situation will be different as it depends on how much power you use, when you use it and the difference between the on- and off-peak rates (which can also change, seasonally). I have an alternative set of rates available to me and it’s neat to see how the theoretical numbers may work out.

I’m not getting too excited about the 75% reduction in daily-operating cost since I’m totally not ready for it. This sample scenario is just for one computer, and a rate-change would apply to all the other things in the house. At the house-level, there are several substantial loads that I don’t necessarily have the leeway to reschedule (i.e., stove, air-conditioner).

For the time being, I’ll focus on the savings that I can get using Rate 1, just with offsetting mid- and on-peak consumption.

[Back to Top]

Step 6 - Window shopping

Without scaling up to whole-house solar, having the MC4 pass-through into the basement does give me some more consumer-oriented options: with a beefy enough power-station, I can start to move towards leveraging the savings while still remaining on Rate 1. Consider some of the larger power stations like the Solix F3000 or Delta 3 Ultra Plus, we are starting to get to the point where there is enough battery capacity and solar-recharge capacity that I can make a meaningful dent into my power consumption.

With either of these power stations (+ single expansion battery), I have enough capacity that I can definitely cover my on-peak usage (approximately 3500Wh) entirely, without relying on any solar recharging. I specifically selected these two power-stations for their above average solar-input numbers: even with just a few hours of usable solar, I can put a substantial amount of charge back into the batteries - likely enough to get me to the off-peak period. There’s also a more than likely chance I can start to add multiple computers, or my network rack to this power optimization as well.

Some exciting stuff - all the more reason I need to play with my C300 in it’s ‘basic state’ to get a better picture of how (and when) I use my power. Obviously, there are other products out there and there’s and entire world of DIY/custom to consider as well. One tempting option would be to get a high-capacity battery and connect that to the C300 (or other power station). The power-station just sees this battery as a ‘solar source’ and I can worry about charging the standalone battery separately.

In addition to getting more real-world experience with this baby-steps C300, it would be beneficial to get per-circuit continuous power monitoring so I can get a better understanding of how and when I use power.

Super long-term, I think the right play is to go with whole-house integrated solar - with enough capacity/generation that I can take a serious look at switching to the alternative electrical rate.

[Back to Top]

Step 7 - Fine tuning

The following

• Every day: 0655 OFF. Turning off while we are still in off-peak pricing

• Weekdays: 1705 ON → 1710 OFF. For weekdays, this represents mid-tier rates and gives me a very quick top-off (if needed) in case it’s a cloudy day and I’ve been pulling more power than typical. Assuming the solar panels haven’t contributed anything, this will put 27.5Wh back into the battery.

• Weekends: 1400 ON → 1415 OFF. For weekends, this is all off-peak. Assuming the solar panels haven’t contributed anything, this will put 82.5Wh back into the battery.*

• Every day: 2020 ON. This is well within off-peak hours and the grid can power the idle-power consumption.

[*] I won’t need this if I’m using just the USB ports on the C300, but I’m playing around with the idea of migrating additional devices to the power-station (like my pedestal fan)

[Back to Top]

Milestone

For the time being, I’ve got my work laptop situation fully handled by the power-station and I’m just looking for small things that I can migrate from grid-power to my power-station. As winter is coming, it’ll be neat to see what changes I need to make (if any) deal with the fewer hours of usable sun.

In the spring, I’ll figure out a way to get the panel’s installed on some tiny angle-stands — for the time being, they are just flopped against the house/fence. I’ll also be looking at what options are available for upgrading to a larger power-station to begin to migrate more things off peak-power.

I’ve got some research to do and a ton of data I need to collect (power usage patterns etc.). This will be an interesting rabbit hole to explore!

[Back to Top]

Appendix - Parts list

• Power station - Anker Solix C300

• Solar panel - Renogy 100W, N-Type, 16BB

• MC4 Cables - Renogy 10AWG (20-ft outside, 15-ft inside)

• RV-style Solar entry gland (3ft 10AWG)

• Cheapo MC4 disconnect tool

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Where we started

• Goals and Objectives

• Rough Design

• Step 1 - Left cabinet

• Step 2 - Right cabinet

• Step 3 - Accessory panels

• Step 4 - USB power

• Step 5 - Speaker mount fixes

• Milestone

• Appendix - Parts list

Where we started

Truthfully, the Mk II retrofit left me in a pretty good state: this desk is super dialed in for how I use my computers. On the surface, everything I want and interact with is generally within my grasp and it’s only after living with the desk for a few months that some (nitpicking?) usability quirks come to light. Since this is a custom desk setup, I can totally go down the rabbit hole of addressing every little quirk.

[Back to Top]

Goals and Objectives

Realistically, there was just an (unending) punch list of small improvements and fixes to implement:

• In order to get my mixer and headphones fitted into my left cubby, I had to give up my pull-out work surface. I wanted to bring this work surface back

• I wanted to revisit the drawers on the right cubby — everything was just a bit too tight (and constantly dragging)

• When I originally built the cubbies, I (intentionally!) didn’t add any bracing so they are susceptible to racking

• The subwoofer compartment was made for a specific subwoofer (the Logitech Z5500). I had since switched to the Logitech Z906 (RIP 5500) which has the port on the opposite side

• I wanted better accessibility when I had to crawl down there

• I wanted to revisit my entire XLR mic and headphone situation so that it would be less frustrating

• The rear-surround speaker mounts (that I built) were not built very well — there was a lot of torque applied to a threaded insert causing it to have a tendency to wobble

• I wanted to have better access to some charging-only USB ports right at my desk

• Previously, I was just punching random holes into the side of the cabinets as I was testing out mounting points for my fan switch and then routing the cables. Instead of punching random holes all over, I wanted to use a sacrificial plate that I could replace if needed

• I wanted to see if there was a way I could improve the work-laptop experience

• Nicole wanted to add some lighting to the DIY leveled floor to minimize chance for tripping or stubbing toes

[Back to Top]

Rough Design

Ultimately, the overall design and dimensions of the existing cubbies were great so it made the most sense to take the existing cubby apart and rebuild it [better], panel by panel - a whole Ship of Theseus moment. This was very much a turbo-project and I didn’t really innovate anything new, so I didn’t end up taking too many pictures.

[Back to Top]

Step 1 - Left cabinet

One of the defining elements of this cabinet is the gaping hole on the side — big enough that I can do in-place hardware swaps without having to extricate the computer from the desk.

The downside of having a massive hole on the side of your cabinet is the hit you take to stiffness and rigidity. It probably didn’t help that some of the panels were also slightly bowed and warped (although I’m unsure if this was entirely due to poor craftsmanship or a side effect of a few years on my ludicrously sloped floor). Either way, I laminated a second panel which really stiffened things up. On the opposite panel, I routed out a small opening - enough for me to crawl under my desk and interact with cables if needed.

When I originally built the cabinets, I was concerned about feeding wires in through the back so I specifically opted not to have any bracing. This time around, I added a minimal amount of bracing at the top and everything is so much stiffer.

View fullsize

View fullsize

View fullsize

The big change is with the drawer. Previously, I stuffed a 12-input (!!) mixer into this drawer but it was always a disaster because I was trying to cram a work surface, a drawer, a computer case and additional ventilation room — all into a limited vertical height. Tolerances were so tight that I had resorted to shaving down some of the audio connectors to squeeze every last millimeter out of it. Even after squeezing every last millimeter of clearance on this drawer, I still wasn’t able to stow my headphones away without folding them up. Woe is me, ha.

For other reasons, I was toying with the idea of swapping out my mixer so I bit the bullet and switched out for the M-Track Duo. This is a much simpler (and smaller!) XLR audio interface. Now I have a ton of extra space in the drawer (that I can leave my headphones in their expanded state). I really miss having a physical, hardware-mute option, so down the road, I may look into getting something a dedicated cut-off mute.

View fullsize

View fullsize

With the drawer/mixer situation sorted out, I had one last annoyance. With my previous setup, when on calls, I would fully open the drawer (since this was necessary to fiddle with the mixer). With the drawer open, I had easy access to my headphones.

One unforeseen consequence of the simpler setup is that I don’t really have a reason to open the drawer as there aren’t really any controls to fiddle with. This means that the audio interface and more importantly, the headphone jack permanently stays inside the drawer. This means that, during calls, I need to keep the drawer open just a smidge so I can snake the headphone connector into the drawer. This annoyance is made worse by the soft-close drawer slides that I opted not to disable.

Ultimately, I needed to find a way to pass the headphone jack/cable through the drawer. Using a 1/4” panel mount, I did just that, and it’s glorious.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 2 - Right cabinet

The right cabinet follows the same process: rebuild the cabinet, panel by panel. Of the two (original) cabinets, this one was much more wobbly — when I took it apart, I realized that the plywood was bowed when I made the cabinet originally.

This time around, I was much more careful with material selection and taking my time with cuts so that problem mostly goes away. The only notable change was relocating the cutout for the subwoofer to the correct side and then adding a small cable access cut-out.

View fullsize

View fullsize

View fullsize

View fullsize

To give myself more breathing room, I removed the “work laptop drawer” and redistributed the space amongst three drawers. More on the laptop situation in a bit…

[Back to Top]

Step 3 - Accessory panels

One of the best things I added on the Mk II was an inline power switch so I can easily turn on/off a pedestal fan I have pointed at me. I remembered how it took me a few tries to get the placement of the switch right (close enough to be convenient, but without constantly bumping it with the arm rest). This meant there were a … few screw holes added as I did trial and error. With this retrofit, I opted to use a sacrificial plate: I can make as many trial and error holes as needed on the accessory plate but that plate would only put two screw holes into the cabinet itself. I added an accessory plate to both sides of each cabinet for a total of four.

View fullsize

View fullsize

View fullsize

View fullsize

This brings me to the last accessory add-on: a laptop holder. By moving the laptop to the outside of the desk, it becomes easier to get to (for docking or removing) but also it’s easier to route wiring too. For the last mile, I use two magnetic attachments:

1. A magnetic 20-pin USB-C adapter (you should be able to use any 20-pin model) to hook up a random USB-hub

2. A magnetic 20-pin HDMI adapter paired with a virtual HDMI dongle. I used an ultra-thin HDMI extension to that I could easily keep everything nearby

With this configuration, I can just pull the laptop away from the dock when it comes time to get ready to go in the office.

View fullsize

View fullsize

View fullsize

Why the separate HDMI dongle? Why not just plug that into the USB hub?

My [non-Thunderbolt] hub has a HDMI port and you would think that I could plug the virtual-monitor dongle into that HDMI port. It turns out though, that the hub’s HDMI port only works as an additional monitor and not as the only monitor. I suspect this is a nuance of using non-Thunderbolt hub. For the way I use this laptop (lid-closed, single-display-only), the only viable option is to use the laptop’s built-in HDMI port.

The bottom line is, laptops are stupid, ha.

Jokes aside, the real problem is the ridiculousness of using a laptop as a faux-desktop (as someone who doesn’t travel for work).

[Back to Top]

Step 4 - USB power

I added a double-laminated beam of plywood at the back of my desk to push my desktop surface out away from the wall a bit (effectively pushing back away from the monitors a tiny bit). By “splitting the desktop surface up” into two pieces, I still retain the ability to remove the ‘main’ desktop for easy access underneath but still have a surface of some sort that would have a degree of permanence - I could easily remove the beam but it is so low profile that there is often no need to.

I added a flush mounted USB charger — now I can run my magnetic keyboard and mouse pad charging cables from this location rather than snaking it them from behind the monitors. I don’t currently have any other plans, but I could use this beam as a way to mount buttons, switches and controls for future projects (I briefly debated putting my 1/4 headphone jack mounted on here).

View fullsize

View fullsize

[Back to Top]

Step 5 - Speaker mount fixes

My rear surround speaker mounts were wobbly as hell since I was transitioning using a 1/4” threaded insert (in cheap plywood) to transition a 3/8” stud from the microphone mount. There simply wasn’t enough material thickness to be sturdy.

To retrofit this, I used the stud and nut to secure a flat piece of steel to the microphone mount. Then, I secured my speaker mount to the steel. Everything is super rock solid now.

I also took the time to tidy up the speaker wire allocated for the rear speakers so they aren’t draped everywhere anymore.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Milestone

This was intended to be a very quick in-between project, specifically to tackle some usability annoyances with my Mk II desk:

• Address the racking that was happening with the cabinets due to insufficient stiffness

• Make accessing the back of the cabinet easier (cutouts, lights)

• Make my ‘mixer drawer’ less of an annoyance to work with

• Pull the work laptop out of the drawer and make the dock-process less burdensome

• Redesign the rear-speaker mounts before they fall off

The most tedious part was the initial cabinet rebuild: taking everything apart and configuring a makeshift desk on sawhorses for a couple days (something you can do when you have a floating plywood top surface I guess). For the time being, these enhancements address all the small day-to-day nuisances and inefficiencies of my desk.

I have some [crazy] ideas for the Mk III desk but I have a lot of things to figure out before I get that far (and I have to save up for it). For the time being, the Mk II+ desk is serving me very well.

[Back to Top]

Appendix - Parts list

Audio

• XLR Audio interface: M-Track Duo. I went with the Duo (vs the slightly cheaper Solo) because I wanted the dedicated monitoring-volume control and I wanted to use the bigger (more robust) 1/4 audio-jacks where possible

• Headphones: Audio-technica ATH-M40X. Audio-wise, I’m still quite happy with my Sony MDR-7506, but I had a situation where I wasn’t sure if the plug-connector on my headphones were being flaky or if it was the headphone-jack on my mixer (this is what prompted the switch to a simpler audio interface). I went with the ATH-M40x as it was a comparable pair of headphones with an easily replaceable cable.

• Right angle connectors: I use both a 1/4 right-angle short stereo-extension and right-angle XLR adapter since I had them handy, just to keep everything in the drawer neat and tidy (and laying flat, without risk of tangling).

• Pass-through audio-jack: I went with a 1/4” panel mount since this was a “high contact” connection (and wanted the more robust plug). To reduce how much the headphone plug juts out, I used a right-angle adapter.

Laptop connectivity

• USB connection: The star of the show is a magnetic 20-pin USB-C adapter which I have connected to a random hub. For me the hub selection is entirely based on the location (and orientation) of connections since I will be ‘permanently’ mounting the hub in place. I also prioritized things like not having a built-in SD card reader since it was something I didn’t want

• HDMI connection: I used a magnetic 20-pin HDMI adapter paired with a virtual HDMI dongle. I used an ultra-thin HDMI extension. The thin (and short) HDMI extension has the right amount of flex and stiffness that I can secure the cable in place

Other

• Power switch: My favorite part of the desk is an inline power switch in easy reach that allows me to toggle a fan

• Under-desk light: I picked up a cheap light bar set (I used the other light on Nicole’s desk). If you’re in Canada, there’s an even cheaper option available although you’d have to rig up a way to mount it

• USB power: I used a simple flush-mount USB charger as I’m only really using this for charging my mouse-pad and keyboard — not exactly high draw devices. I use magnetic USB cables to hook everything up

• Floor RGB: I reused one of my old RGB strips for one of the leveled floor platforms and bought a cheapo strip for the other. I’m running everything at very-dim so I don’t need anything fancy

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Where we started

• Goals and Objectives

• Rough Design

• Step 1 - Prototyping a base

• Step 2 - Outfeed support

  • Top attachment point

  • Bottom attachment point

  • Main attachment panel

  • A splash of colour

  • Mounting the folding arms

• Milestone

• Mixups, surprises and lessons learned

Where we started

After the last round of upgrades (and a few upgrades since then), I’m rapidly approaching the end of the line terms of small, easy improvements I can add to the saw.

View fullsize

View fullsize

[Back to Top]

Goals and Objectives

I originally set out to do two things:

1. Stop the integrated mobile base from interfering with the dust hose/fittings

2. Add some kind of outfeed support

For the hose/mobile-base issue, the problem arises because the crossbar of the (connecting the front and back halves of the mobile base to a common actuation lever) is directly inline with the opening for the dust port. Technically, everything fits — but just barely — if you tap anything funny, the hose will just pop off (and it can definitely happen mid-cut). I can’t just clamp the hose to the fitting because this is the hose I disconnect to use on my bandsaw, jointer, planer etc. When I started (foreshadowing), I had some additional vague requirements for the replacement mobile base:

• The current integrated base, when transitioning from ‘up’ to ‘down’, is a very jarring ker-chunk process. While I’m sure the table saw can handle it, I would like to reduce this if possible

• Ideally, I wanted to level the saw (at the least, in the ‘down’ state)

• I wanted to add additional shelving underneath the right wing and also, on the back side of the saw (which would then become the outfeed support)

• I wanted to keep the table saw as close to the current altitude as possible (i.e., reduce the amount of lift that any platform, wheels, stands might add)

• Although I don’t do it often, I wanted to keep the transition between ‘up’ and ‘down’ as simple as possible

• I wanted to reduce any chance of ‘stubbing’ the wheels, feet or mechanisms

The motivation for the outfeed support is pretty straightforward: to be able to cut longer things without it falling off the table and/or bugging my very gracious wife to act has a human outfeed support.

[Back to Top]

Rough Design

Since I wanted to the outfeed to be connected to (and move with) the saw, I needed to tackle the mobile base first. Originally, I wanted to leave the integrated mobile base intact: I would leave the table-saw in the ‘down’ position and place the entire saw on a platform that would then go up and down. My first order of business was to figure out how to handle supporting, levelling and moving the saw would look (i.e., “the feet/wheels”). There were three archetypes I was considering:

1. Popup casters (like this).

   • $80-ish

   • 800-lb (ish) weight capacity

   • This is the simplest option: some models allow you to remove the entire wheel mechanism (when in the ‘down’ state), completely eliminating any chance of stubbing anything

   • Downside: to way to level the saw (in the ‘down’ state), I would have to figure out a way to do this

   • Downside: in almost all scenarios, transitioning between ‘up’ and ‘down’ would result in a substantial ker-chunk

   • Potential downside: most of these models (and particularly ones with removable plates), rely on two screws attach to the base. Even if I found a model using four screws (or bolts), that feels like a ton of torquing force applied to each foot.

2. Levelling casters (like this)

   • $60-ish

   • 2200-lb (ish) weight capacity

   • This single unit contains a levelling wheel and caster all in one, all in a fairly tight footprint

   • There’s no ker-chunk when transitioning between

   • Downside: it’s possible for ratcheting models to accidentally ratchet while moving the saw around (causing the levelling to unintentionally adjust). There are models without the built-in ratcheting mechanism that don’t have this problem but that is even more tedious — now you have to use a wrench on each wheel to transition

   • Downside: it’s tedious to transition between states, you have to manually adjust each foot

   • Potential downside: most models have a maximum height adjustment span of about 3/4”

3. Floor locks (like this) paired with dedicated castors

   • $130-ish (for the floor locks) + $60-ish for casters to pair with it

   • 800-lb (ish) capacity

   • The floor lock is exceptionally beefy and pressing on the go-up/go-down lever is super easy

   • There is a slight amount of give in the floor lock (perhaps 1/4” compression, when weight is loaded on it — more on this later)

   • It’s possible to reduce the amount of ker-chunk to near zero: by keeping the height-gap between the extended-floor lock and the wheel to a minimum. In this case, the wheel would be smaller

   • Downside: there isn’t really much of a way to level anything

   • Potential downside: it is possible to unintentionally trigger a transition through accidental tap

   • Potential downside: since you have four locks and wheels, you have twice as many things that you can potentially stub

4. Buy a mobile base (like this)

   • $360(!!) (plus I know down the road, I’d sink another $200 into the extension)

   • 1500-lb (ish) capacity

   • Off the shelf solution that has mobility, levelling and isn’t tedious to transition between

   • There is a cheaper model that also features all-swivelling wheels, but this option doesn’t allow me to add an extension of any sort

My original plan was to just get the mobile base and move on with my life, but the significant increase in cost made me go down the rabbit hole of exploring options. As an alternative, II originally set my mind on Option 3. For the last ten years, I haven’t had a levelled saw, so I figured, maybe I can live without that (although I did come with an idea to address this)

View fullsize

View fullsize

[Back to Top]

Step 1 - Prototyping a base

I opted to go with a 58x40 base: left-to-right, that would not overhang the saw (reducing any chance of stubbing as I walk around it), but on the back side that would give me 8-12” of additional room to build up an outfeed support.

At one point, I debated building out the lower support box — this would allow me to integrate levellers between the main body and the lower support, but all of this is still a prototype that may or may not work, so I didn’t want to go down that rabbit hole. If this worked out, I would entertain the option of rebuilding the lower box (with an integrated dust port that would make connecting the saw so much easier)

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Next is to make the base platform: a sheet of plywood with a couple 2x4s running across to stiffen it

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Last is to attach the double-laminated ply and do a test run of the floor locks…

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

And this was a total bust.

• The floor locks did their part and can handle the weight

• The 2x4 did it’s job: there’s very little flex left-to-right. Surprisingly, there is only a little bit of flex front-to-back

The reason this fails is because there’s just not enough overall lift (and/or my tolerances were too tight) — I had about 1/4” to work with (which was what I wanted), but there’s

• Flex in the base

• The slope of the garage

• The 1/4” sponginess in the floor locks when weight is applied

So in the end, I’ve got maybe 1/16th of clearance across the floor — and this is on a brand new build with no long-term sag. I had some options if I wanted to salvage this:

• I could try and put some 2x4s front-to-back to stiffen that up (— I would have to plan that carefully not to interfere with the dust hose)

• Using another sheet of plywood and some scraps, I could turn the entire base into a torsion box that would virtually eliminate the sponginess. But this would now start to bump the height of the floor closer to 2”

• I could compensate by building a custom base entirely (which would let me incorporate levelling as well)

All of these were unknowns that I wasn’t in the mood to explore. Although this was a dead end, I’m super glad I explored it — if I hadn’t, I would have second guessed myself for the next decade if I could have saved a bunch of money by building things out myself.

[Back to Top]

Step 2 - Outfeed support

Top attachment point

With that out of the way, I bought the Bora PM-3550 mobile base and installed that according to the instructions. Although this was mostly straightforward, it was a hell of a pain to lift my table saw into and onto the mobile base, solo. I eventually got the saw in place, dialed in (and levelled!). So time to move onto the real project.

First, I needed to create a mounting point on the back of the saw running left-right. I was able to make use of the bolts holding the rear-fence rail to the table-saw body to hang some angle-steel to make a shelf.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Bottom attachment point

Next is to find a way to make an attachment point at the bottom - I wanted to leverage the mobile base (that I paid dearly for, ha). This way, the outfeed automatically moves with the saw and I don’t have to do any fiddling.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Main attachment panel

With the vertical plate in place, I can add a removable panel that I can use to attach our folding outfeed to.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

A splash of color

Now is a good time to do a quick sand and stain before I can’t access areas behind the folding shelf supports. I was able to get some better pictures of how everything fits together as well.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Mounting the folding arms

The folding arms I bought work but they are really finicky to install. When the back side is fully secured down, you can still swing the entire arm side to side by 3/4” or so. Which means it’s really hard to get a pair of them aligned. Also note: the back screw holes are big enough for 1/4 bolts, but the top ones are not.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Milestone

The primary objective of this project was to eliminate connectivity issues with the dust hose and to have an folding outfeed support and despite the struggles and setbacks, this was ultimately a success. I did some soft-touches on the outfeed support (sanded the top to 800-grit and rounded over the bottom exterior edges).

After getting everything together, I did a thorough clean and tune-up of the saw, so we’re good to rock for the next project!

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Mixups, surprises and lessons learned

Although the original idea for the mobile base was a flop, I’m glad I tried it out — I think it could have totally worked, but this would turn into a much bigger project — essentially integrating a table saw into a mobile work station essentially, and that wasn’t what I wanted to do.

The folding arms though were a bit of a mess to get installed. It’s super hard to describe but the top panel kept moving diagonally when they shouldn’t have moved at all (or if anything, vertically). I suspect this was due to the amount of play that is in the folding arms (which exists until you get all the screws down), so it’s just a one-time frustration.

I didn’t expect the folding arms to open 90-degrees (some models advertise that they can — I didn’t have time to sift through the options when I ordered). I briefly considered adding a tapered shim to the under side of the folding top but that would require removing it… and then more frustration when putting it back together (using the same screw holes) … and it doesn’t close properly.

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Where we started

• Rough Design

• Step 1 - Installing the umbrella stands

• Step 2 - Mounting the chain-link

• Step 3 - Drip irrigation planning

• Step 4 - Drip irrigation hookup

• Milestone

• Mixups, surprises and lessons learned

• Appendix - Parts!

Where we started

The previous growing season highlighted two problematic areas:

1. Watering:

   • The cheap soaker hoses failed (which meant we had to resort to manual watering) however,

   • The super-leafy nature of zucchini meant that it was really hard to [easily] get water to the base/roots

   • We often had some standalone planters out in front of the mega-planters (say, for some flowers) — there was no way to integrate the automatic watering for these, so we’d have to manually water them anyways

2. Plant supports in general: this was a rabbit hole of problems.

   • The initial problem was that the electric fence posts were not nearly as sturdy as I thought they would be or rather, I grossly underestimated the loading that our tomatoes would exert (mostly because I had never had tomatoes survive long enough to be problematic before)

   • As a stop-gap, I bought some heavy-duty stake options. Individually, these were great — this solved the flexing problem, but this was an O(n) solution to support all of the tomatoes.

   • I thought I’d be clever by running some taut lines between the stakes and then supporting the tomatoes across the span. Obviously that was going to sag as there was no easy way to keep the two end-stakes from caving in on each other (I wasn’t about tether things down with guy-wires)

View fullsize

View fullsize

[Back to Top]

Rough Design

Watering

I think the root cause of the soaker-hose failure was running it without any pressure reduction: measuring the pressure at the final fitting was giving me 50-60psi - nothing on the packaging of the soaker hose suggested that it needed a pressure reduction.

This is an easy fix: buy a pressure regular to bring the pressure down to 25psi. In theory, by getting the pressure down, this should keep the soaker hoses from failing in the future. I had bought an extra set of flat soaker hoses so I could put those in the planters and be done with it, but I was never really happy with watering pattern. There was no elegant way to get the coverage I needed with a soaker hose (and especially a flat hose) — they just don’t do twists and turns well.

I debated going one step further: switching to round soaker hoses and then using fittings to do the turns. Getting a kit was fairly affordable, but it was the fittings that were the death of me: they were expensive, hard to find, and I couldn’t be sure if they would fit. This really, was the kick in the pants I needed to reconsider drip-irrigation.

When I had previously looked at (and previously used) drip-irrigation, I was turned-off by the pain it was to work with: having to jam a dinky 1/4” hose onto a barbed fitting was not fun. Looking at this from a “I may have to reconfigure the watering for next season” (and thinking about the aggravating labour) was a no-go for me.

Thankfully, I found that you can get push-connect fittings; and with hose being cheap, I would be less bothered if I had to reconfigure my watering layout next season.

As a final just-in-case, I bought a small articulating watering wand should I have a need to have to manually water something. The wand is handy because last year, I discovered that zucchini is prickly.

Plant Supports

I had the opportunity to buy some 4ft (1/2” diameter) and 6ft (1” diameter) fibreglass stakes made by Miracle Gro at surplus prices. Compared to my bulk 4ft (1/4” diameter) fibreglass stakes, these are definitely tanky stakes. Definitely good to have for some individual plant supports (say, our cantaloupe or watermelon). But I wanted something more MinMax for the tomatoes.

The beefiness of the T-stakes was sufficient, the challenge really, was keeping the two posts from collapsing in on each other.

Simply put, if there was a way to jam a horizontal bar at the top, between the two T-posts, this would keep the posts from collapsing on each other. The tricky part is how to actually make that stretcher; I had some options

• Obviously, if welding was something I could do, this problem would mostly go away but I’m working with my skill-set

• Take a large diameter PVC or ABS pipe, drill a hole at both ends, on one side only — this would allow the pipe to sit on top of the T-post, without falling through. Using some strapping, we can probably strap the pipe down onto the T-post. The problem with this is that the entire structure is dependant on the pipe not warping/sagging due to heat from sun-exposure (stuffing cut-down lumber into the pipe should nearly eliminate the sag). In theory, this solution would work but I wouldn’t know until I built and tested it — and large diameter piping is expensive

• Using some end-rail t-clamps we could probably rig up something using EMT or even another T-post

• As I was writing up this post, I discovered that T-post brackets are a thing and I think this, with some pressure treated wood, would be great. I could even put in four T-posts and build a cage of sorts

• I briefly considered working with cattle panel and then said hell-no

Whatever solution I came up with had an additional constraint: it must be done in such a way that it can be dismantled so that I can tarp over the planters in the winter).

The solution I came up with was to just build a one-segment chain link fence:

• The fencing hardware is already built for outdoors

• Parts the mostly standardized

• The chain-link portion could act as the support lattice structure for the plants as they grew taller

The only tricky part that remains is figuring out how to disassemble this for winterizing the planters. For that, I needed to add a socket into the planter box essentially. Although potentially an eyesore for some, a couple of 5-gallon buckets with concrete could totally solve this problem: the fence would be “outside” of the planter-box, so the winterization effort would be negligible.

I had a thought that perhaps umbrella stands could fit the fence posts: some umbrella stands will support 1.9”/2” diameter umbrellas, and my main-post was 1-7/8” (1.875”) — close enough that I wanted to triple check with my own eyes before committing to it (spoiler: it fits).

[Back to Top]

Step 1 - Installing the umbrella stands

Moving away from using a T-stake means that I have to do some more thinking about how to get the umbrella stand into the planter box — without losing a ton of volume/surface area to the umbrella stand. There are a few options:

• A ‘normal’ umbrella stand (ballpark 18” diameter)

  • Roughly 255 sq. inch footprint

  • Fence post is minimum ~9” inset from the edge of the planter

  • Roughly $50 each for me

• A ‘half’ umbrella stand

  • Roughly 122 sq. inch footprint

  • If you place the umbrella stand smart, you can have the fence post only a couple inches from the edge of the planter

  • Roughly $60 each for me

• A ‘deck mount’ umbrella stand. This is just the ‘socket’ piece — I would have to bolt this to ‘something’ and bury that. The advantage is that I get to control what I attach it to. I debated going with a plastic cutting board but ended up going with brick because I already had extras on hand.

  • Variable footprint

  • Depending on where you mount it (relative to the base) you can have the fence post only a couple inches from the edge of the planter. Also by adjusting where and how you mount it, you can better control which axis the base resists tipping on

  • Roughly $30 each for me. There is an additional cost of mounting hardware and the brick (or cutting board etc.) as well

For cost savings, I ultimately went with the deck mount option.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

One thing that I will have to keep my eye on is the bolt used to secure the umbrella/post: “the thread replacement screw is made of high-quality black plastic and stainless steel” — but after only a few days, it’s showing signs of rust.

[Back to Top]

Step 2 - Mounting the chainlink

While you can do this solo, don’t. Having an extra set of hands makes this so much less stressful. Even with an extra set of hands (and a couple extra ladders), this was still a handful. Maybe there’s a good reason people don’t build chain-link fences suspended in the air? I wasn’t able to capture the individual steps because we simply had our hands totally full. You can get the gist of what needs to be done on YouTube. As a highlight though:

• Measure the span between the end-post-assemblies (measure from the inside-of-the-cup) and using a grinder, cut the top-rail to the desired size. We did this in a few passes, slowly getting to the final size. A cordless bandsaw would have made this so much easier (I don’t have one - yet) but after a couple test cuts, we got some pretty clean results: I head the grinder steady (using the guard a a stable reference edge) and Nicole slowly rotated the rail

• After getting the size right, I hit the cutoff end of the top-rail with a flap disc to clean it up and sprayed a bit of black Plastidip to try and protect it

• We got super lucky in that we had an extra length of chain-link fence (this is referred to as the ‘fabric’) available and I didn’t have to buy any — it would have been $180 for a 50ft span (!!)

• To size it, Nicole held one end of the fabric against the main-post as firmly as she could and I stretched it out to the other end, pulling against her until I could only hand-squeeze the diamond grids about 1/4”. We made note of which link it was and unravelled the fabric at that link

• With Nicole holding the span of fabric midair, I struggled to get the tension-band installed. We first did two tension bands on one side first (just to get it attached), then we randomly ran fabric-ties to help take the weight before then moving to the far side and struggling to get the tension band on that side. This is particularly challenging because we simultaneously need to pull against the main post (in order to tension it) but we couldn’t really pull too hard because the post isn’t really “anchored”. Again, this is probably why people don’t built fences suspended midair, inside planter boxes…

• We then looked up how to actually tie the fabric ties and then tied every other link since we are planning on weight-loading the chain-link with tomatoes (and really, the ties come in a pack of 50)

• I debated running a tension-wire on the bottom but decided to use a top-bar there as well, essentially turning this into a big frame. There is a hope that come end of season, we can ‘lift the entire thing out’ (without having to disassemble anything) and then in the spring, we can “just put it back in” — I’m not about to take it out now to find out. Again I ran fabric ties on nearly every other link across the bottom as well

View fullsize

View fullsize

[Back to Top]

Step 3 - Drip irrigation planning

The biggest advantage of going with drip irrigation over a soaker hose is that I can add random drops at will, if needed. When picking out parts, I specifically wanted to start with a kit and then add more stuff to it (even overlapping). This lets me build out extra parts and try new things if needed without having to worry about running out of parts (and they are fairly cheap anyways — at least the entry level cheapo stuff I bought).

A lot of the videos and guides I’ve seen have all of the ‘configuration’ hooked up at the hose-bib. Doing things this way has a lot of advantages:

• Since everything is in series, you only need to buy one of everything

• Hiding a black 1/2” mainline hose is infinitely easier to do than a bright green garden hose (and substantially cheaper). I would also get to ‘reclaim’ my garden hose

• With strategic use of tees, elbows and shutoffs, I can have control per-planter

View fullsize

View fullsize

However, I did not go with this method. Instead, I kept all of my existing runs and made the drip-irrigation changes after the 4-way-manifold.

View fullsize

View fullsize

This has the additional expense of having to buy +2 pressure regulators and hose connection assemblies but it gives me a couple advantages:

1. Reclaiming the garden hose isn’t a priority since I already bought another one anyways — the hose that connects to the last port on the manifold has enough reach to allow me to run the sprinkler in the yard if needed (and it is certainly easier to ‘stuff the hose back in a milk crate’ than it would be to neatly spool the hose back at the house)

2. More importantly, by doing things this way, I have the option of swapping over to soaker hoses should this not work out in the long run (and if it does work out, I’m only out the cost of a couple pressure regulators)

I suppose, I could have done a combination of the two, trading the cost of two additional regulators and hose connectors vs the cost of tees, elbows and shutoffs:

If I like the drip irrigation (preliminary opinion suggests that I will), I can certainly revisit the plumbing of the planters next year. For the time being, by doing things the way I did, should things not work out — I can very easily transition back to using soaker hoses (now with pressure regulators in place).

[Back to Top]

Step 4 - Drip irrigation hookup

Going with push-connect everything made this super relaxing to work with: eyeball, cut, push, repeat. If I end up staying with this, I may get some 1/4 elbows to reduce some of the arcing of the 1/4 tube but for this season, it’s fine.

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Milestone

This was intended to be a very quick project — and it was. Everything came together in a matter of a couple of days (I spent most of the time waiting for packages to arrive more than anything). Now that I’m done (for now), it’s super satisfying to step back and see how everything works together.

As the plants get bigger, I’ll have to dial up the amount of water being pushed out (or run the timer for longer — but that assumes all of the plants grow at the same pace), but for now, this is in a good state.

Assuming that everything works out as well as they should (on paper), I have some plans for the next refresh

1. Revisiting the plumbing between the house and the planters — you can see the bright green hose in the background ha

2. Going down the rabbit-hole of tying this into home automation — soil moisture sensors and local-first water controllers

3. Right now I don’t have a filter in place — because I didn’t feel like buying three filters (and didn’t know if putting it in-series, above the 4-way-manifold would be problematic). When I revisit, I’ll probably add a filter.

Exciting!

[Back to Top]

Mixups, surprises and lessons learned

This project mostly went without a hitch but there are a couple small things of note:

1. While I am super happy about how the chain link fence turned out, had I known about T-post brackets, I think I totally would have gone down that route. While it certainly would be a ton easier to build that, I think I would lose out on the versatility of the chain-link fence — you can see how the dinky 4-ft stakes are braced into the fence above, making them super sturdy.

2. One thing that will be problematic is the water timer: I have a mechanical timer which works great — if you need a ‘big chunky time span’ (the tick mark is at 15-minutes). But shorter bursts of water (say 5-minutes) don’t work so good: the spring doesn’t even engage. I’ll need to switch to a digital (or at least, non-spring-driven) timer to improve upon this.

[Back to Top]

Appendix - Parts!

Chain Link Fence

The parts generally come in ‘black’ and a grey galvanized look — I went with the black as it fit better aesthetically (although it also is the more durable option). Since I’m only building literally one panel, the cost difference doesn’t matter.

• Fence

  • $22 each - Two 7-1/2 ft, 1-7/8 main post

  • $24 each - Two 10-1//4 ft, 1-1/4 top rail

  • $2 each - Four rail-end-assemblies

  • $3 each - Two tension bars for 48” fabric

  • $1 each - Six 1-7/8 tension bands (you need at least three tension bars for 4ft fabric and four tension bars for 5ft fabric)

  • $1 each - Two 1-7/8 post caps

  • $9 each - Pack of fencing ties

  • $128 each - 4ft x 50ft chain link fabric (Free for us since we had enough on hand)

• Umbrella stand base

  • $30 each - Two deck mount umbrella stands

  • $22 each - Pack of 1/4 x 1-3/4 stainless Tapcons

  • $7 each - Pack of stainless 1/4 washers

  • $7 each - Two Oldcastle Townsend wall blocks (Free for us since we had them on hand)

Total cost: $240 (for us)

If we hadn’t, by sheer luck, had the chain link fabric on hand, I would have opted to go with the taller 5ft fabric and I would have reached out to fencing companies to see if I could get a measly 6.5ft span of chain-link fabric instead of having to buy a whole roll. Alternatively, I would have considered

• $37 - PVC lattice

• $48 - 16ga welded rabbit fence

• $70 - 12.5ga welded fence

Drip Irrigation

I probably could have saved some money going with just the kit and/or a different parts

• $65 - A starter kit that used push-connect fittings

• $18 - 1/2” mainline to 1/4” 4-port tee

• $23 - 1/2” mainline to 1/4” 2-port tee

• $17 - 1/2” mainline to 3/4” garden hose adapter (my kit came with one, but I needed two more)

• $15 each - Three 25-psi pressure regulators

• $24 - 1/2” mainline elbow

• $28 - 1/4” in-line shut-off

• $10 - 1/2” mainline end-cap

• $8 - Back-flow preventer (not convinced I need this, but oh well, it was cheap and my topology allowed me to get just one)

• $22 - Bulk pack adjustable pass-through emitters

• $35 - Bulk pack of adjustable emitters

Total cost: $330 (with lots of extra parts)

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Where we started

• Goals and Objectives

• Rough Design

• Step 1 - Phone mount

• Step 2 - Dismantle the headboard

• Step 3 - Cubby storage

• Step 4 - Center divider

• Step 5 - Adding the switch

• Step 6 - Adding the slats

• Step 7 - Finishing touches

• Milestone

• Mixups, surprises and lessons learned

Where we started

Overall, my bed-frame has been great:

• Using glue (where practical) means that it hasn’t creaked at all in the last three years

• Almost surprisingly, the cheap tablet mount still works great

• Increasing the size of the sub-frame (and/or the overall weight increase) means that we haven’t over-cantilevered the bed-frame

[Back to Top]

Goals and Objectives

I have three broad aspects I wanted to tackle:

1. I wanted to find a nicer way to mount (and charge) our phones (and watches), specifically without having them sprawled on the nightstand — which inevitably leads to knocking them onto the floor

2. I wanted to ‘clean things up’ a bit more so that I could blindly schedule a robovac to regularly clean the room. This means that I would have to find a solution for the various cables spilling out from the sides of the bed as well as our super lightweight garbage cans that would get knocked over

3. I wanted to revisit storage; to take a step back and look at how we were using the night tables and see about improving things

I didn’t want this to be a huge sprawling enhancement project and I mostly wanted to work within whatever materials I happened to have on hand where possible. I also had to make sure the bed was “reasonably” bed-worthy each night, so I had to tackle things piecemeal.

[Back to Top]

Rough Design

I didn’t really enter this with a specific plan since I had forgotten the exact state of the headboard and nightstands that I had built. I knew that I had cut some corners when rushing through the final steps when making it, but I wasn’t sure how much I could reuse.

At the start of the project, my biggest concern was reconciling the nightstands and the garbage:

• I wanted to keep with floating nightstands — and I wanted to keep their top-surface roughly where they were (altitude-wise)

• I wanted to get the garbage off the floor

The problem is that I naturally gravitated towards keeping the garbage can directly below the night stands. So getting the garbage off the ground whilst keeping the nightstands mostly where they were would be impossible (or at the least, be very challenging) as they would intersect each other.

I did briefly consider something like a Yamazaki side-table-trash can (either buying it or building something similar) - this would allow me to still toss things in the garbage whilst keeping the night table roughly in the same place.

I decided that I would start by adding some storage cubbies behind the headboard: this would allow me to migrate all but the alarm-clocks off the nightstands and I would tackle this problem when I got to it.

[Back to Top]

Step 1 - Phone mount

We’ve had magnetic phone holders in our cars for awhile now without issue and I’ve rigged up a few DIY phone mounts using magnets and plywood in the past, so had the idea of adding a sliding magnetic phone mount to the headboard. This was a super easy enhancement: you have quite a bit of adjustability left-right but also up-down by changing which gap to thread the phone mount through.

Parts used:

• 130lb rare earth magnet. Using a 1-1/4” forstner bit, I set the drill press for 5/16” depth for a flush fit

• 1/4-20 thru-hole star knob (alternatively 1/4-20 washered wingnuts)

• 1/4-20 x 2-1/2 machine screw. I set the machine threw into a threaded insert to keep it bound to the plywood — one less moving part

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 2 - Disassemble the headboard

This was the proverbial ‘point of no return’: I removed the headboard and the attached nightstands and would go down the path of seeing how much of it was built reasonably-well versus what I would have rebuild/redesign. Thankfully, the main framing element, I built reasonably well so I didn’t have to revisit it.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

I didn’t stop to take photos but I also swapped out the old RGB strips for a WLED compatible 5m strip. While I certainly wasn’t unhappy with the old strip, the new strip is definitely smoother (60leds/m versus 30leds/m) and can go way brighter if I choose to. The biggest advantage is running locally off my home network and not being at the mercy of cloud connectivity and the latency that comes with that.

New WLED-compatible strips, viewed from far-away

New WLED-compatible strips, viewed from ‘over the bed’

I debated putting the new strips into channels and diffusers but opted not to since I didn’t really want to use up the few channels and diffusers I had left. The end result, at least in my bedroom and application, is perfectly fine without it. For the previous RGB strips, I had preinstalled clips, so I was able to reuse those.

Parts used

• WLED. I had previous experience with this strip specifically when making the LED slat wall for Skype Station

  • Controller. This controller generally ‘just works’ — there are some quirks around updating the firmware on it and then having it take it’s sweet time before the web-server stops timing out, but once things are all done, the controller is generally reliable

  • 5m RGB strip (60led/m). The density here is perfect for by application and I was able to wrap the three visible sides of the sub-frame with just the one strip

  • 10A max power supply. No complaints, this ‘just works’

[Back to Top]

Step 3 - Cubby storage

I vaguely recalled seeing some kind of ‘storage cubby behind the headboard’ in the past (future me tells me it is most likely based on the IKEA BIMNES or similar). Since I plan to have slats on my headboard, I needed a system that would work with that. I happened to have some leftover double-laminated plywood (ironically a carry-over from my very first bed-frame build) so I was able to use this to make extra heavy duty cubbies. At this point, my I had the following ‘rough plan’ (totally not to scale)

At this phase in the process I was mostly working with whatever scrap and leftover materials I had on hand. The idea was that I would have a cubby on each side that butted up against the vertical 2x4 and was solidly bolted to the mounting frame. With the top and bottom of the cubby box being made of double-laminated ply, this was beefy enough that I could then consider structurally mounting things to that — perhaps a cantilevered nightstand? To be determined…

In the middle, I wanted to add a panel so that nothing could really fall behind the bed and then perpendicular to that panel, I would attach a small shelf. The original intent for this shelf was to place books, kindles, tablet, flashlight etc. I would provide a way to manage cables to/from this shelf and provide power. Implementation to be determined…

But first, to build those cubbies.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Overall this phase is pretty low-drama and I eventually noticed that the gap between the cubby and the 2x4 was because I didn’t fully notch out the top and bottom (this was quickly corrected with an OMT).

[Back to Top]

Step 4 - Center divider

Was making fairly good progress here and I got lucky as I happened to have a perfectly sized piece of OSB I could use for the divider. I debated not using this as there would be a visual change going from plywood to OSB but it felt silly not to take advantage of this material that I would otherwise never use for anything else.

Generally speaking, this isn’t a huge (aesthetic) deal I suppose — I was more concerned about the texture and spent quite some time absolutely sanding it down to 220-grit.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 5 - Adding the switch

It was awesome that the original bed had switches to control lights but it was a bit annoying to awkwardly reach for the switch while in bed. Also — locating the switch where it was before was a constraint (or consequence, I’m not sure) of having the nightstands the way I did them originally. This was an opportunity to have a ‘clean slate’ and revisit things entirely.

I opted to put the switch in the center panel: since it’s located past the pillows, in theory, it shouldn’t be accidentally triggered and moving the switches there removes one more need for some kind of extension or night table. I briefly debated putting the switch on the side of the bed but I was afraid it would interfere with where the nightstand/garbage would go - having it on the center panel is so much cleaner as well.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

I’ve stopped at adding just the switch — the light box remains a dangling tag-along that I’ll need to mount to the slats — once I figure out what my plan is for the slats. At this point in time, the plan is:

At this phase of the build, I’m starting to consider the slats; I really want to remake the slats, but:

• [Pro] I can finally address the fact that when I made the bed originally, I sent screws in through the front (rather than from behind)

• [Pro] By re-cutting the slats, I have a bit more control over the sizing of the slats to better fit the space

• [Con] That seems like a silly waste of materials

• [Con] While it’s not challenging to do, it requires sinking more time into this mini-project that I really just want to be done and over with

• [Con] I really don’t want to spend the time to cut the relief needed to mount the lamps

I ultimately decided to go with the existing slats and ‘just see how it looks’ — I can always re-cut the slats, but for now, I wanted to press forward.

[Back to Top]

Step 6 - Adding the slats

This step gets a bit tricky (even though there’s not much to show for it) as I have to juggle both the unwieldy slats, figure out the spacing and take into consideration the center opening — I don’t want the slats to really intrude into that space, otherwise I wouldn’t be able to access it easily. The plan was to start from the top, with the upper soundbar shelf and then work my way down, using shims to maintain a consistent gap between slats. My wife was awesome and helped handle the slats and spacing while I waffled around and then attached them

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Getting the top shelf and then the top slat as properly as I could is super important: everything else registers off it! I ended up running pocket holes from the double-laminated top shelf into the topmost slat just to tie everything together tightly. We used plastic shims in order to dial in the gap between the slats; there were two things to keep in mind:

1. The gap had to be big enough for the magnetic phone holders to slide between

2. I had intended to use the gap in the slats as a faux tissue dispenser — the gap in front of the tissue box needed to accommodate that

From a distance it’s not as apparent, but the gap for the tissues is slightly bigger. I also took the time to finish mounting the lamps to the slats.

[Back to Top]

Step 7 - Finishing touches

This last section was a seemingly endless punch-list of little things to do and add. I spent the most amount of time doing cable management for all of the wires behind the bed, of which there are a ton:

• 2x for integrated power bars (on the side of the 2x6 mounting frame)

• 2x bed lamps on switches

• 1x for USB ports on the visual shield

• 2x for the sunrise alarm clock and Google Nest Mini

• 1x for heated blanket

• 1x bed motion lights

• 1x for WLED

• 2x for soundbar & subwoofer

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Parts used:

• Wall mounted garbage can. I had previously used this garbage can when making my Coffee & Tea bar and it worked great there

• 4-port powered USB hub. I went with a rectangle-style rather than the more common (and less annoying to install) round-style because I wanted a more low-profile unit

• Cheap tablet mount. Paired with a DIY magnetic mount, this works great

• Sunbeam heated blanket. With the design of the bed-frame, it’s super easy to just leave the wires and controller preinstalled and disconnect the power during off season

• Philips 3520 sunlight alarm. I bought this years ago to act as a silent alarm clock and it’s still kicking

• Power. I run all of the electronics on the bed through a UPS that I have under the bed. The bed-frame and all of the electronics sits on a extra large rug which means that all of the wiring ‘travels together’ (when I pull the bed off the wall) and I only have to worry about a single power cord (from the UPS to the wall)

  • Cyberpower ST900U standby UPS. Still going strong, this keeps everything connected running in the event of a power blip

  • Long power bar. This allows me to widely space everything out and keep things organized. While the UPS can handle the load, I opted to keep the heated blanket on surge-only — if there’s a power outage, losing bed-heat is not the end of the world.

[Back to Top]

Milestone

Overall, the (original) bed-frame has been great and hasn’t given me any major grief and while this set of small enhancements addresses a few small issues I’ve discovered along the way, truthfully, the primary motivator was to ‘get stuff off the ground’ so that I can regularly (and blindly) schedule robovac to do it’s thing.

The various enhancements have the end result of entirely removing the need for having night tables and with that out of the way, getting the garbage off the ground was a no-brainer. A bit of cable management later and the space is robovac friendly. I ultimately decided to cluster all the ‘stuff’ in one area and hopefully the robovac just treats that area as a obstructed blob instead of working it’s way in there and getting stuck (if that’s the case, I can set a no-go zone).

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

I did a trial run of the robovac (without doing any prep- or tidying beforehand) and it ran successfully without snagging onto anything. It’s time for the residents to move back in…

For a ‘quick’ project that was mostly dictated by what materials I had on hand, this was pretty successful. Getting rid of the nightstands and pushing everything ‘inward’ visually opens up the room a bit more as well.

As for the robovac that kicked this all off? I have a Eufy L60 SES assigned to upstairs duties.

[Back to Top]

Mixups, surprises and lessons learned

The biggest takeaway here is ‘spend the time the first time to do things right’. When I rushed through this originally, I knew that it was a bad idea to send screws through the front of the slats (doubly so, since I did things properly on my first bed). Even with relatively color-matched screws, it still bothers me. To a degree, the OSB bothers me a bit but much less so, ha!

I do slightly regret putting the pocket-holes on the visible face of the cubbies as that added a bit more (unnecessary) work to plug them up — spending a few more minutes thinking about this probably could have prevented this.

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Where we started

• Goals and Objectives

• Rough Design

• Step 1 - Mounting in a custom frame

• Step 2 - Test run, adding wheels

• Step 3 - Swapping the wheels

• Step 4 - Finishing touches

• Milestone

• Appendix

Where we started

I bought a Ryobi pressure washer (Model #RY141900, discontinued now) in 2015 when I saw it on sale for $40 new. Ever since opening it up, there were a few things that bothered me:

• The hose reel is just a means to store the hose — you had to fully remove the hose from the wheel to use it

• The soap dispenser never really worked that great

• The way the lance holder was setup, the lance added to the effective footprint of the pressure washer…

When I talk about the ‘effective footprint’, I mean that, when you look at the pressure washer at glance, you get a mental picture of roughly how much space it occupies (or rather, 'how much you expect it to occupy’). But then now try squeezing the pressure washer out of a tightly packed shed and you’ll find that things like the lance, or the wheels that stick past the frame catch onto every little thing.

Also, the entire design of the pressure washer cart is built around pulling — which works great as you pull the pressure washer from the shed, across the lawn, to the driveway.

But what about in the shed? If you want to tuck the awkwardly shaped pressure washer into an awkwardly shaped gap between tires, outdoor power equipment and bins of landscaping supplies? That (generally) requires pushing the power washer. And if your shed is a disaster, you’re going to be trying to push the cart while outstretched. Aggravating to say the least.

But for the price I paid, clearly, I can forgive a lot.

[Back to Top]

Goals and Objectives

I recently bought a short swivel pressure washer gun and compared to the medium lance the pressure washer originally came with, this has totally opened my eyes to a better ergonomic experience. This was the kick in the pants I needed. The three biggest things I wanted to address:

1. Make the cart more manoeuvrable, more push-friendly — even at the cost of making it a pain to pull across the lawn

2. When you look at it and get that mental snapshot of how much volume the cart occupies: that’s roughly how much volume it occupies — no random things jutting out

3. Fix the stupid hose reel situation: the pressure washer hose should just be permanently plumbed in

I originally set out with the idea of storing the pressure washer in the shed as I did before, I was open to keeping it in the shop if the manoeuvrability was improved enough.

[Back to Top]

Rough Design

When I started this project, I didn’t really have a rough design in mind — I was just playing around with whatever scrap materials I had on hand. The only two things I knew for sure was having a mounted hose reel and switching to four castors.

[Back to Top]

Step 1 - Mounting in a custom frame

At this phase, I wanted to get the pressure washer and the reel mounted as a singular object. I hadn’t put any wheels on it yet but I mentally planned on having the castors mounted for the vertical/storage orientation. The thinking was the vibration from the pump running might cause the cart (on wheels) to runaway.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Back to Top]

Step 2 - Test run, adding wheels

Here, I put some wheels on the cart (vertical mode) and everything behaved roughly as I expected it would: horizontal mode was super stable and vertical mode was sort of tipsy but manageable, without introducing additional footprint (from castor wheels spinning beyond the perimeter).

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 3 - Swapping the wheels

Ultimately, I didn’t like the tipsy nature of the wheels in the vertical configuration — it wasn’t worth risking tip-over; and the option of extending the wheelbase would change footprint in horizontal-mode. I opted to try putting the wheels on the horizontal-orientation just to see if vibration would be a problem.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 4 - Finishing touches

The bones of the pressure washer ‘rebuild’ is complete and I wanted to tackle the [power] cord management by making use of the original cord-wrap. I then had the crazy idea of making a gun holster so that I could disconnect the gun during storage so it wouldn’t be dangling around as a weight at the end of a hose.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Milestone

This project has mostly been a success, I set out to reduce the effective footprint and make everything so much more usable and I think I nailed it. The pressure hose reel makes everything so much nicer to use. It is less manoeuvrable on grass (as you would expect) but thankfully, I had a sidewalk that runs down the side of my house to the back so it’s a very minor inconvenience - the compact footprint means that I keep this in the shop now so I don’t have to lug it across the lawn.

View fullsize

View fullsize

[Back to Top]

Appendix

I used a bunch of quick connect fittings (both for the pressure and garden-hose sides). For the garden hose side, I used some cheapo fittings that happened to be a perfect fit with my existing Yardworks set. For the pressure side, I ended up walking into a Peavey Mart and Princess Auto with the parts in question and spent half an hour figuring out what magic combination of parts I needed to tie everything together.

• Powerfist Pressure washer reel. This is just a cheapo wheel that allows me to skip having to plug the pressure washer hose into the pressure washer each time. Also super handy to wind up the hose when I’m done

• McKillan’s Short pressure washer gun with swivel. Coming from the stock Ryobi medium-lance, this is a joy to use — less hand fatigue, better control and doesn’t get in the way all the damn time

• 4ft Pressure washer whop with swivel. This lets me connect the reel to the pressure washer directly. Depending on what connectors you have where (quick connects, threaded connections, male vs female, etc.) you’ll have to find what works for you

• 50ft Uberflex and 25ft Flexzilla hoses. Generally 50ft is enough for me to do everything I need so I don’t usually connect extra 25ft. Both are fantastic hoses and feature the same anti-kink construction

• Universal garden-hose swivel. This allows me to ease the entry angle for the garden hose feeding the pressure washer. In this application, I think the all-plastic build should be fine, but if you want a bit more peace of mind, you can get one with an with an all metal construction

• 6-in-1 pressure tip, this is just a cheapo on that has worked okay for me

• Measureman pressure washer gauge. Totally unnecessary, but cool to see the data

• Bed armor paint. I used this as it was available, but I’m not convinced it’s particularly tough — however, in my application, it may be sufficient.

. This i[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Where we started

• Goals and Objectives

• Rough Design

• Step 1 - Reclaiming the base

• Step 2 - Building the carcass

• Step 3 - Building the top

• Step 4 - Finishing the top

• Step 5 - Drawers

• Step 6 - Door

• Step 7 - Finishing touches

• Milestone

• Mixups, surprises and lessons learned

• Appendix

Where we started

A few years ago, I tried to build my own tool-chest of sorts (partially as a learning exercise). After moving to this house (and after I ended up getting a real rolling tool chest), we repurposed the tool-chest as a drink station of sorts. There was really nothing special about this setup: it was a bunch of drawers on a mobile base and the top was definitely not supported well so it was very sag prone.

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Goals and Objectives

In addition to the sag across the top (since I didn’t know to add bracing when I originally built it), there were three other things I wanted to address now that I have a moment to revisit this:

1. We recently got a new robovac (the Eufy X10) and this new unit is a bit taller than our outgoing model. It’s tall enough that it doesn’t fit underneath the cart anymore, so I definitely wanted to increase the height to ensure that the robovac can clean under there. The alternative would be to make a fixed base flush on the floor to limit crud building up under the coffee bar.

2. When we setup the drink station, we just threw a hodgepodge of stuff into the drawers - I wanted to make a more-focused use of the space

3. We use a water-boiler which is great because we have readily-available hot water. Since it’s not hooked up to a water line in any way, we have to refill it every few days using a pitcher. I really wanted to find a way to make this ‘better’

[Back to Top]

Rough Design

I originally planned on using a pair of stacking bins to store things like potatoes, onions, pop bottles etc. I would then allocate drawer space for all the ‘expected’ things:

• cups and mugs

• tea bags and tea supplies

• dried goods: sugar, whitener, hot chocolate, etc.

• storing all of the different racks and accessories for our toaster oven

• dog treats

• napkins and towels

View fullsize

View fullsize

This is already a ton of stuff to cram into that footprint. Of course, I wanted to do ‘one more thing’ — have some way to make the refill process of the water boiler a bit less tedious. I envisioned using a 5-gallon water jug as a reservoir and ‘somehow’ getting the water to the boiler. Somehow.

That somehow started off as one of the many commonly found cheapo 'desktop water dispensers’ but I ended up going for something a bit more robust — a plug-in water dispenser pump. Of course, the tricky part now, is ‘where do we put this 5-gallon jug’.

• If we put the jug on the outside, then I get to keep all drawers and storage I had originally hoped for — but now there’s a 5-gallon jug on the floor

• If we put it inside the coffee bar, then everything is cleaner but something has to go

In either case, this was a problem for Future Me.

[Back to Top]

Step 1 - Reclaiming the base

First step was to breakdown the existing cart and re-purpose the base: the base was reasonably flat, true and sturdy so there was no reason to rebuild that. I swapped out the wheels for 5” softer rolling ones. While I was looking at this, I cleaned up some of the uneven surfaces as it was relatively low-effort. I did have to extend the width of the cart ever so slightly, so I added some solid wood side strips. This brings my width out to an even 48”. I triple checked and the base was definitely 48”. #foreshadowing.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 2 - Building the carcass

Next up, I’m now building a back, two sides and a middle divider. For the most part, this process was suspiciously smooth.

• Break panels down with the tracksaw

• Pre-sand everything to 220-grit

• Mark where the screws go

• Glue, pins and screws

• Done

The back panel was 48” wide so it was super easy to cut to size. #foreshadowing.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Now about that foreshadowing: the plywood was actually 48-1/8 wide — not the 48 that I assumed it would be (and built the base for. You can’t really tell that the two sides hang over the edge ever so slightly unless you’re looking for it. But this was a rookie mistake — I should have double checked that measurement. Since there’s a center divider, I wasn’t concerned about any weight distribution issues (and the sides are still 95%+ supported anyways). Lesson learned for next time.

[Back to Top]

Step 3 - Building the top

Now that I’m rebuilding this coffee bar, I wanted to keep the aesthetic the same as my dining table; thankfully, I still had a bunch of walnut leftover from that build.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 4 - Finishing the top

The big challenger here was finding a way to increase the height of the faucet so that the faucet pours straight into the water boiler - and to make it pretty. I opted to make a walnut riser block to encapsulate the faucet I had bought and bring it to the necessary height. The challenge here was to ensure that I could service/replace the plumbing components if needed.

Making the riser

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Drilling & cutting the top

Now the scary part: I needed to make very permanent, risky cuts and holes in the laminate top. Making the cutout for the power strip was a lot more unnerving because I only had 15/64” (6mm) of trim overlap — not a lot. Thankfully this was all stress and no drama.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Mounting the faucet

The faucet I bought came with everything I needed and thankfully, at this point, everything was ‘plug and play’

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Mounting the pump

Now that the faucet is in place (and I did a quick plumbing test), I can mount the the pump inside the cabinet. To keep things simple, I opted to mount the pump to it’s own board and then screw the board to the back panel.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Finishing touches

I bought a pair of super sized silicone mats to keep the top deck safe as long as I could. Using an X-Acto knife, I made the cutouts for the faucet and power bar for a perfect fit.

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 5 - Drawers

So far, this has been the ‘easy stuff’. Moving onto drawers is something I dread as I’m not very good at it (need more practise really). This time around things were a lot smoother than they used to be — I didn’t have to re-drill mounting holes for the slides at all, so that’s huge progress. The biggest lessons to take away (at least from my experience) is:

• Don’t try and maximize every last millimeter of space with drawer — it’s OK that there’s breathing room between drawers

• I made the drawers first and then only after all of the drawers were in place, then, I made the drawer fronts

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 6 - Door

I threw together a quick plywood panel door to cover up the plumbing bits. I almost screwed up when it came to mounting the cabinet portion of the euro hinge (due to inexperience). I was just going to slide the hinge in place, with a 1/8” gap and YOLO it. Thankfully, I had done a test door hinge installation and I needed ~1.5” gap (between the edge of the open door and the edge of the cabinet)

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 7 - Finishing touches

There were a few small quality of life touches I added to the coffee bar.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Milestone

Once done, this was a fairly satisfying project — I had set out to make a better-built cart that holds all my tea & coffee bits as well as housing all of the stuff I need for the toaster oven in one place.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

I’m done for now — maybe one day we’ll get a water line run here; theoretically, that should be a very easy connection to the faucet

[Back to Top]

Mixups, surprises and lessons learned

Shockingly this was a fairly uneventful build with only a few screw ups

1. I assumed the plywood back would be 48” wide and sized the base for this super exact size when in fact, it was 48-1/8”

2. When cutting the opening for the power outlet on the back panel, I should have cut this bigger to accommodate sliding the cart left-right if needed. I expanded the existing opening and thankfully it was low-drama

3. The door appears darker — it’s not just a trick of the camera, it’s visible to the eye too. This is because the door was an afterthought — by the time I got to it, I cut the door from a different sheet of plywood versus the drawer fronts

4. I have a power bar on the inside of the cart: I have the power for the pump, toaster oven and the LED light connected to it. The latter two require passing a cable and plug through the back of the cart — I forgot to drill a hole when I had easy access to everything and had to do that afterwards

5. Speaking of power bars, I would have also considered using a double-sided power bar to keep the cabling inside the cabinet tidier

6. In hindsight, I probably should have figured out what the thread-pitch was for the faucet and put a nut on there to keep the screw clamp from backing out. I still may do this if it bugs me enough (or if it loosens).

[Back to Top]

Appendix

Tech

• LED Sign: this one. I specifically wanted one that included tea and would power on when power was provided (without having to push a button)

• Smart switch: TP-Link KASA KP125M. The Kasa ecosystem is decent. So far definitely has failed-less than SmartLife.

• Power bar: 4 LEAF recessed power bar. I’ve used this one before without complaints.

Appliances

• Toaster oven: Breville BOV900BSS with bamboo cutting board. The cutting board is stupid expensive for what it is — especially since I can build something similar. But I bought it for the custom silicone feet more than anything

• Water boiler: Zojirushi DCC50XT. Can’t say enough good things about this

• Coffee grinder: Cuisinart DBM-8. It’s okay, loud as hell. I’m not a fan of how readily coffee grinds go everywhere but I don’t know a lot about grinders, maybe this is normal

• French press: OXO Brew Venture and ESPRO P7. The OXO is my daily driver it’s super well built and I don’t have to worry about breaking glass; the ESPRO has more capacity and better filtration for when we have company or I want to make more coffee

Plumbing

• Faucet: TEKXDD Drinking Water faucet. I picked this one because it was cheap, had all the fittings I could need and included a water line.

• Pump: FlowFluid Bottled Water pump

• 5-gallon jug: I went to my local water store and just bought a bottle of water — the pump I bought has no issues working with nearly any type of bottle

Stuff in the Drawers etc.

• Glass mugs: Zwilling Sorrento double-wall glass mugs. I love these but you do have to be a bit delicate with them and their roundness can lead to tipping over if you’re not careful

• Insulated mugs: Yeti Rambler 14oz and Milwaukee Packout 30oz. The lids for these two are interchangeable: the Milwaukee lid holds heat better, the Yeti lid is less aggravating

• Insulated travel mug: Zojirushi SM-SHE60BZ (20oz) and Zojirushi SM-KHE48AG (16oz). Can’t say enough good things about these in terms of insulation value

• Spoons: Callyne spoons. Nothing particularly noteworthy: I wanted a whole bunch of little spoons specifically for drinks without having to raid my normal dessert spoons

• Tea kettle: Zwilling Sorrento tea kettle. Very nice tea kettle, but as before, need to baby it a bit

• Tea steeper: Bottom dispensing tea-steeper. I’m not sure if ours is a DavidsTea or not (doesn’t really matter, they are all the same-ish). Works pretty well

• Tea organizer: Clear tea organizer

• Containers: OXO Good Grips 2.6L and OXO Good Grips 1.6L. I debated doing a full product comparison for various other brands of products (in particular, the President’s Choice Click-n-Store) but [a] didn’t have time to do it and [b] the OXO stuff has been around for a long time and will likely continue to be around for a long timer — other brands have product lines that come and go.

• Silicone mats: 25.2” x 17.7” silicone mat. Nothing to it, it’s a silicone mat and can be cut with a sharp knife

• Dispenser: Honey Can Do KCH-06074. Seems really stupid, but being able to consistently pop 1/4, 1/2 or 1 teaspoon in a squeeze without spilling everywhere is kind of awesome

• Garbage can: Mbillion 1.05-gallon hanging garbage can. I was looking for something tiny, was pretty and had the least amount of front-to-back depth

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

Quick Links

• Starting Point

• Goals and Objectives

• Rough Design

• Step 1 - Revisiting the wall component

• Step 2 - Remaking the bottom panel

• Step 3 - Remaking the middle panel

• Step 4 - Tweaking the top panel

• Step 5 - RGB

• Step 6 - Side panels

• Step 7 - First power up

• Step 8 - Revisiting the drawers

• Step 9 - Speaker mounts

• Step 10 - Magnet all the things

• Step 11 - Misc touches

• Milestone

• Mixups, surprises and lessons learned

• Appendix - Parts List

Starting Point

The original Wall-Desk is still one of my most ambitious projects ever, encompassing a few different hobby sectors and by far and large, the desk is still, very effective. As such, I don’t really want to reinvent the desk so to speak.

The first catalyst for revisiting the wall-desk was actually when I built a pre-levelled platform to compensate for the severe sloping in my office: the unforeseen five-inch platform raised the altitude of my left/right cubbies and by proxy, raised the desk. While I got lucky with the connections at the floor (none of those were blocked by the platform), the new altitude of the desk sides meant that the desk top would now interfere with the wall-mounted keystones and speaker mounts that were attached to the wall (since the wall was not included in the five-inch platform).

The second catalyst came in the form of a new monitor purchase or rather the power brick that the monitor uses — there was no way I was going to be able to neatly mount that behind the monitors, so I needed to find a way to put that somewhere.

As a bare minimum, I probably could have gotten away with rebuilding the side cabinets to lower the altitude of the desk, but this was a good opportunity to revisit some enhancements.

View fullsize

View fullsize

View fullsize

View fullsize

Something cool to see is how my cheap RGB strip fared over the years; with some minor exceptions, I ran the RGB strip at 100% brightness on a single color (blue) for five years. That’s over 40,000 hours if you’re counting. The fact that my strip still turns on and provides something is kind of impressive; but it’s definitely earned it’s well-deserved retirement (or at least to be deployed with a different color). The strip is the Govee H6163 which I’m pretty sure if all kinds of discontinued now.

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Goals and Objectives

I booked vacation time to work on this, so everything needed to be back up and running ‘enough’ by the time my vacation ran out. I had roughly two weeks to tackle this; I also didn’t want to ‘reinvent the desk’ so to speak: I wanted to reuse as much as I could from the existing desk.

As for the more specific considerations:

• I wanted to be able to level my triple monitors more easily: when you look at the back side of a monitor to see where the VESA mounts are, some (most?) monitors have the mounting point in the dead center of the monitor. With a lot of cheaper monitors (like my side displays at the time), the mounting point sometimes shifts towards the bottom. This shift makes aligning monitors trickier (if you’re using fixed wall mounts). You can use a VESA slider to give yourself some vertical wiggle room, but this is tedious and annoying

• The original wall supported 4x USB-A and 1x USB-C port integrated into it; I wanted to add a whole bunch more USB ports into my wall (without resorting to using hubs); this would allow my to support future PC upgrades

• Based on how I used the wall over the last few years, I wanted to move where some of the connection points were on the desk. In particular

  • I wanted to run the networking down through the top of the faux wall; this would allow me to run the cables from my network rack up into the drop ceiling

  • I wanted to move the rear-speaker connections to the side of the wall, and higher up (to reduce the chaos in the corners)

  • The XLR connection for my microphone could be moved out of sight and routed to the side of the wall (where the boom arm is mounted)

  • I wanted to revisit how I was mounting the speakers to the wall as well

• I wanted more outlets, both on the face and the right-side of the desk. This would also allow me to connect everything to the wall, and then connect the wall to the UPS with a single cord

• On a related note, I wanted to reduce the amount of crap on the desk. I accept that I probably can’t get away from having to account for keyboard, mouse and mousepad but I wanted to find a way to reduce some desk clutter:

  • Capture card

  • SATA dock for backups

  • Google Nest Mini, Amazon Echo Show 5

  • Clock

  • Speaker control pod

  • A label printer

• I wanted to replace the RGB strips with something that I could have local control with (and I wanted to take repair, replacement and upgrading into consideration)

And a big one: I wanted to support an additional PC on this wall (a real computer, not an anemic work laptop). In short, I wanted to add a whole bunch of connectivity to the wall, including supporting a whole new computer and simultaneously reducing the amount of stuff that will live on the desk.

A note about ‘minimalist’ trends

“Reducing desktop clutter” and “minimalism” are trendy buzzwords that influencers like to push around. My drive to reduce the amount of [tech] stuff on the desk has nothing to do with minimalism - I want to reduce the number of devices I need to unplug/relocate if I need to take the entire desk apart for whatever reason.

[Back to Top]

Rough Design

At a high level, I was going to do three things:

1. Change two of the three monitor arms from fixed monitor mounts to gas-shock arms. In theory, I can leave the center monitor on a fixed mount and having adjust-ability on the left-right monitors would allow me to get everything aligned

2. Add a bunch more outlets to the wall

3. Add a boatload more keystone connections

This allows me to keep most of the faux-wall elements — I would mostly be changing out the cutouts on the ‘face’ and the sides of the wall. This should keep the cost of materials down.

It was quite chaotic to look at all of this on a single diagram, so I broke things out by category.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 1 - Revisitng the wall component

When we initially got the office setup, we knew the floor was sloped but we were in a crunch so we shimmed the faux-wall up as best we could and screwed it to the actual wall. When I recently levelled out the floor in my sitting area, I swapped out the shim for a more ‘proper’ levelling foot. Now that I’m going through the effort of rebuilding the wall, this is a good time to just fix it more properly by pre-levelling out the base.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

This was a good time to cut/drill all the access through-holes for wires I specifically cut everything (in the 2x4s) oversized so I wouldn’t have to think about alignment — when I cut the openings in the plywood ‘skin’, I’ll need to be much more careful.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Regarding the screws that we used to initially attach the faux-wall to the actual wall, I didn’t want to make new holes and I didn’t want to wear out the screw holes so I needed to find a way to make this faux-wall (and any additional unit) “compatible” with the existing screw holes as well.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 2 - Remaking the bottom panel

The bottom panel definitely needs to be re-cut as there were substantial changes to the number and placement of keystone panels. I ended up doing it a super hard way by cutting everything out using an oscillating multi-tool; I know you can drill four holes and connect them using a jigsaw, but I simply didn’t trust (myself, the tools, nor the process) to be able to make the cuts super precisely (since this is a finish surface). The overlap of the trim panels I had on hand was something tiny (3/8? 1/4? I’m not sure) and I’ve seen drill bits skate more than that...

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Spoiler: in my original plan, I was going to run the speaker connections on the right hand side of the wall but at some point, I decided I didn’t want to do that and I would come out the face of the wall as I did previously. So I’m missing two keystone panel cutouts that I would have to make later.

[Back to Top]

Step 3 - Remaking the middle panel

Although most of the middle panel would be ‘hidden’ due to having another panel on top of it, the placement of the keystones required me to make the panel anew. I probably could have figured a way to reuse the existing panel, but I wasn’t going to expend the brainpower to deal with it at the time. Same as previously, I cut everything much more generously because there’s no reason to fight for stupid-exact tolerances on cutouts that won’t be seen in any way.

One change I did have to consider was switching two of the fixed monitor mounts for gas-spring mounts which had a different bolt pattern.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Once the woodworking phase was complete, I continued with the wiring phase — the middle plate is the most complicated since there’s a ton of stuff coming in/through it (from below and now, from above too). Thankfully, a Rubbermaid bin was a good height to bring it to a close enough that I could work on it.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 4 - Tweaking the top panel

I was able to reuse the top panel and just make some cutouts to it which saves me a bit of material and time. For some reason (it made sense in my head), I opted not to fully secure the middle (nor top) panels down and just leaned/propped them up. It’s a bit complicated to explain, but I did have to go back and forth with taking the middle and top panel off to fish wires around. As I got closer to the finish line, I did run some temporary screws to hold the middle and top plates in place.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

Although I could have used the cubby as it was, I opted to knock out the divider and have a large compartment up top. This would afford me plenty of space to put all the tech stuff I wanted to have up top. While I had the cubby on the worktable, I reinforced the microphone mount area as I noticed it sometimes had a bit of flex.

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 5 - RGB

Coming off the LED Slat Wall project, I knew I was so totally done with chopping up strips of RGB up; the size of the faux-wall (roughly 48” across) was a perfect use case for using two 1m RGB strips. Since I was buying individual strips anyways, I opted for a 144 led/m strip. The only gotcha with these strips is whether they will fit inside the aluminum channel I use. Thankfully they do — you do have to carefully push them in there though, there’s probably 0.5mm wiggle room.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 6 - Side panels

The side panels are the last bit of the desk that needed to be done to finish the ‘woodworking’ component. Due to the way I made things with the Mk I, the 48” span of the panels does not run flush with the frame. In the Mk I design, I just screwed it down regardless, but I like to think I’m a bit more refined now. With the Mk II, I routed out a gap at the front so that if there is a bit of overhang on the panels, they wont interfere with the side panels and I can still screw the side panels flat to the frame.

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 7 - First power up

The downside about this wall project is that you can’t really test connectivity until everything is already in place; or at least, you can’t do so easily. This initial power up is just to verify the core bits work before I look at rebuilding vs modifying the desk portion.

I ended up getting really close with the placement of the keystone panels on the middle plate. With the cabinets the way they were now, there was maybe 1/4” interference between the plate and the desk — totally serviceable, but I was planning on redoing the cabinets anyways (or at the least, revisiting them).

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 8 - Revisiting the drawers

I debated rebuilding the drawers from scratch, but truthfully, I didn’t feel like it yet — I wanted to use the setup for a bit (especially get through a PC upgrade) before revisiting a total rebuild. I opted to take the cabinet apart, shorten it a bunch and put it back together. How hard could it be?

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

The drawer aspect was extra difficult because I was cramming the same amount of drawers into (substantially) less space. I was able to get things to fit but just barely.

[Back to Top]

Step 9 - Speaker mounts

My previous DIY speaker mounts were okay but they had a few problems with them

• They weren’t beefy enough so they were splitting at the connection to the wall (this could have very well been low quality plywood I was using at the time)

• While I was able to adjust left-right, it was a bit cumbersome

• I wasn’t able to adjust up-down at all

I ended up just buying a few cheap speaker mounts and I also had a similar set on hand that I picked up from a surplus store. For this phase, I only needed to mount the Front Left, Center and Front Right speakers to the wall.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

For the rear speakers, I specifically didn’t want to have speaker stands that I would inevitably knock over or bump into; previously I used a drawer slide to have a way to retract the rear speakers but I wanted to try something different this time around. I tried using a pair of low-profile microphone boom arms: these boom arms came with a 3/8 stud on them — luckily I happened to have some 3/8 T-nuts so I was able to adapt my previous speaker mounts to them. The boom arm is rated for 3lbs — my rear speakers are a bit under 2lbs so, on paper, it should hold. Giving it a bit of a ‘push test’, the arms seem to be okay for now, but time will tell.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 10 - Magnet all the things

I’ve been using magnetic charging cables for our for a long time now: it’s been wonderful to not have to think about whether something is Micro-USB or USB-C — it’s the same magnetic charging connector for both. A special bonus is that the magnetic cables handle power and data.

I ended up getting a Logitech G915 (linear) - more to try it than anything else - I’m still very much in love with my G510 (but I do recognize that my stockpile of backup G510s will eventually run out one day). I do like having a wireless keyboard — not because I’m opposed to wires, but the wire on my keyboard is so long (and it’s not detachable); the distance between the end of the keyboard and my keystone is 18” so having a 6-ft cord is just an exercise in frustration.

The way I have things setup, I connect the keyboard via wireless and I keep a magnetic adapter plugged into the back. When the battery runs low, I just connect it to a magnetic charge cable I keep around my desk anyways. Since the charge cable is power-only, there’s no transition between wired and wireless — the keyboard continues to work using the wireless dongle while charging.

View fullsize

View fullsize

I also opted to put a magnetic adapter on my charging mouse-pad. Since the mouse-pad itself can act as a mouse receiver, I have the option to either

• Connect the mouse-pad to the computer using a data link; the mouse connects to the mouse-pad or

• Connect the mouse-pad purely to power; the mouse then connect to the computer using it’s original nano-receiver

For the most part, there’s no functional difference (connectivity wise) between these two options. The change for me is that I have a few more options on how to connect the mouse-pad to power source (and reduce the number of cables draping across the desk). Being a magnetic cable, I can just take the mouse and mouse-pad in one go.

View fullsize

View fullsize

For my Stream Deck, it’s a bit more of nuanced implementation; in my specific case, I wanted to connect the Stream Deck using USB-C (the cable that ships with it has USB-A on the computer-side). I paired a USB-A to USB-C adapter along with the magnetic adapters to make the whole thing quickly detachable. It was a bit of work to stuff the long cord into the included mount, but I got it in there.

[Back to Top]

Step 11 - Misc touches

One of the goals of this rebuild was to reduce the amount of stuff on the desk that I would have to carefully unplug/move if I ever needed to remove the top to access the bottom/back of the desk. For the time being, I opted to shelve the Amazon Echo Show 5 so that still leaves the Google Nest Hub. From some scrap plywood and some magnets, I just attached this to the network cart, just at the edge of my periphery (I suppose I could also make a magnetic mount for the Show 5 if I ever decided to revisit that).

View fullsize

View fullsize

View fullsize

The desk features a big access panel with an internal shelf, at this time, I keep just a monitor brick, my WLED controller and a USB capture dongle. On one side, the capture dongle connects to the USB feeder ports in the bottom left and up at the desk level, it transitions to an HDMI port. This way I can directly connect a camera to the wall.

View fullsize

View fullsize

I also added a remote [physical] power switch to my desk; at this time it’s linked to a fan I have pointed at me, but in the future I might consider getting an additional one to remotely trigger a light or something. Depending on how often I may bump into this with my armrest, I may adjust the altitude of the switch, but for now it’s working out great.

View fullsize

View fullsize

View fullsize

I normally keep some kleenex in one of the segments of the upper cubby but since I removed a divider, I lost the allocated space for that. I made a quick tissue box holder from some scraps and mounted it up top.

The speaker control pod was the last thing to be mounted up: I debated making a quick shelf for it off to the side, but I do actually need to have reasonable access to it to switch the inputs and possibly to turn it on/off. I ended up using some steel strapping and some VHB tape to keep it secured right overhead.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Milestone

All in all, I think this desk rebuild turned out well and has breathed new life into my office. I do have to give credit to my wife though, for putting up with the absolute disaster zone that was the office for a week while this was under construction. In the next month or two, I’ll be revisiting her wall-desk and levelling her floor area in one go. Thankfully her wall-desk is a bit less involved cable-wise, so it should go by much easier.

No KVM or USB switcher

I specifically did not want to include any kind of KVM or USB switcher to toggle between the three machines. From PC-A to PC-B, I only need to migrate the two nano-recievers for my mouse and keyboard and the work laptop can connect via Bluetooth (keyboard) and VNC. Remoting is also an option to pass keyboard/mouse through as well. For PC-A and PC-B, the monitors are directly connected.

Reducing Clutter

Reducing the number of ‘permanent residents’ on my desk was one of the motivators for this rebuild: I started out with nine or so devices that were parked on my desk regularly:

• Keyboard

• Mouse

• Mouse pad

• Stream Deck

• Google Nest

• Amazon Echo Show 5

• Label printer

• SATA/USB Dock

• Speaker control pod

After the Mk II rebuild, I’ve got this number down to to just four: keyboard, mouse, mouse pad and Stream Deck. Of these four devices, only mouse pad and Stream Deck require ‘unplugging’ before moving them from the desk but given that they are connected by magnetic cables, it hardly registers.

What’s in store for the Mk III?

• Looking ahead, I will almost certainly be rebuilding the side drawers once I settle on the next system build. I will certainly be swapping out the stupid soft-close drawers for traditional, undampened drawer slides.

• Expanding on the idea of connecting the keyboard and mouse pad with power-only connection, I’d like to explore alternative ways of getting USB-power to these devices without having cables strewn across the desk (perhaps a strategically placed USB grommet).

• I would also like to have more ‘hidden USB keystones’; by hidden I mean, that they can be in the upper cubby, be placed behind the monitors or even below the desk — dongles for my keyboard, mouse and wireless headset don’t actually need to be visible

• You might have noticed that the monitors aren’t fully aligned with each other — this is because the VESA mount on the right monitor is located near the bottom of the monitor. This seems to be a fairly common thing with ultra-cheap monitors. In comparison, the VESA mount location for my two other monitors is located dead-center. I thought that the gas strut monitor mount would allow me to compensate for this — and it does — but I’ve maxed out the declination and it’s still out of alignment. I’ve been low-key planning another monitor upgrade this year so maybe that problem will sort itself out.

[Back to Top]

Mixups, surprises and lessons learned

This was a surprisingly uneventful project with only a few minor setbacks. The first was having a WLED control brick itself during an update (thankfully, I hadn’t gone down the rabbit hole of configuring things yet - this was just during initial setup).

The second was a bit more annoying: ultimately, I discovered that a fairly new network cable had a short in it. In hindsight, I definitely noticed something was off months ago, just never thought to check the cable itself (since it was relatively new and not abused). Thankfully, the network cable in question was outside of the wall (it was one of the feeder lines providing connectivity to the wall) so it was an easy thing to replace.

This brings up an interesting point: it is really challenging to ‘test as you go’ with a build like this. For a lot of it, I have to be nearly completed before I can really do any testing. I suppose I could bust out a laptop to do some spot checking of keystones as I go but really, a keystones generally ‘just work’.

Testing some connections, like the capture dongle require testing multiple jumps too and given the chaotic nature of the build, it’s not entirely viable.

Thankfully, I actually had no issues with any of the connections within the wall.

[Back to Top]

Appendix - Parts list

To make things even more complicated, the cables noted below are self-contained within the wall-desk — each keystone endpoint will require an additional cable in order to connect that port to the corresponding port on a computer to ‘feed’ it. Think of the wall as a docking station for the computer.

Networking

• 8x CAT6 keystones — I happened to buy a few pre-loaded wall-plates that came with them

• 4x runs of CAT6 inside the wall

Audio

• 8x 3.5mm stereo keystones (6x for both ends of 5.1 signal, 2x for both ends of a stereo signal)

• 1x 5-speaker binding post plate (to connect the surround speakers)

• 2x RCA keystones (you don’t need different colors, but it helps)

• 1x Female Stereo to 2-RCA + 1x 3.5mm stereo cable with one right-angle (this allows us to connect the stereo signal to the second input on the Z906)

• 3x 3.5mm stereo cable with two right-angle (this allows us to connect a 5.1 signal to the primary input on the Z906)

• 1x Right-angle XLR adapter

• 2x XLR wall plates (one male, one female) — you can get something like this and separate them

• 1x XLR cable with one right-angle on the male or female end (the specifics depend on what gear you’re trying to connect together

• 1x cheapo microphone boom arm

• 20x Banana plugs (need them for each speaker as well as for the connection to the wall)

• 10x Banana plugs for spring loaded inputs (used on the speaker end)

• 5x Round 2-way speaker box terminal

• 3x Cheap speaker mounts

• 2x Low-profile microphone arms

• 14AWG speaker wire

USB

• An ungodly number of keystones, 30x USB-A and 10x USB-C if I‘m counting correctly

• 15x USB A-to-A cables

• 1x USB C-to-C 20Gbs/DisplayPort-AltMode (this is used as a video cable)

• 4x USB C-to-C cables (you could get five of the previous but these are cheaper since we don’t need the enhanced video spec)

• USB capture card (can go with a cheap unit or something fancier)

Video

• 3x HDMI (2.1 or 2.0) keystones

• 3x HDMI (2.1 or 2.0) cables

• 3x DisplayPort 1.4 cables

• 1x Fixed monitor mount

• 2x Gas strut monitor arm

• VESA vertical adjustment adapter (if needed)

• Right angle DisplayPort adapter (if needed)

RGB

• 2 WS2812B RGB Strips (144 led/m)

• 1x WLED controller

• 1x 10A power brick

• 18AWG RGB extension wire (if needed)

• 2x Aluminum channel with diffuser

Power

• 1x Inline extension switch cord

• 2x Desk power bar with USB

Misc

• Large wall mounted USB clock

• 6-port wall plates

• Blank keystone insert

• 1x Access Panel (8x8)

• 3x Cable pass through wall plate

• 10+ Low Voltage boxes

Two Special Call Outs

There are two things that definitely made this project so much easier to tackle: the Milwaukee M12 Installation Drill/Driver and CordIDs from Dotz. The former is pretty straightforward: an ultra compact and versatile drill/driver that excels in ‘close quarters’ actions. The latter provides a means of labelling the tires but due to their clip-on/clip-off nature, provides you with three game changing benefits:

1. You can easily flip the clip so all your labels are oriented the same way, you can also rotate the clip to show either side

2. You can clip them onto cords after you’ve already wired them if needed (so you can focus on one thing at a time)

3. The smooth plastic surface means that labels both stay on there if you want them to but can be peeled off if you need to update it

I used probably close to a full bulk pack (100) of these CordIDs on this wall; at the very least, each wire in the wall gets two labels (one at each end) with some additional ones getting two more (in the case of video cables, they are labelled on both ends, inside and outside of the wall. It’s so nice to not have to guess what a given wire is — it’s labelled.

[Back to Top]

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

After build the LED Slat Wall last weekend, I opted to convert the existing desk into a side table of sorts. In the interim, I pulled out one of the ubiquitous plastic banquet tables and pressed it into service as a computer desk. Even when directly clamping onto steel under-frame, over the course of the week, I could see the plastic top deforming under the clamping force of the monitor-arms.

Guess I’m building a desk then.

Quick Links

• Goals and Objective

• Rough Design

• Step 1 - Making the top

• Step 2 - Making the legs

• Step 3 - Connecting the legs and top

• Step 4 - Installation

• Step 5 - Positioning

• Milestone

• Appendix

Goals and Objectives

By far, the primary objective was to get a desk that addressed three things:

• It needed to be sturdy and stable — I was planning on mounting monitors to the desk, I didn’t want to deal with a tipsy, unbalanced desk. To make things harder, remember that everything is on carpet as well

• I wanted the desk to fit into the rest of the aesthetics of the room

• I wanted to substantially reduce the cable clutter behind the desk

As a bonus, I wanted this to be a quick build from materials I had on hand.

[Back to Top]

Rough Design

I originally planned on making a desk similar to the existing one: square ‘wheels’ out of 2x4s connected with a few stretchers and with a plywood topper on top. After doing the slat wall though, I was really on a streak for nice, hard, crisp edges so I ultimately decided against that. I wanted something that was more ‘skeleton’ like.

Since I’m paranoid about tipping, I debated adding an anti-tip leg - something to catch the desk and forcibly keep it from tipping backwards. I wasn’t sure if I was going to need it or not, but I kept this as a backup plan.

[Back to Top]

Step 1 - Making the top

A little while back I had bought a [damaged] sheet of melamine (with red-oak veneer on the opposing side); the sheet I bought was supposed to be 5/8 but with some of the post-damage swelling, it was closer to 3/4 in thickness and there was a definite bow to it. I knew I wanted to laminate two layers of this together to make the top — but which faces to sandwich together? I originally wanted to have melamine exposed on one side and red oak on the other — and make the desk top symmetrical so we have the flexibility to flip the top if desired but I don’t have any melamine glue so that plan was dashed.

I definitely didn’t want to rely on mechanical fasteners so this decision ended up being straightforward — the only option available to me was to glue the two wood-surfaces together. My next problem though was that the sheet had a substantial bow in it — I would have to cut slits into the sheet so that it could lay flat. I marked off slits on both surfaces (making sure to offset them so that a slit would not be across from another slit) and after cutting a bunch of slits, the sheet laid flat for easy gluing.

View fullsize

View fullsize

To trim out the perimeter, I debated going with walnut for the incredible contrast but I didn’t have the right lengths handy (and wasn’t in the mood to rip and join smaller strips together). I kept this simple and just used plywood. I cut the edging 1/4” wider than needed and very carefully glued the edging to be perfectly flush with the top side; I had planned to come back and flush-trim the bottom edge, but ultimately didn’t since I just didn’t trust the router to not make a disaster of the melamine. I might revisit that down the road as it’s easy enough to do.

[Back to Top]

Step 2 - Making the legs

Based on my existing desk, I wanted to hit a top-side height of roughly 29-inches, working backwards, I figured out I needed to cut 27-1/4” legs give or take (the carpet does allow for some extra give). To ensure things were as rigid as possible, I made the legs from plywood glued at 90-degrees. I ripped the two necessary strips so that, after joining, they would be equally sized. To keep everything as consistent as possible, I gang-cut all the strips where possible — and the off-cuts made for perfect spacers to position the legs on the top!

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 3 - Connecting the legs and top

With the individual legs done, the next step was to connect the two sets of legs with a pair of left-right stretchers and then somehow, connect the table top to the legs. Note that the connection between the two pairs of legs does not need to be super strong — the weight is being loaded down through the legs directly. The stretcher is there to minimize any sagging in the table top and to keep the two sets of legs physically connected to make it easier to handle. I opted to screw-only here, in case I ever wanted to take this apart.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 4 - Installation

Being comprised of two separate parts made this super easy to bring inside and install in place. As part of the desk upgrade, I put a trailer bed mat down to be a chair mat: figuring out which way to spin the mat and the underlying plywood mat (which I took from my original office desk build) took us the better part of an hour. In the end, we didn’t need to cut anything (which was great, I wasn’t looking forward to trying to cut the mat). The last major thing to do was figure out where to place the monitor arms — which then determines where to drill the grommet hole.

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

View fullsize

[Back to Top]

Step 5 - Positioning

In order to get the positioning, we needed to work backwards a bit

1. The camera was centered against the center of the wall

2. The monitor was centered on that same centerline

3. The came centerline would focus on the user’s left eye

   • The floor mat is then adjusted to be centered relative to the user

   • The desk is shifted to be appropriate for the user’s relative position

4. The second monitor is then place at a slight angle

5. The elevation of the tripod and monitors is adjusted to frame the user

6. The microphone is placed out of frame

View fullsize

View fullsize

[Back to Top]

Milestone

This was a fairly quick project where things mostly fell into place and I’m super grateful for that. I had actually originally planned on doing this desk as part of the LED Slat Wall build last week (albeit as a simpler version) but the wiring woes chewed up all my time.

So far, I don’t think I need to add any kind of anti-tip leg (although it’s not hard to retrofit that down the road if needed).

[Back to Top]

Appendix

• Monitor arms: Charmount Single monitor arm gas-spring mount (max load, 19.8lb, 2.3” grommet capacity, 3.3” clamp capacity). Since the grommet is just an M8 bolt, you could theoretically replace this with whatever custom length you need (if you needed more than 2.3” capacity). I specifically went with two single-mounts rather than a single double-mount as this gives me better flexibility on placement and better weight/force distribution

• Mat: 4x6x5/8 trailer bed mat. Alternatively, you could also go with a 4x6x3/4 horse stall mat

[Back to Top]

---

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Skype DeskDIYDPFri, 24 May 2024 14:45:00 +0000https://www.minmaxgeek.com/blog/led-slat-wall6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:664e090f20327c1ff74dae7eWe’ve had Skype Station for a few years now and it’s been great: it gives Nicole a low maintenance and (fairly) reliable place to do her video calls (I say ‘fairly reliable’ since it’s also a bit of an AV guinea-pig setup for me) and can be a separate working space when needed. Since moving to this house, I did a minor refresh but most of the changes have been fairly minor. While the hanging-backdrop solution we have is pretty good, this is something I’ve been wanting to tinker with for a bit so while Nicole was out for the weekend, I set out to do a ‘quick turbo project’ (spoiler: it was most definitely not quick turbo).

---

Quick Links

• Starting Point

• Goals and Objective

• Rough Design

• Step 1 - Painting the base layer

• Step 2 - Making cleats

• Step 3 - Making slats

• Step 4 - Routing slots fo channels

• Step 5 - Wiring RGB (aka misery)

• Milestone

• Mixups, surprises and lessons learned

• Appendix

---

Starting Point

With respect to the backdrop, not much has changed since I originally built Skype Station:

• The actual hanging background is a 5x7 ‘shower curtain’ essentially

• We used a pair of Jumbo Command Hooks to hang 1/2” EMT conduit on the wall. We using a combination of clips, twist-ties and string to hang the backdrop off the conduit

When I built the original Skype Station, we were renting so I wanted to minimize the number of ‘holes in the wall’ so this setup worked great. In hindsight, if we were to do this again, I would opt for a wider backdrop — the wider the better. This way, we had more flexibility with regards to how we setup the camera to frame everything. Being 5’ wide pigeon-holed me into a specific camera placement.

[Back to Top]

Goals and Objectives

At a high level, I wanted three things

• This needed to be a fairly quick project — Nicole was out for the [long] weekend and I wanted to have it done by the time she got back

• I wanted to cover most of the wall so that I could have flexibility for [camera] framing

• I wanted to give it a ‘cool factor’ but without it being excessively distracting

• It needed to be ‘fairly affordable’ for the given amount of ‘cool factor’ it provided

My dream scenario would be to have some TVs mounted as faux-windows on the wall — something like what Drew Builds did; I’d configure it either as three 55’s in portrait mode in a 1x3 configuration or perhaps four smaller 48’s in a 2x2 configuration. For the sheer awesome-factor, I could totally stomach the cost of buying some cheap TVs, but the ongoing electrical cost (and extra heat they put out) killed the deal for me.

 

---

 

To keep heat and power usage in check, I opted to go for a slat-wall with RGB strips. For strips, I specifically wanted to look at DIY strips using something like WLED rather than off the shelf strips from a company like Govee. While I haven’t specifically run into a lot of grief with off-the-shelf RGB strips per-se, I didn’t like that everything was dependant on cloud-connectivity. Being cloud-dependent has a few drawbacks:

1. You’re at the mercy of cloud connectivity

2. All your commands need to go up to the cloud and then back down to the RGB strip which then means

3. That you’re mostly limited to whatever lighting animations and configurations that are provided out of the box

[Back to Top]

Rough Design

This project is broken down into four phases:

1. Paint the wall a dark color. This provides a bit of a moody base tone

2. Make the slats. A lot of online videos have you use 1x2 strips but since I have an easy way to cut it, I’m just going to use plywood

3. Cut slots to accommodate the aluminum channel

4. Mount the slats and wire up the RGB strips

At the start of this project, I expected the process of making the slots for the aluminum channel to sit in to be the biggest hurdle. I debated between using a plunge router with a track, a router on a router table or just cheating and using the table saw to cut dadoes:

• Plunge router: advantage that I can set the final depth once and just use the turret stop to do multiple passes; downside is that this is a herky jerky process and I didn’t have a lot of wiggle room to work with — I’m routing 3/4” slot on something only 1½” wide

• Router table: the advantage here is that I would feel a lot safer slowly dropping the strip onto the bit and I was okay with figuring out how to mark start and stop points; the downside is that I didn’t feel entirely comfortable with the process of making progressively deeper cuts and landing the same slot again and again (likely just a lack of experience thing)

• Table saw: mentally this is the easiest process — figure out where the slot needs to go and use the miter saw to cut above and below that point and use the table saw with a dado stack to cut the groove cleanly in one go; the downside is that I would waste time gluing it back together and there’s always the risk of the glue-up being slightly askew

For wiring, this was the easy part: my plan was to leverage solder-less connectors and spade bits to wire everything up. #foreshadowing.

[Back to Top]

Step 1 - Painting the base layer

I’ve never painted anything that ‘mattered’ before in my life. Sure I painted the shop, but that was white-paint, on white-primer on white-drywall, and it’s a shop — it’s not required to be pretty. Here I am, taking a wall that is (roughly) white and making it black. On a room with carpet. Without Nicole finding out.

View fullsize
Masking off the wall View fullsize
Additional coverage on the sides View fullsize
Cutting in View fullsize
Rollered in the fill

I went with Behr Marquee matte-paint in Limousine leather. My working area was about 70 sq-ft and I was told that the small can would give me about 100 sq-ft. When I bought it, I wasn’t sure if I needed two coats or not, so I picked up a tester as well just for a bit of buffer because I didn’t want to have to go back and get more paint mid-project. Spoiler: I didn’t need two-coats and I didn’t need the extra sample container.

For masking, I went with blue tape because I didn’t want to risk pulling off any good paint on the edges and I used a YOLO mix of dollar-store artsy brushes, foam brushes, a cheapo bristle brush and random 10mm roller. No idea if my approach was right, but I think it turned out okay (and it’s mostly going to be covered anyways) and the most important thing for me was that I didn’t get any paint on the floor or walls.

[Back to Top]

Step 2 - Making cleats

One thing that makes my wall a bit different from most of the ones online is that mine is [mostly] hung on a French cleats which means that I could remove the entire thing off the wall if I absolutely needed. In truth, I do have a middle brace piece that isn’t a French cleat and I did have to pin a few slats to that to keep them from wobbling, but the pins can be pried off if the need arose. In hindsight, I would have done three sets of French cleats so that I definitely could have taken it off the wall without effort.

I painted the cleats (and the screws used to hang them) black to make this blend away together.

View fullsize
Ripping cleats down using the track saw View fullsize
Almost! View fullsize
Cut to length View fullsize
Old habits die hard View fullsize
Two quick coats of paint View fullsize
Center aligning View fullsize
Aligning the bottom cleat View fullsize
Middle support brace View fullsize
Painted the screw heads

[Back to Top]

Step 3 - Making slats

At the start of this project, I had about 2½ sheets of ¾-plywood and I had settled on roughly 1½” wide strips with a ¾” gap between them; doing some math, I needed somewhere around 42 strips — depending on whether I was planning on centering the slats or aligning to one of the edges (and I guess, depending on how accurate I was with lining things up). Because I like to live dangerously, I opted to make 43 slats to give me a full extra spare.

I used the tracksaw to rip the plywood down to a manageable size and then batched everything out on the table-saw. This was a bit messier than making all of the cuts with the tracksaw but it was so much faster and I could play into the strength of the table saw and get dependable cuts. I explicitly sanded all surfaces to 220-grit, in particular, going as far as batch-sanding the plywood edges together. Time spent here getting a smoother finish makes the staining process go so much faster (because you can simply swipe with a brush for coverage rather than having to go back over rougher surfaces).

Ripping the sheet to a manageable width

I measured out enough width for about 8 strips at a time making to add some extra to account for blade kerf as well. I set this width using a combination square so I could go batch out ~8 strips at a time

Crisp

Even from a distance, these are so crisp and clean (doubly so compared to 1x2s). Credit to the 90T blade I used.

Initial end

The slats are cut down from an 8' length of plywood, so one end was cut with the tracksaw

Cut to final length

I cut the initial plywood 1/4" longer than planned so I cut them to final length on the miter saw

Sanding

Time spent sanding at this point makes everything go smoother downstream. I debated adding a roundover to all the vertical edges) but I was digging the super crisp look of the slats

Splash of stain

My tried and true 'Early American'. The staining process was exceptionally smooth - everything went on with a single stroke and wiping it dry was splinter-free

Roughly visualizing

Brought all of the slats upstairs to visually confirm whether I should go with an odd-number or even-number of slats: to center align a slat on the wall or a gap-between-slats.

[Back to Top]

Step 4 - Routing slots for channels

An easy method to add RGB to the wall would be to surface mount the strips (either on the front or on the back, for glow) but I definitely didn’t want to do that. Another easy option would be to place the RGB channels in the gaps between the slats but the inconsistency would drive me mad. I could also put a line of RGB at the top/bottom/both running the width of the wall, that would be easy but it didn’t have the same cool factor.

All this to say, I was going to route out slots in the front face to place the aluminum channels and then run wiring from behind. Not only that, it was going to be a bit of a ‘dynamic’ shape. I borrowed a router table and did a few cuts but it still felt hella uncertain (and I was on the clock for this project so 'testing around’ wasn’t viable); so after a few slots, I opted to setup a routing jig (since all of my slats are identically sized).

There is an argument for squaring off the corners of the slots but I opted to leave them round because [1] it’s less work and [2] the extra buffer area gives me room to drill a pass-through hole.

View fullsize
Routing jig View fullsize
Test slots View fullsize
Laying out which slats get RGB View fullsize
Batching out the cuts View fullsize
Staining the slot as well View fullsize
Cutting the channel and cover View fullsize
Visualizing in place

[Back to Top]

Step 5 - Wiring RGB (aka misery)

Now that I’m done the stressful part of routing out slats, I thought this would be the easy part. I thought that the very first connection would be the hardest; sadly, it turned out to the be the easiest and most straightforward.

This was pretty straightforward and I was in high spirits: on one end (of the JST connector), I put spade-connectors which allowed me to run a custom length of wire between the start of the RGB strip and the controller.

View fullsize
Initial connection View fullsize
Cut my first length of RGB View fullsize
Routed out a channel on the back side

For the first connection, I didn’t have to worry about connecting anything to the RGB strip itself — it came with a pre-connected wire. I thought I would save some time and make it so that I could easily replace burned out sections of RGB down the road by using solder-less connectors. Buy cutting the solder-less connector in half and using spade-connectors, I could insert a custom span of wire which would allow me to wire up the RGB strips up and down the slats. Again, at a high level, this looked a bit tedious but certainly not problematic.

I figured for the nest few hours, I would go through the following process:

• Determine the length of the RGB strip needed, cut it to length

• Drill the necessary hole in the slat to pass wire from front to back

• Hot glue the aluminum channel in place

• Cut a solder-less connector in half, feed it through the hole, add spade connector to it, clip the RGB strip into the solder-less connector

• Take the other half of the solder-less connector, feed it through the downstream side, add a spade connection to it, clip the RGB strip to the solder-less connection

• Verify everything works, add heat-shrink to the space connections (to keep it from shorting out), tape the light down, pin the slat on the wall

The first few slats (with RGB) mostly worked out like this and then suddenly, I would see some extremely odd behaviour: the last few strips would refuse to light. If you cycle the power, there would be a 50% chance it would work. I initially doubted my crimping job — maybe the crimp connectors were having a hell of a time dealing with 22awg wire? So I busted out the multimeter to check continuity and voltage and then discovered my second oddity: there’s no continuity on the data line! Like even between two adjacent RGB nodes.

I would made very painful progress (about 1-hr per light segment): things would work, they would stop working, then they would look like they worked but when you try and change the color (testing the data signal), some LEDs would not change. I crimped and re-crimped several connections a few times; I even switched my connections entirely to spade crimps; I thought I would save time and material in heat-shrinking by using the inline-coupler crimps — but I needed the spade connection so that I could take things apart to troubleshoot.

I eventually got smart and chopped a single LED and put that on its own solder-less connector with spade connections on it. This is where things got extra tedious. Every time I would do anything — connect, disconnect, heat-shrink, run a wire, etc, I would have to power things up to make sure ‘things still worked’. Progress was extremely slow here because I would finish a few slats… just to have a slat that was ‘known to be working’ stop working and I’d be back at square one troubleshooting connectivity.

View fullsize
It started great View fullsize
Hiding wires in plain sight View fullsize
Hanging the wires out of the way View fullsize
Connecting one span to the next View fullsize
Hot glue to stabilize wire View fullsize
More cable management View fullsize
Crimp splice connections View fullsize
Switching over to spades View fullsize
Single tester View fullsize
Tester in action View fullsize
Recess View fullsize
Checking as I go View fullsize
Very painful progress

---

I eventually discovered that the problems I was having was entirely because of the solder-less connectors. In this magnified shot, you can see the metal contacts that slide over the copper connections on the RGB strips. This was an exercise in extreme frustration as I did troubleshooting.

• I verified continuity from the copper contact to the far side of the spade connector. Good tone. This means that each copper contact is touching the prong and there’s a solid connection from the metal prong all the way to the spade connector. I verified this for all three copper contacts

• What if the springy connector is lifting off? While there was power in the system I used a small screwdriver to gently mash the connector onto the contact. No change. I used three screwdrivers to carefully mash all three contacts (without touching each other no less). No change.

• I tried flicking the connector. I tried squeezing gently. I tried squeezing hard. No consistent change.

• I tried re-seating the RGB connection. No dice.

Also, nothing at the time could explain why things would work/not-work and then, with a power reset, they would not-work/work. I would later deduce that just because something lights up does not mean it’s working. I found that the best test was to set a scrolling animation: for an animation to work all three pins (power, ground and data) needed to have a good stable connection— if any given segment stopped animating, then you knew something was up and you could tell where it had fallen over.

This extra knowledge was helpful but also extra infuriating because things would work for hours (including full color changing animation) and I would cycle the power (using a power bar as to not disturb any wire connections no less) and there would be a random break. Upon closer inspection I did discover one (of the two!) problems: when cutting the RGB strips, I had cut slightly poorly on one of them so the copper contact was substantially smaller resulting in intermittent disconnects.

Fixing this was a fairly easy (and fortuitous) fix: I happened to have enough leftover RGB that I could swap out the ‘misbehaving’ length with a longer one where I intentionally cut it poorly (but in favor of making sure there was enough copper). This led to the second ‘discovery’: even when there is a good contact patch, everything is cinched down, there’s continuity, the damn connection still doesn’t work.

I had discovered there was a specific connection where you needed to pinch the solder-less connection with a certain amount of force. Too little or two much and there wouldn’t be a data connection.

At this point, I just gave up.

I took a yellow and a blue crimp connector, broke off all the metal bits (to prevent accidental shorting) and wrapped them in white electrical tape (to not throw off any coloring). I put the blue one behind the solder-less connector and I put the yellow one in front and forcibly jammed the channel’s diffuser on top of it. That was enough force.

Why not solder?

I debated this (and at one point busted out the soldering iron to do a quick test) but I really like the ability to unclip a section of RGB and plug in a new one if it fails — this was certainly something I did during troubleshooting as well. Can you de-solder? Yes, but there’s an additional complication the aluminum channel wasn’t wide enough to fit the quick connector in place (which didn’t help things) — so I had to put things on an angle (and soldering on an angle in a cramped channel was a no go).

---

[Back to Top]

Milestone

With the exception of a couple fiddly connections and the fact that the strips took way longer than I wanted, this was a reasonable success. Nicole gets a bit of a refresh on the room, there’s cool factor and the extra width of the backdrop means that I’ve got more flexibility on where I can place the camera. And technically speaking, I can still hang a curtain-style backdrop if a specific need arises.

Using WLED to control the LEDs is a breath of fresh air though: there’s so much you can configure and everything works snappy fast because it’s all local to my home network. There’s an almost overwhelming colors, effects and settings to play with but it’s all laid out in a very friendly way.

[Back to Top]

Mixups, surprises and lessons learned

By a landslide, the biggest grief point was working with the ‘quick’ solder-less connections. In their defence, it doesn’t help that the connectors don’t fit neatly in my aluminum channel which required me to put them in on an angle. This additional rotation introduces a pull-out force that the quick-connect needs to resist (and intermittently fails to do). Lets look at some sizes:

• My RGB strips are 10mm wide; at the very start of the run I have a pre-attached power lead. The connection (mostly heat-shrink) makes the strip 10.2mm across at this point.

• Fully closed, the quick solder-less connector is 14.5mm across; when open enough to access the RGB strip, the connector requires 14.8mm clearance.

It’s possible for the solder-less connector to clip in neatly at the top of the channel — but then I’m unable to slide/clip the diffuser over it. I suppose I could cut out a small relief on the diffuser to allow it to slide over, but I didn’t think of it while I was putting this together and there’s always the risk of shattering/cracking the diffuser.

 

---

 

Just brainstorming, I think that switching to super deep V-channel (18.4mm internal width) would have solved all my problems:

• Wide enough that I could even consider putting in two sets of RGB strips; obviously wide enough to fit the quick connectors without fuss

• Tall/deep enough design that the diffusion would still smooth out the LEDs

The only change I would have to do would be to route a V-groove rather than a straight/wide groove. To fully recess this channel, I would need to route roughly 21.3mm down. I could probably get away with something slightly more shallow, say 20mm.

This functionally means that I would be routing entirely through 3/4” plywood. At the top, I would need to be roughly 42mm (1-5/8” wide!) across. I could probably manage to do this if I increased the width of the slats to 2” or so. Being entirely through-cuts, I could probably use the table saw to cut this a bit more safely.

There is a less-intense V-channel, which would only require routing 11.3mm down and being 22.6mm (7/8” wide) across. This would be a lot easier to route, but the diffusion would be notably less.

 

---

 

If I was better at soldering (in tight confines no less) then perhaps soldering lead wires directly onto the RGB strips would be the way to go. By including a bunch of slack, there would be enough room to pull the strip out and replace it if needed down the road. I could still leverage spade or molex crimp connectors behind the scenes as well.

One definite advantage here is that I would be entirely working with 18-gauge wire which is so much more pleasant to work with than the 22-gauge wire that the solder-less connectors used. I might need to use some kind of cable clamp on the back side of the slat to keep the extra loose wire in check.

 

---

 

In hindsight, another option would be to just throw entire RGB strips at the problem: each RGB strip comes with a very nice to work with initial connection pigtail (that I could just use spade connectors with). I would simply buy a whole bunch of shorter RGB strips and chop them down to the length I needed, discarding the excess (I would save the excess bits for different projects that were less fussy). I could even go with brighter/smoother 100 LED/m strips. Of course the downside here is that it would cost a lot more. But given how much I swore during this project, it might have been worth it…

View fullsize
Side profile of the aluminum channel View fullsize
Technically it fits View fullsize
From another angle View fullsize
Need to rotate

[Back to Top]

---

Appendix

Electronics

• WLED compatible controller

• 5V 10A power supply

• 5m 60 LED/m RGB strip

• 18-gauge, 3-color wire. For my specific wall, I ended up needing to order two of these

• The solder-less quick-connects. I ended up chopping these in half — when I bought them, it was cheaper to do this than to buy the ‘single end’ version. You can figure out how many you connections you need based on the layout of RGB segments: you’ll need one at each end of the RGB segment

• A boatload of spade connectors. Each end of a connection requires three pairs of spade connectors

• Also a boatload of heat-shrink. The specific spade connectors I had on hand were not fully insulated so I used heat-shrink on all of the connections to keep them from touching

Woodworking, construction and tools

• Roughly 1.5 sheets of plywood for my specific space

• 1 quart of paint + primer in one, I went with a matte limousine leather

• 23-gauge pin nailer to initially hold the slats — they are strong enough to hold the slats but if you need to, you can brute force pull the slats out

• 18-gauge brad nailer to really really secure the slats

• a hot glue gun to secure the aluminum channel to the grooves as well as keeping some wires from flopping

• a heat gun to work the heat-shrink

[Back to Top]

---

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: LED Slat WallDIYDPMon, 13 May 2024 00:45:00 +0000https://www.minmaxgeek.com/blog/dust-cart-mk36020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:664185a54e91c843af038293I have a fairly effective setup for small-tool dust collection: a high-powered vac with a cyclone all mounted on a cart but something that’s been bugging me for a bit now is just how damn loud the vac actually is. I find myself reaching for hearing protection whenever I have to run the vac because the tone is just so annoying. I’ve been on the lookout for a definitively quieter setup. Recently I picked up a ‘quiet vac’ so this was a good opportunity to revisit the dust collection cart.

---

Quick Links

• Starting Point

• New vac (and rant)

• Step 1 - Platform

• Step 2 - Verticals

• Step 3 - Support dolly

• Step 4 - Accessories

• Milestone

• Appendix

---

Starting Point

I’ve been using my Dust Collection Cart Mk II and recently I swapped out my main vac hose and changed up the main cyclone separator as a series of small improvements. On paper, I mostly hit all the check-boxes:

• [Relatively] high-powered/high-cfm vac

• Nice 2.5” hose (anti-static too!)

• Highly-effective cyclone separator

• By using custom-length flex-hose to connect the vac and cyclone, I minimized ‘excess hose’ as best as I could

Having all of this mounted vertically was a blessing and a curse: by stacking the cyclone, I was able to reduce the footprint that this cart occupied but it was ungainly and top-heavy. The addition/upgrade to the 16-ft hose did not help the situation — I had resorted to sprawling the hose along the floor. The important thing though, was that it worked and so there was never really a reason to substantially revisit this.

There were two things that bugged me about this cart.

1. The cart is still bulky. Even though I had stacked the cyclone atop the vac to minimize the footprint, there was no getting away from the fact that the vac was a behemoth to begin with (16-gallon!). This is particularly frustrating because I use the cyclone, so I have little to no use for the vacuum’s actual capacity

2. Running the vac is loud. Sure, it cleans really well, but other than short bursts, I found myself reaching for hearing protection just to try and cut down the tiring drone of the vac

I was able to find a new vac that all but eliminated the noise, so this was a great time to revisit the cart to see if I couldn’t address the bulky concern.

[Back to Top]

New vac (and rant)

Before we get started…

Compared to some other brands that I have looked at vacs in the past (i.e., Ridgid, Milwaukee), DeWalt really struggles to with product information pages. It feels like all of the product pages are just poorly auto-generated and/or there’s a disconnect between dewalt.com (US) and dewalt.ca (CA) — multiple products that definitely do exist and are available for sale in Canada, just don’t have any existence on the Canadian website. Ironically, the single StealthSonic vacuum listed on the Canadian site is nigh-impossible to actually buy…

DeWalt has a series of wet-dry vacs marketed as ‘StealthSonic’ and as the name suggests, the focus is on being hella-quiet. From what I can tell, there are ~four vacuums that fall under the ‘StealthSonic’ marque:

• DXV16P-QTA (US) / DXV16P-QTA (Canada) :

  • ‘6.5HP’-class vac, 16-gallon, ‘cart-style’ vac, 20-ft cord

  • The vac includes some 2.5” accessories so presumably the connection is 2.5” (which is expected for this class of vac)

  • According to the product page (at the time of this post), this doesn’t include a hose so there’s no way to verify. None of the product pictures show a hose.

  • The US site indicates that you get 1 (or 2?) bonus contractor kit(s) but it’s not clear what that kit includes

• DXV12P-QTA (US)

  • ‘5.5HP’-class vac, 12-gallon, 20-ft cord

  • The vac includes some 2.5” accessories so presumable the connection is 2.5” (which is common for this class of vac, but not guaranteed)

  • The product page does say that it includes a ‘crush resistant’ hose

  • The US site indicates that you get 1 bonus contractor kit, but it’s not clear what that kit includes

• DXV12P-QT (US)

  • The same as the above, but with a ‘sturdy hose’

  • Also, you don’t get a bonus contractor kit

• DXV09-QTA (US)

  • ‘5HP’-class vac, 9-gallon, 10-ft cord

  • The vac includes 1-7/8” accessories

  • The included hose is a ‘crush resistant’ 1-7/8” hose

  • The US site indicates that you get 1 bonus contractor kit, but it’s not clear what that kit includes

 

---

 

Based on the information “available” on the product pages, you would be justified in thinking that the DXV09-QTA (9-gallon) would be eliminated for me because I’m looking specifically for 2-1/2 vacs.

But it’s a bit more nuanced than that:

• The included accessories are 1-7/8

• The included hose is 1-7/8, mostly:

  • The hose is 1-7/8

  • The tool-end (where you connect the nozzles and accessories) is 1-7/8

  • The vac-end (where you connect the hose to the vac) is 2-1/4 (which is the ‘actual size’ for 2-1/2 hoses)

There was recently a ‘DeWalt demo day’ at a nearby tool store so I took the opportunity to swing by and see for myself. Embarrassingly, the DeWalt rep was adamant that ‘DeWalt vacs only connect to DeWalt tools’. I had to coax him into the idea that “hey maybe we can go and actually check the demo unit on the shelf”. Sigh. I’m sure this isn’t representative of their entire team but gah.

In either case, I specifically showed up with a 2-1/4 x 2-1/2 coupler (the type that threads neatly onto 2-1/2 flex hose) and individually checked the vacuums.

For my use case, I only needed to confirm that my flex-hose coupler physically fits into the vac any included hoses/accessories were going into long term storage anyways. For regular consumers though, this is a bit of a mixed bag

• The vac can take 2-1/2 hoses, so you could (at additional expense) upgrade down the road — but I imagine if potential buyers knew that in advance, they would have just opted for the 12-gallon vac to begin with.

• If you had just looked at the hose connection and made your decision based on that (like I did ha), when you got home and found that you got all 1-7/8 hose & accessories would feel a little bit like a bummer

 

---

 

TL;DR — I bought the 9-gallon StealthSonic (DXV09-QTA) vac because it accepts 2-1/2 hose.

[Back to Top]

Step 1 - Platform

Instead of building cart around the cyclone and vacuum, I wanted to integrate the cyclone onto/into the vac as this would reduce the overall bulkiness. I lucked out — the DeWalt vac has a nice flat platform that I can lever some plywood against.

View fullsize
Marking the centerline View fullsize
Rough cutouts View fullsize
Cutouts View fullsize
Test fitting View fullsize
Shimming to keep it level

[Back to Top]

Step 2 - Verticals

Technically, I could have stopped there with just the platform for a bucket — add a couple tie downs and some bungee cords to keep it from tilting and call it a day. I wanted to do something a bit different to reduce the amount of tipsiness that the cyclone introduced (particularly with the 16-ft hose connected to the front of it).

View fullsize
Starting with some bowed scraps View fullsize
Additional screws View fullsize
Setting the vertical T's View fullsize
Tying them together View fullsize
Couple of scraps View fullsize
Support arm View fullsize
Another angle View fullsize
Offcut cap

Once the verticals were in place, I needed to find a way to keep the bucket from sliding and tipping. Initially I had thought to use two bungees to strap over-and-down, in a crisscross fashion but my wife had the clever idea of just strapping the bucket to the verticals. Strictly speaking, it’s less secure, but a quick test showed that it’s perfectly fine — and I could always add another bungee at the bottom if I needed it.

View fullsize
Figuring out where to add the tiedown View fullsize
Lined up right at the 'neck' View fullsize
Works quite well

[Back to Top]

Step 3 - Support dolly

Everything up to now is levered by the handle of the vac: with a relatively light load (empty bucket), the plywood base handles it but I can envision it sagging over time. Also, having the plywood cantilevered so far makes me generally feel uneasy, so I wanted to support that mass by connecting it to the ground. Initially I thought to use a spring-loaded gate-wheel, but I wanted to work with materials I had on hand.

View fullsize
Making couple of smaller T's View fullsize
Top plate View fullsize
Pre-drilling View fullsize
Bottom brace View fullsize
Two wheels added View fullsize
Nearly perfect shim

Once I attached this dolly to the platform, it really reduced the amount of ‘chatter’ as I moved the vac and cart around. There is one concern though: that an accidental swipe would just twist this platform off. While I can’t prevent the leg from breaking off away from the vac, I can minimize the chances of it breaking towards the vac by adding a standoff. You guessed it, using a plywood T.

View fullsize
Plywood and 2x3 View fullsize
Just a single pocket hole screw View fullsize
Adding a cord wrap

[Back to Top]

Step 4 - Accessories

The Ridgid vac I used before had onboard storage for accessories and this vac doesn’t so I needed to find a way to get this organized. When I build my miter station, I specifically left a bit of an overhang between the levelled base and the cabinets so I opted to hang the accessories there.

View fullsize
Planning out the holes View fullsize
Drilling from one side View fullsize
Flip over to finish on the other side View fullsize
Adding a splash of color View fullsize
Pre-marking the other spots View fullsize
Attached to the cabinet

[Back to Top]

Milestone

This was a spur of the moment project triggered by the transition to a new vac and overall I think it was quite a success: the vac cart takes up less space (both literally and ‘visually’) and I’m able to bring the hose around with the cart rather than dragging it along the floor. As a bonus, the brace that I used to connect the vertical T’s makes for a good grab-handle.

View fullsize
View fullsize
View fullsize
View fullsize
View fullsize
 

---

 

The burning question is: how does it sound? I made a quick comparison clip before building the cart and it’s a pretty stark difference.

Quick noise comparison between my old Ridgid WD1680 and new Dewalt DXV09-QTA vac.

[Back to Top]

---

Appendix

• Vac: Dewalt DXV09-QTA (9-gallon, ‘5HP’ with 2.5” connection on the body of the vac)

• Primary hose: Milwaukee 49-90-2004 (16-ft, 2.5” connection); I used a 2.5” flex cuff with a hole drilled in it

• Other hoses: 2.5” flex hose with 2-1/4 by 2-1/2 quick coupler to connect the vac to the cyclone, also a Dust Right FlexiPort kit for tool-hookup

• Cyclone: Dust Deputy 2.5

• Misc: Milwaukee 49-90-2030 (flexible, long reach crevice tool)

[Back to Top]

---

 

 

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Dust Collection Cart Mk IIIDIYDPMon, 11 Mar 2024 11:30:00 +0000https://www.minmaxgeek.com/blog/leveled-floor6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:65ef122e13a4c7064c6c75ffWe have our main computer office in the basement, where everything is cooler (and the drop ceiling makes it easier to route cables) but one ‘annoyance’ about our specific basement is massive slope — on two axes. The slope is substantial enough that, not only will my chair roll away, the chair will auto-rotate on its own. This has been bugging me since we moved into this space, and I’m finally getting around to seeing if there’s anything that can be done.

---

Quick Links

• Starting Point

• Rough Design

• Step 1 - Properly support the faux-wall

• Step 2 - Layout

• Step 3 - Cutting tapers

• Step 4 - Torsion ‘box’

• Step 5 - Finishing touches

• Milestone

---

Starting point

I’ve never really paid any thought to whether or not a floor was flat or not (other than the garage) because even in places I’ve lived with a slope, it was a nearly unnoticeable slope.

Our office is in the basement, which is a room you would reasonably expect to have a slope (towards a drain). The previous owner, however put down a floating vinyl floor on top of the basement, without levelling it (and in the process, covering the drain as well — assuming there was/is one).

Where I sit, the floor slopes on two different axes which makes it super fun because not only does my chair roll away on me, it tends to auto-rotate on its own.

View fullsize
Left-to-right slope View fullsize
Even at my desk

Until now, I’ve used a piece of 1/2” plywood, shimmied and padded to sorta-kinda-somewhat keep the chair in one spot. Since it’s not really supported across the span, the plywood has formed a bowl-like depression. Which kind of helps to reduce the tendency of rolling but not-quite — and it’s not a long-term serviceable solution (and all kinds of crud accumulates under the platform).

[Back to Top]

Rough design

There are two major constraints:

1. Whatever solution I implement needs to be ‘renter-compatible’: I don’t want to make this a permanent change — so no direct-concrete, no cutting up the existing flooring. While I would love to just pay someone to come in and re-level the basement, that would be a massive (and intrusive) change - much more than I am willing to deal with at this time (or possibly ever).

2. My computer monitors are mounted to a faux-wall, which in turn, is secured to the wall which means that the monitors are effectively, at a fixed height. As such, there is a limited amount that I can ‘raise the floor’ before the monitors are no longer at the ideal height. If absolutely necessary, I could buy something like a vertical VESA slider (I already have one on my center monitor) but would prefer to avoid the additional expense if possible. On a related note, I want to minimize the ‘raise the floor’ amount as much as possible, also to reduce the inevitable ‘stub your toe' factor.

If I had more verticality to work with, this would be a super easy project: I would treat this like my toe-kick cabinet bases and make a tapered 2x3 or 2x4 frame supported by adjustable hockey pucks and sheathe everything in plywood. Something like this:

View fullsize
Making a leveled base View fullsize
Adding hockey pucks (upside down) View fullsize
Sheathed in plywood

While cutting tapers on framing lumber versus sheet-goods is nearly the same amount of effort, having the mid-spans supported by hockey pucks is a huge time-saver because you can have much larger spans in between framing members — the height can be hand adjusted by spinning up the puck up and down. The gotcha for me is my limitation on verticality:

• 2x3 as framing member (2.5” tall)

• 0-2” for tapered support

• 0.75” for plywood

All this puts me at 5”+ of verticality which I just don’t have the space for. To save on space, instead of having ‘framing + taper’, I could go with just the taper. The tricky bit here is that it’s much less forgiving, and I would have to cut a lot of them — since the tapers would be supporting the weight.

I loosely considered having a flat base, building up tapered walls on four sides and filling the span with playground sand, or small gravel (or even levelled concrete). I’m sure that, once packed, would handle the weight distribution perfectly fine but it was a lot of what-ifs that I was unsure of.

 

---

 

Ultimately, in an effort to reduce verticality, I went with tapers directly supporting the top layer of plywood. Since it would be a nightmare to try and secure the tapers to the underside of that top-layer-of-plywood, I would need to have a bottom layer of plywood as well. Ultimately, this ends up being a large, tapered, torsion box

[Back to Top]

Step 1 - Properly support the faux wall

When we set things up originally, I just jammed some random scraps of wood under one edge of the faux-wall to get it to be level (and secured the faux wall to the actual-wall). The bits of wood sticking out were now getting in my way so I wanted to replace it with something tidier. I used an adjustable levelling foot to take the load and also opted to throw in a ‘just in case’ shim (since the load is now sheared across the screws).

View fullsize
Original support shim View fullsize
Support foot View fullsize
Just in case shim

Since I built the faux-wall, I know that the entire perimeter is framed with 2x4 and the roughly-four-foot span can (and obviously has) taken the load this way.

[Back to Top]

Step 2 - Layout

Once the wall was cleared out, I used a laser to figure out what kind of footprint to make the levelled platform. I did some spot measurements to figure out where and how much to raise the floor in order to get it as level as I could left-right and front-back.

Something to keep in mind is that you need room to roll the chair back in order to get out of it! The amount of front-to-back you need also depends on the depth of your desk as well (my desk is about 24” deep). I originally started with 6-ft (front-to-back) but with some spot testing, I was able to reduce that to about 5-ft.

My left-right span was 6-ft so I ultimately ended up with (roughly) a 6x5 rectangle (since the damn wall comes out by about 1” on my span there).

View fullsize
Figuring out the footprint View fullsize
Self-leveling laser to the rescue View fullsize
Spot testing some shims View fullsize
Testing a taper

[Back to Top]

Step 3 - Cutting tapers

Now the super tedious task of cutting exact sized tapers. I initially started cutting tapers to be placed front-to-back (starting on the right, with the least amount of taper needed); everything was level (front-to-back) but something didn’t feel right. I cut the left-most taper (running front-to-back) which was the biggest taper and then checked level left-to-right and it was way off. This is sort of expected because I have a much bigger slope running left-right than I do front-back.

To correct this, I switched my tapers to run left-right instead. I cut some scraps of plywood to try and get a loosely even spacing of tapered supports. When I was done with roughly cutting and placing the tapers, I threw my level across it to measure diagonally and thankfully, it was reasonably level.

View fullsize
Cutting tapers with the tracksaw View fullsize
Getting started with tapers View fullsize
Tapers all cut

[Back to Top]

Step 4 - Torsion 'box'

Even with super powers: to be able to see through the plywood and be able to lift the entire sheet to place it perfectly atop the tapers, it would be super challenging to attach the top sheet of plywood to the tapers because as you to go to the right, the tapers get smaller and smaller — so there’s nothing to glue/pin/screw to. Even if I were to somehow pull this off, it means that the entire top-plywood is balanced on some vertically aligned strips — which will shear off over time causing everything to collapse.

This means I need to have a bottom plate: the tapered strips can easily be glued + pinned to the bottom plate and then the top plate can easily be screwed to it. By only screwing the top plate, we can remove it for whatever reason if needed. This was an easy step to do, but fairly tedious as I had to switch between multiple sizes of brads and pins.

View fullsize
Pocket holes View fullsize
Starting the layout View fullsize
All done gluing and pinning View fullsize
Final shimming View fullsize
Torsion 'box' done

Like the bottom plate, the top plate is a 4x6 panel + 1x6 panel. I opted not to secure the top plate down — it’s wedged in there between the bottom plate and the faux wall. I may screw it down at some point but for now, it’s accessible just in case I want in there for whatever reason.

[Back to Top]

Step 5 - Finishing touches

By sheer luck, I happened to finalize the platform to be 6-ft wide which was a perfect fit (left-right) for a trailer mat. I was stressing about how to 'dress’ the top plate — whether to use some type of flooring panel, or to have a sacrificial layer of wood that I might coat in some kind of bed-liner or something else.

The trailer mat is great — it’s heavy duty and can handle the point-load of my rolling wheels and actually makes it somewhat ‘difficult’ to roll: I have to make a conscious effort to roll my chair — a very welcome change from the ‘roll whether you want to or not’.

The trailer mat is 5/8” thick and doesn’t cover the full span of the platform, so I did have to find a way to shim the back side, at least temporarily somehow). By sheer luck, I had enough scrap 5/8 plywood matched the height perfectly.

View fullsize
Trailer mat View fullsize
Computer is up View fullsize
Random scraps of 5/8 ply

[Back to Top]

Milestone

For a weekend project, this was great: my chair doesn’t roll anymore and I didn’t have to buy anything (yet). I’m going to get some 5/8” plywood to even out the gap between the trailer mat and the faux wall but that’s not a priority. It’s so nice to be able to just put my feet down without consciously trying to hold the chair still, in one place. When I get the plywood, I’ll probably cut a taper to dress up the front and side edges, just because.

Material wise, this project cost me almost two sheets of plywood for the sandwich and 1/4 to a 1/3 sheet of OSB to cut the tapers. Your mileage may vary depending on how many tapers you cut and how many times you screw up cutting tapers. I did use a whole bunch of Handi-Shims which I can’t recommend enough. There are some gaps still (because of humps and dips in the floor) but I’ll slide some long composite shims to support it.

[Back to Top]

---

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: DIY Leveled-floorDIYTechDPWed, 20 Dec 2023 22:10:00 +0000https://www.minmaxgeek.com/blog/mini-network-rack-mk36020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:658368159c21535d7f345ea6I’ve had my network rack for a couple years now, really about a year in its current configuration and I’ve been wanting to revisit it to address some concerns and annoyances. Since taking everything apart and would be a pain, I wanted to wait until I had a few other upgrades before I went down that road. I’ll cover the networking/technical updates in a future post, but for now, this post will cover the DIY angle of the network rack rebuild.

---

Quick Links

• Starting Point

• Rough Design

• Step 1 - Switching out the base

• Step 2 - UPS enclosure

• Step 3 - Handle bar

• Step 4 - The rear patch panel

• Step 5 - Making my own rack mount

• Step 6 - POE Injector mount

• Step 7 - NAS shelf, finishing touches

• Milestone

• Mixups, Surprises and Lessons Learned

• Appendix - Parts and equipment

---

Starting point

Before embarking on an upgrade/rebuild, I wanted to highlight my top three favorite features of the Mk II rack

• All of the connections were self-documenting: by using CordIDs, I can unplug any and all of my cables and know at glance which device and which port (on the patch panel) it should be plugged into.

• Almost all of the components have individual (analog) power switches associated with them. Being able to manually power down elements of your network is incredibly handy if you’re dealing with a power outage scenario (and need to cut the power footprint to maximize UPS runtime) and it’s also way faster to just walk up to a device and toggle the power switch versus logging into a management interface and telling the device to reboot

• The steel side panels allowed me to magnetically attach everything to the rack (both on the inside and outside surfaces). The top-plate is also magnetically attached, so it’s super easy to pop the top off and I can then snake around inside the rack

Of course it’s not without some downsides.

• I had some doubts about the stability of using a washing-machine mobile base to support everything on a very sloped floor. Specifically, I was concerned about flexing on the base plate

• I really, really hate the second PDU I added to the rack. On a day-to-day basis, it bothered the hell out of me seeing the 1.25U PDU which then reminded me of the dumb design where rack-mounting was clearly an afterthought.

• So all of my cables are connected to the patch panel through the back side (like they would be in a traditional rack). But as I use this more as a cart, this is problematic: [1] anytime I want to swap something out, I have to snake my hands around on the inside to disconnect the device and [2] if I want to disconnect the entire cart, I have to take everything apart for the same reason

[Back to Top]

Rough design

I had three broad considerations I wanted to keep in mind for the new cart:

• Redo the base to remove any concern about sagging or flexing in the base plate

• Try to keep the overall height of the network rack down if I can - I park the network cart in front of a cabinet with the breaker panel and it would be fantastic to be able to access the breaker panel without having to move the cart

• Make room for two UPS units and where possible, consider any room for a future NAS unit

View fullsize
Option 1 - Minmalist View fullsize
Option 2 - Going wide View fullsize
Option 3 - Embrace the cart

I really liked the idea of Option 2 but it would definitely make connecting the network elements together tricky — even if I punched a hole through the two sides connecting them. I ultimately went with Option 3 because I wasn’t sold on being able to fit everything I wanted now (and in the future) into a small rack.

The big change would be adding rear-facing patch panels and permanently connecting them to the front-facing patch panels. This way, I don’t have to snake around inside the rack whenever I want to remove a device. It turns the back side of the network rack into a dock.

[Back to Top]

Step 1 - Switching out the base

The previous base was a panel of 1/2” plywood that I ‘kind of’ clamped onto an adjustable base and while I don’t have any overt reason not to trust it, over time, I just ‘had a feeling’ that due to the uncentered weight was causing the wood to deflect and eventually result in tears.

To replace this, I had a universal tool base that I’ve had kicking around for a long time (and truthfully, have been tripping over). To try and get ahead of deflection, I opted to make give the base a bit of a torsion-box. Being designed as a mobile base for heavy workshop tools, it was easy to bolt this platform to the base.

View fullsize
Loosely assembled the mobile base View fullsize
OSB base View fullsize
Trim off the tongue/groove View fullsize
Clean reference edge View fullsize
Quick sanding View fullsize
Quick test fit View fullsize
Marked the drill holes View fullsize
Rummaged around for scraps View fullsize
Cutting to length View fullsize
Roughly visualizing the placement View fullsize
Using tape for alignment View fullsize
All glued down View fullsize
Added a threaded bolt View fullsize
Studs mounted View fullsize
Test fitting the 'pretty' top View fullsize
Marking the center View fullsize
Quick notch View fullsize
Quick coat of stain

For actually bolting the platform to the mobile base, I used 2” machine screws and star knobs. At some point down the road, I may swap them out for a pair of nuts but for now, the star knobs are great for being able to quickly get the board on and off.

[Back to Top]

Step 2 - UPS enclosure

I laminated some 3/4 ply together to make some side panels; the idea is that the side panels of the rack will rest on these, transferring weight to the base. I connected a couple quick stretchers to keep the two panes co-planar. During test fitting, I realized that one of the braces blocked the power cord for the UPS so I drilled some access holes to be able to feed the plug through it.

View fullsize
Laminated some scraps View fullsize
Glue and pocket screw the stretchers View fullsize
Test fitting View fullsize
Marking clearance hole View fullsize
Hole saw to cut access holes View fullsize
It fits! View fullsize
Test placement

[Back to Top]

Step 3 - Handle bar

One of tricky things of the old base is there was no easy spot to grab anything in order to move the cart. I wanted something I use to manoeuvre the cart around. I was going with a pair of wall-mount racks and wanted a solid, physical way to transfer the weight load to the ground — I am not going solely rely on anchoring the racks to my vertical back posts — I would prefer to use that more for ‘alignment’.

I raided the scrap bin to get some plywood to make a ‘bushing’ to separate the two racks - yet provide a continuous, solid contact surface to transfer weight down.

View fullsize
Scraps View fullsize
Gluing together the stretchers View fullsize
Adding a front handle bar View fullsize
Test placement

As you may have noticed on the previous step (in the final picture), the rack has a support brace on the top side — which meant that I had to cut a small relief on the side strips so that everything would lay flush. After connecting up the handle bar, I ran a round-over bit make the handle bar aspect of this glorified spacer, nicer to touch.

I didn’t get any photos of it but using 3/4 OSB, I laminated together a frame (like in my previous rack build). Even though the weight of the rack is directly transferred downward, the frame gives me something to bolt everything to, if only to keep the rack from sliding. I had to wait until I had the handle bar together so I could stack the second rack — in order to get the final height needed for the back support frame.

[Back to Top]

Step 4 - The rear patch panel

I bought a small 2U wall-mounted rack in order to mount the back-side patch panel. I had originally had a vision of recessing this into the frame (somehow!). By recessing this into the frame, the front and the back patch panels would be closer - meaning I could use 2-ft cables to connect them. Logistically, I couldn’t think of a clever way to recess the panel (without adding more height), so now, the back patch panel was 4-inches further away than I had originally intended, which was a bit of a problem since I had already bought 2-ft cables…

Starting point
Test install
Swings open
Easy access to the back side of the panel
Initial attempt
Getting the cables ID'd up
Starting to feed the cables through
I gave up
Front side hooked up

[Back to Top]

Step 5 - Making my own rack mount

Since it was just a matter of a couple inches of clearance (and I had nothing to lose), I took a side quest to try and make my own rack-mount. Using scrap plywood and the rack-mount I already had handy, I carefully marked out the hole locations and then drilled out the holes using a drill press. I debated using threaded inserts to then hold the bolts captive (turning them into studs), but I didn’t want to risk tear out with the holes so close and honestly, for a 2U rack like this, it’s not a big deal.

The custom rack route worked great, allowing me to get the back side of the rear patch-panel close enough that I could use my 2-ft cables . Of course, had I anticipated this, I would have considered 3-ft cables and avoided all of this drama to begin with.

View fullsize
My reference View fullsize
Playing with scraps View fullsize
Binding the two pieces together View fullsize
Drill! View fullsize
All done View fullsize
Predrilling View fullsize
Perfectly set screws View fullsize
All assembled View fullsize
Mounting the rack View fullsize
Mounted! View fullsize
Relabeling the ports View fullsize
Pause View fullsize
Tidying up the front

[Back to Top]

Step 6 - POE Injector mount

Previously, I had mounted the POE injectors one at a time, independently because, for my main AP, I was just using a POE injector I keep around for testing purposes, not the injector I wanted to use long term. This was a good opportunity to get around to getting a dedicated injector and mounting everything on a common plate. I kept the injector mount for the low-power AP separate because [1] ran out of room on the mounting plate and [2] thematically, it’s kind of its own thing anyways. I did opt to directly screw the low-power AP-injector into the back frame though.

Scrap plywood
I got ahead of myself here
Marking out where the magnets will go
Saved by the drill press
Laying out the mounting plates
Drilling a cable pass hole
Opened the hole
Permanently mounting the low-power injector
Ducks in a row
Everything in place
Tidying up the side
Ordered dedicated injector

[Back to Top]

Step 7 - NAS shelf, finishing touches

For the NAS shelf, I opted to dismantle my previous two-tier shelf and reuse it here, rotated sideways. I also opted to route a power bar for the NAS units so that I wouldn’t have to snake the around inside the network rack to access the power. The extra few inches to get to power doesn’t sound like much, but I have an upcoming change with the way power is handled in the rack so I wanted to get ahead of it this way.

While I was here, I took the time to finally mount everything down to the base plate! Until now, everything was just relying on mass/friction to not really move. I had taken the time to drill pocket holes to secure things down but due to a comedy of errors, every single one of the pocket holes had been covered up by something else and I was just too lazy to undo it — so I went with some angle brackets I had lying around.

View fullsize
Test fitting View fullsize
Securing the network rack to the base View fullsize
Cable pass View fullsize
More angle brackets View fullsize
Mounting a power bar View fullsize
Finalizing View fullsize
Capping the shelf

[Back to Top]

Milestone

This was a fairly quick project as I was able to leverage existing pieces (the NAS shelf) and the fact that all of the cables are individually labeled (for both ends) made it easy to be able to switch between “powering everything down to work on it” and “okay we need the network up because the weekend is over”.

I have to thank my wife for putting up with zero-network-at-all for hours at a time while I tinkered with this. Oh, and the massive puke of cables and tech that sprawled in the basement as I took everything apart ha-ha.

In the near future, I might buy a pair of Noctua 200mm USB fans to act as top-exhaust. For the time being (and for the last few years) I’ve not had any temperature issues with the rack mostly-closed up. If there ever is a heat issue, I can simply remove the top plate as it is just magnetized down to the rack.

At the moment, I have nine empty bays which sounds like a ton but I imagine once I take a closer look at what I can rack mount, they will fill up fast.

[Back to Top]

Mixups, Surprises and Lessons Learned

There were three notable screw-ups

1. I intentionally drilled some pocket holes so that I could secure the UPS enclosure down to the platform. I also drilled pocket holes on the back support from (which is secured to the UPS enclosure) to secure that to the platform. I managed to face the pocket holes for both into each other and thus, had to resort to using angle brackets to actually tack the network stack down

2. For the UPS enclosure, I really should have opened up the cable pass through holes with a bandsaw/jigsaw to reduce future headache. As it stands, I can get power cords in and out of the enclosure but it’s a bit annoying.

3. The last careless mistake was not drilling the holes for the magnets before assembling the shelf for holding the POE injectors. Another careless oopsie.

The only other hurdle was the issue of the rear rack-mount using up too much of the limited 2-ft length span I had planned for the wiring, something that could have been avoided with 3-ft wiring or just DIYing the rear rack.

 

---

 

For the mobile tool base, I’m simultaneously really happy and really annoyed with it: it’s quite sturdy, well built and I love the fact that it has adjustable sizing but I hate, hate, hate the fast that the back wheels are fixed. I have this same hatred for rolling tool chests as well. Having fixed real wheels isn’t an issue in a large automotive shop but in any smaller/home scenario, I want to be able to shuffle and snake the tool chest (or network cart) in small tight spaces. Instead of just scooting the cart over, I have to do the game of snaking it back and forth incrementally. Honestly, if I see the mobile base go on super sale again, I might legit consider buying another one just to cannibalize the front wheels to mount on the back of this network cart. Totally infuriating.

 

---

 

Regarding the rack-mount cable concealer, I’m kind of torn. On the one hand, I really do like that it gets the random jumble of cables tidied up but at the same time, I really miss being able to directly see the CordIDs for each cable. Thankfully it comes with a quick-snap panel, so if I needed to see the label, I could do so quickly enough so for now, this is just an irk for me.

[Back to Top]

---

Appendix - Parts and equipment

• Racks

  • StarTech 12U Wall-mount Rack. Since I already had a 12U, it made sense for me to go with another one as one of my candidate designs had them stacked side by side (thus, requiring the same rack). The only thing to note is that the brace bar on the top can interfere with all but the most shallow devices (read: small PDUs). In my case, I was planning on leaving the top as an air-gap anyways, but something to be aware of. Otherwise, fantastic rack mount, no complaints.

  • StarTech 2U (Hinged) Wall-mount Rack. Even under the hypothetical assumption that I had bought 3-ft cables and could use this, I’m kind of torn on this product. Like it’s fine, it does the job but the hinge is not confidence inspiring - just having the two patch panels hanging off the hinge has some serious deflection. In hindsight (and unique to my scenario), considering how easy it was to make my own 2U rack, I think I might give something like this a pass.

  • ETS 24-port loaded patch panel. This is one of the more cost effective units and comes with everything ready to go out of the box

  • Rackstud Series II Duo. An alternative to cage nuts, this is totally worth every penny. The ‘Duo’ only works for 1U equipment (which by coincidence, all of my gear is) but it’s a massive time and frustration saver. Having studs (instead of nuts) allows your to temporarily hang your equipment off the stud while you reach for nuts, instead of playing the balancing act of trying to support equipment whilst lining up holes.

  • Monoprice SlimRun CAT6a cables. Available in a variety of colors and lengths, I went with these to try and keep the cable mess under control. Note that for POE devices, anything ‘downstream’ of the power-injection source (whether it be an injector or a POE switch), you should try and use non-slim cables — for better heat management.

• Organizing

  • CordID tags. I use these extensively to label every single cable, knowing exactly where to plug (both!) ends of a random cable is pretty awesome

  • 1U network cable organizer. As I mentioned above, I’m not entirely sold on this for my specific use case (where I want to see the CordIDs) but for most people, I imagine this will be great. I might suggest a 2U organizer as it allows for a bit more flexibility (and maybe even the possibility of using CordIDs) — but for my specific application, I only had 1U available. The front plate can swivel from the top or the bottom and can be removed so it’s actually quite a pleasure to work with.

  • Bulk Velcro. I bought a huge roll of bulk Velcro to keep things my cables all grouped together and organized where possible

• Magnets

  • 130lb rare earth magnets. I use these magnets to make magnetic mounting plates - the solid steel surface of the wall rack gives me plenty of area to mount everything

  • 25lb magnet hooks, I use these to gently route cables around on the rack and take up slack. These generally don’t need to have a ton of holding power as you’re just gently routing things, so finding magnet hooks at your local surplus store will be fine too

• Power

  • ADJ Pow-R Rack. This rack has it all — switched power outlets, two always-on power outlets (front + back) and front-USB

  • Pyle Pro PDBC70. This is a more basic PDU that just works and provides nine switched rear-outlets

  • Splitter extension cord. You can get them in 2-split and 3-split and these are great for when you have just a few items you want to power on/off together. In my case, my main modem, fiber-switch (allowing me to toggle between main-router and low-power router) will always be powered on together, so I can use this splitter to link them to the same port on the PDU

  • Short power cords: 1ft ‘outlet saver’ extension cords, C13 ‘PC power’ cords and C7 ‘figure-8’ unpolarized power cords. One way to reduce the amount of cable management and wrapping needed is to literally reduce the physical amount of cable required to hook things up. Shorter cables allow you to get near exact-length power runs.

• Mastercraft Universal Tool base. I happened to use the Mastercraft model as I had bought one years ago on super sale but there are any number of adjustable mobile tool bases that might be suitable alternatives. I did see that the Portamate PM-3550 comes with all-swivel wheels but that price tag is something else — you could totally buy two bases and cannibalize parts for less. You can also easily put one together using just four swivel locking castors as well.

• A small 4V screwdriver: I have an older version of this 4V Black & Decker — makes working with rack nuts/screws so much less annoying as otherwise, I’m just sitting there spinning screws forever. Even since switching to rackstuds, the low power of this screwdriver means I can just cinch the nuts down on top of the RackStuds without concern that anything is going to break or strip.

[Back to Top]

---

 

 

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Mini Network Rack Mk IIIDIYDPFri, 01 Dec 2023 13:50:00 +0000https://www.minmaxgeek.com/blog/shop-improvements-2023-26020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:655bb7046ea91968241120baAfter getting the overhead blade-guard mounted dust collection on my saw in the previous update, I was totally blown away by just how much dust could be directly captured at the source — reducing the amount of time and effort spent cleaning up at the end of a project. I’m not a big fan of the blade guard assembly on my saw but the effectiveness of the aftermarket blade-guard mounted dust collection was just so compelling, it was worth using — at least for a few months.

In that time, of course, I wanted to keep an eye out for a different, over the blade mechanism to handle dust collection. Recently, I found a compelling option so it was time to revisit dust collection in the shop.

---

Quick Links

• Table saw dust collection

• Starting point: blade-guard dust collection

  1. Why the odd requirement?

  2. Installation

  3. Thoughts

• Miter saw dust collection

• New dust hose

• Ceramic (and poly) all the things!

---

Table saw dust collection

Starting point: blade-guard dust collection

In my previous update, I got around to installing over-the-blade dust collection I got from Etsy and the results were/are incredible - I can comfortably say that ‘nearly all’ the dust was captured. For most types of cuts, it’s easily 99% — the only dust that has a chance of escaping comes right at the end as the blade comes out of the trailing edge of the board.

  View fullsize
Just before View fullsize
Just after  

While the dust collection performance is phenomenal, having the guard present, can be a nuisance at times:

1. Whenever I’m performing a non-through cut, I had to take the guard off anyways. Removing and replacing the guard is, a huge pain in the butt, regardless of how “quick and easy” the table saw instructions tell me

2. For some/many types of cuts, I would really like to use something like the MicroJig Gripper but with the bulky guard in place, it’s difficult/impossible, so again, I end up removing it to do the special cuts

3. If you’re making very thin (but long) rip cuts, say 1/8th or 1/16th — fragments of your cut can get sucked up into the chute and get ‘stuck’ requiring you to stop what you were doing and dislodge it before continuing

Despite all of these annoyances (none of which were really the fault of the dust-collection adapter), the phenomenal dust collection was totally worth the hassle of working with the guard. I did keep an out for some form of dust collection that ‘come down from the ceiling’ — without wrapping around the right hand side of the saw like nearly every overhead dust arm ever.

[Back to Top]

Why the odd requirement?

Nearly 100% of the time, when you look for “overarm dust collection” you get a ‘thing’ that

1. Hangs over the blade but then

2. Pipes/routes to the right side of the table saw, often down low. Some models will (after going right), snake back left to hook up to the dust port on the saw

3. In many cases, they will bolt to the right side of the saw in some way as well

My dust collector happens to be on the left side of the saw (and I have hose running across the ceiling and dropping down) so the idea of having piping run right (and then down), only to then ultimately go left (or up) was silly to me. Many of the options also felt like they were saw-specific (I’ve love to one day, upgrade this saw) or at the least, required me to commit to a specific rip-capacity (I would totally consider an increased rip capacity) so I didn’t want something that would hamstring me in the future.

Ultimately, I wanted an overhead guard where the piping goes up then connects left, or even have the piping go left a bit then up (or down). But no, I’m pretty sure every option on the planet has things going right (not to mention, very few options tie to the ceiling).

I did find one really neat option, the PSI TSGuard can be mounted to a ceiling. There is a downside: the dust collection guard is built for 2.5” hose and more importantly, I got Canada’d. I had almost resigned myself to the fate of having to DIY something when I stumbled upon an option by Centurion.

Canada’d

A condition that happens when a product is either unavailable or delayed in Canada, or costs three times more, or comes with a ridiculous shipping cost.

Or more than one of the above.

It’s not exactly what I wanted (a pipe going straight up, or left-only) but it was the closest option I found that I could actually order.

[Back to Top]

Installation

I have a bulkhead running across the ceiling over the table saw and with foil-lined insulation, there was a whole lot of ‘who knows what is behind the drywall’. Installation went in three stages

1. Mounting some plywood to the ceiling — this would allow me to secure piping to the plywood and not have to fret so much about whether I was mounting to wood or not,

2. Some kind of a vertical drop-down bar that I could mount the guard to and

3. Replacing the dust-hose with rigid piping and tidying everything up

Since this was a relatively big change, I took the opportunity to swap out the ribbed hose for straight pipe where I could: I left the final connection to the dust collector as a flex-hose so that I could easily disconnect the dust collector

View fullsize
Starting with the first link of solid pipe View fullsize
Couplers, couplers, couplers View fullsize
Testing with some extensible hose View fullsize
Swapped out for flex hose View fullsize
Cutting out a relief for the ceiling (ha) View fullsize
Lego View fullsize
Adding in an additional plate View fullsize
Switching to flex cuffs View fullsize
Lining up the guard View fullsize
Recessed bolt View fullsize
The drop beam View fullsize
Mounting plate View fullsize
Temporary holding View fullsize
Final leg View fullsize
Solid piping View fullsize
Straight pipe continues View fullsize
Blade guard, mounted

[Back to Top]

Thoughts

After getting the guard initially mounted, I did some rough testing (using an excessively long length of flex hose); compared to what I was used to with the guard-mounted collection, the dust collection was promising but not great. Once I finished the transition to straight pipe, I spent a bit of time tuning the pitch and elevation of the guard and as you can see, there’s a substantial improvement.

View fullsize
Ribbed hose, just before View fullsize
Ribbed hose, just after View fullsize
Straight pipe, just before View fullsize
Straight pipe, just after

I have a few complaints about the dust guard:

• It doesn’t come with any instructions or anything. There’s a lot of custom stuff you have to do to mount this guard so I understand being light on the hand-holding but I was hoping for a bit of guidance on setting the elevation of the mount (i.e., how high to mount the unit above the table saw surface)

• This is an extremely ungainly unit — this is something you find out the hard way when you’re trying to visualize how high above your table you might want to mount the guard. So you get to play a balancing act with random boxes (or jack stands) to try and guess the elevation

• Although this is only 13lbs, it feels needlessly heavy (magnified by its unbalanced nature). Was it necessary to use steel for everything? It’s not like the tubing is under immense crushing pressure or anything; I feel like this wouldn’t have been such a pain to balance (despite the lopsided weight distribution) if it was half the weight

• The use of 3” hose and 3” tubing is dumb. All it means, is you now need to have an adapter to bring it up to 4”. The adapter was also incredibly loose. There is a screw that use to tie the two together but was it that much of an ask to get a better friction fit? I ended up using a couple pieces of foam tape to take the play out

All of the above issues mostly arise while you are installing the guard. I did have a few grips that persist even after everything is installed:

• The blade guard uses a parallelogram arm-linkage; being a parallelogram, it’s difficult to introduce a pitch in the actual guard i.e,m to have the trailing edge be angled higher than the leading edge (to make it easier to move material past the cut). This is a minor gripe which may go away as I get more comfortable with the guard

• Adjusting the tension on the linkage is obtuse: you have to skewer a hex key through the back side whilst holding a nut captive with a wrench. Is it too hard to make this lockable with a front/top mounted screw?

• The guard provides a clamping knob that lets you lock the guard in the down position — but there is no way to lock it in the up position. You can move the guard out of the way but once you let go, it drops in down a bit

• For my specific use case, the tube running left-right seems wider than necessary. For my use case, I was hoping to have the least amount of right-ward travel from the guard to the dust port

I think the dust collection from this over-the-blade solution is really close to the performance from the blade-guard solution (with a small advantage going to the blade-guard) but there’s a substantial advantages in usability that can’t be ignored:

• There will always be cuts you simply can’t use dust collection for; being able to retract the chute out of the way (or just moving the saw) is super straightforward

• For those that need to see the cut as they make it? The clear guard here is a winner

• This will work with dadoes!

• After a cut is complete, being able to gently tug the guard to drop it to the table (boosting suction) collects nearly all the dust that may have escaped

• This system will be universal to any table saw I might have

[Back to Top]

Miter saw dust collection

I build my current miter station with dust collection and containment in mind: the saw is enclosed in a box that tries to keep any dust that isn’t collected by the saw from spilling. First order of business was to take the front covers off an do a clean. This was the first time taking the covers off in a year or so (although I do regularly shove a vac nozzle back there to try and keep things tidy), and we can see the containment aspect is working pretty well.

Left
Right

Since I had everything disassembled, I used the opportunity to laminate the hardboard with some packing tape; the hope is that it makes the surface easier to blast with compressed air. I spent a bit more time this time to (very generously) seal the innards with duct tape to reduce the chances of dust falling sneaking out.

View fullsize
Packing tape View fullsize
Reinstalling the plates View fullsize
Getting started with taping

To connect my saw to a vac, previously, I ran a 2.5” flex hose, from below, up through the table, to the saw. This obviously, requires a hole in the table for the hose to go through. I was extremely careful in drilling the absolute smallest hole I could that would allow the hose to pass through but inevitably, some dust does sneak through. Since I had everything apart, I wanted to see if I could seal up the gap.

I bought two 2.5” dust ports that would sandwich the hole, removing any chance for dust to sneak through. Additionally, this allows me to decouple what happens ‘above the table’ from below the table’.

View fullsize
Not quite enough clearance View fullsize
Oscillating tool to the rescue View fullsize
Just enough clearance View fullsize
Hose, attached View fullsize
Repeat above the table

Doing this brings two usability improvements:

1. I no longer needed to have bits of plywood in the dust box to control the hose (above the saw)

2. When I need to empty the cyclone, I can just disconnect the cyclone without having to worry about hoses flopping and falling all over the place

[Back to Top]

New dust hose

Since doing a minor rebuild of my dust-cart, I found that the cyclone that that I initially thought would be ‘good enough’, in fact, wasn’t. It does pre-separate but I was finding it was closer to 30%. Since I use that cart as my primary “clean up cart”, I switched to the more familiar Dust Deputy 2.5. I initially retrofitted a generic bucket lid to work with this but found it just wasn’t sturdy enough (even with a plywood backing), so I opted to get a reinforced lid which, as expected, worked perfectly (you can also get everything as a single kit). Compared to the previous iteration of the dust deputy (which I have on my miter saw), this unit has a higher airflow.

I’ve been using the new cyclone setup for about a year now and two things have really started to bother me over the time:

1. How utterly useless the stock vac hose is and

2. Why it is so damn hard to find a nice, 2-1/2” hose without spending an absolute fortune

With my cyclone mounted above the vac, the hose starts off about five feet off the ground; with a stock seven foot hose, I’m constantly having to drag the vac-cart around with me wherever I wanted to clean. So I knew I wanted something that gave me a bit more than the seven-foot reach I had. So I went through some options:

	A	B	C	D	E	F	G	H	I
1	Brand	Model	Hose Length	Hose Diameter	Connector A ("Tool End")	Connector A Removable?	Connector B ("Vac End")	Connector B Removable?	Notes
2
3	DeWalt	DMV9315	15	1-1/4	Universal [DeWalt] tool connection	Yes	DeWalt dust-extractor connection	Yes	Advertsied as crush-proof
4		D279058	14	1-1/4	Universal [DeWalt] tool connection	Unsure, likely yes	DeWalt dust-extractor connection	Unsure, likely yes	Advertsied as crush-proof
5		DMV9316	15	1-1/4	Universal [DeWalt] tool connection	Yes	DeWalt dust-extractor connection	Yes	Antistatic
6		DMV9314	8	2-1/2	2-1/4	Unsure, likely yes	DeWalt dust-extractor connection	Unsure, likely yes
7		DXVA19-2501	7	2-1/2	2-1/4	Unsure, likely no	2-1/4 with locking tab	Unsure, likely no	Product page for the DXV12-QTA states this is the included hose, but no product page exists
8		DXVA19-2500	7	1-7/8	1-7/8	Unsure, likely no	1-7/8 with locking tab	Unsure, likely no	Product page for DXV09P-QTA states this is the included hose, but no product page exists
9		DXVA19-2556	7	2-1/2	2-1/4	Unsure, likely no	2-1/4 with locking tab	Unsure, likely no	Product page for DXV12P-QT states this is the included hose but no product page exists
10
11	Dustless Tools	14251	12	1-1/2	2-1/4	Unsure	2-1/4	Unsure	Advertsied as Crushproof, kinkproof
12		H0946	25	1-1/2	1-1/2	Unsure	1-1/2	Unsure	Advertsied as Crushproof, kinkproof
13		14291	25	1-1/2	2-1/4	Unsure	2-1/4	Unsure	Advertsied as Crushproof, kinkproof
14		H0038	25	2	2-1/2	Unsure	2-1/2	Unsure
15
16	Rigid	VT1720	7	1-7/8	1-7/8	No	1-7/8 with locking tab	No
17		LA2570	10	1-7/8	1-7/8	No	1-7/8 with locking tab	No	Advertised as flexible, kink-resistant
18		LA2522	20	2-1/2	2-1/4	No	2-1/4 with locking tab	No
19		VT2534	10	1-1/4	1-1/4	No	2-1/4	No	Advertised as flexible, kink-resistant
20		LA2520	7	2-1/2	2-1/4	No	2-1/4 with locking tab	No
21		VT1721	7	1-7/8	1-7/8 with air relief valve	No	1-7/8 with locking tab	No	Expandable hose, collapses down to 2ft
22		LA2521	13	2-1/2	2-1/4	Looks like yes	2-1/4 with locking tab	No
23
24	Milwaukee	49-90-2006	9	2-1/2	2-1/4	Yes	2-1/2 with nubs	Yes	Advertised as flexible, kink-resistant and antistatus
25		49-90-2004	16	2-1/2	2-1/4	Yes	2-1/2 with nubs	Yes	Advertised as flexible, kink-resistant and antistatus
26		49-90-1984	16	1-7/8	1-7/8	Yes	1-7/8 with nubs	Yes	Advertised as flexible, kink-resistant and antistatus
27		49-90-1996	9	1-7/8	1-7/8	Yes	1-7/8 with nubs	Yes	Advertised as flexible, kink-resistant and antistatus

The only real candidate from my quick survey was the 16-foot Milwaukee hose which hit all the checkboxes: good length, 2-1/2 hose, antistatic, premium build. I didn’t actually notice the nubs on the hose until after I got home. The nubs are on designed (presumably) for the hose to securely latch onto the Milwaukee vacuums however, I was hoping to connect this hose to my cyclone which was roughly the same size so it would not fit.

Initially, I bought a 2-1/2” flex cuff to clamp the Milwaukee hose to the cyclone

View fullsize
MIlwaukee vac hose View fullsize
Flex cuff to the rescue

Since the hose ends (both) can be unscrewed, ideally, I could just buy another standard 2-1/2 end and use that but the hose end does not appear to be purchasable as a standalone product. I thought perhaps I could accommodate the nubs my drilling into the flex cuff and that worked brilliantly.

View fullsize
Marking two opposing points to drill View fullsize
Test hole drilled (21/64") View fullsize
Slipped over the nub View fullsize
Secured with clamp View fullsize
Slipped onto the cyclone

The flex cuff is a bit of a tight fit, so if I ever want to pull it off, it will take me a moment to do so but overall, I don’t think I’ll really need to remove it often. With the massive 16-ft reach, I can clean the shop without constantly having to move the vac cart around (and inevitably backing the cart over the power cord which is infuriating).

Compared to the 7-ft Ridgid hose I had before, this is a dream to use, not just for the extended reach but also for it being anti-static. The only notable downside is that everything about this hose is heavier and I will need to be mindful that repeated drops from operating-height to the concrete floor will likely end up with a cracked connector. Also, the additional hose can be a bit unwieldy. These are minor quibbles to put up with for the vastly improved experience.

[Back to Top]

Ceramic (and poly) all the things!

After going through all the effort to improve the dust collection efforts throughout the shop, I did one more final step of craziness: I coated all the surfaces in the shop:

• Metal surfaces were ceramic-coated: I did all the major metal surfaces (tool chest, table saw, dust collector etc.)

• For the wood surfaces, I applied a coat or two of polyurethane and then sanded to 800-grit - I did all the drawer faces and surfaces of the cabinets and miter saw station and the garage door

The hope here is that dust will be less likely to stick to things and when it does, I can just hit it with some compressed air and not have to bust out the brush or damp rag.

View fullsize
Ceramic'd View fullsize
Sealed and cleaned View fullsize
Cleaner than new View fullsize
Ceramic on the steel supports View fullsize
Polyurethane on the wood door View fullsize
Tool chest

I also opted to ceramic coat additional items and surfaces like the rigid piping for the over-the-blade dust collection as well as the plastic scoop I use to empty out the cyclone. Anything to try and keep things dust free…

[Back to Top]

---

 

 

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Small Shop Improvements (2023.2)DIYDPTue, 31 Oct 2023 12:00:00 +0000https://www.minmaxgeek.com/blog/drill-rack-mk-ii6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:653fc2390eda0e1521b2106dSomething I’ve procrastinated since moving into this shop was making specific shelving for this shop - I’ve made do with shelving that I brought from previous shop(s). This weekend I took some time to make a new drill rack - it was an opportunity to use up materials cluttering up the shop and I was hitting the capacity limit of my previous drill rack.

---

Quick Links

• Starting Point

• Goals and Objectives

• Rough Design

• Step 1 - Cutting strips of plywood

• Step 2 - Determining the spacing

• Step 3 - Batching out the T’s

• Step 4 - Attaching the top

• Step 5 - Attaching the back

• Step 6 - Finishing touches

• Milestone

---

Starting point

I built my previous drill rack back in 2015 and it has held up great in the interim years. As I add more tools, it’s time to take a second look at the drill rack to both increase capacity but also to change the design to be better suited for this (more permanent) space.

[Back to Top]

Goals and objectives

When I built the original drill rack, I didn’t have the drawers/storage that I do now — so I kept all the extra drill related stuff on it. This facilitated the design of having the high side walls to keep things from falling out. In this shop, I have drawers dedicated to holding drilling accessories and screws so I having a more streamlined design will allow me to better use the limited vertical space that I do have.

I wanted to make the top of the drill station flat without side walls so that I can use this as a shelf of sorts to store tools that I don’t use often. As such, I want the top to be relatively wide - which is great as it lets me support hanging more tools than my current setup. As a rough starting point, I’ll target something around 10 slots and can go from there based on the materials I have on hand.

[Back to Top]

Rough design

I debated on going with an adjustable drill rack using a pair of t-tracks to allow the dividers to slide to fit whatever it is you’re holding but [a] I didn’t have any t-track on hand and [b] in the eight years I’ve had the previous drill rack, not once did I feel like i could have benefited from adjustable width.

The basic design will be based around a collection of inverted T-shapes.

There are three measurements that I need to settle on

1. The space between the inverted-T bars (red) — this will balance the total number of tools you can hang for a given width vs the size of the tool you can hang in the holder. For example, a big impact wrench will require a lot of space vs a compact driver

2. The height of the vertical portion (orange) — most drills and drill-shaped tools will fall within a fairly tight spread of heights (measured from above the

3. The amount of overhang (purple) — this will be somewhat tool specific; some tools will have a more or less of a deviation between the width of the handle (usually the narrowest part) and the width of the motor housing (usually one of the bulkiest parts).

There is one more additional measurement (not shown), which is the overall depth of the drill rack: a deeper rack will allow you to hang drills with longer bits (say, pocket hole bits) attached to them.

[Back to Top]

Step 1 - Cutting strips of plywood

I settled on a front-to-back length of roughly 13-5/8” — this would allow me to fit even the longest drill, chucked with a pocket-hole bit comfortably. I gathered up a bunch of scrap plywood and batch cut that.

View fullsize
Plywood scraps View fullsize
All batched to common width

You kind of have to have a rough idea of the dimensions of your inverted-T so you know roughly how much plywood to set aside. I started with a plan to use roughly 3-3/4” for the verticals (red) to start.

[Back to Top]

Step 2 - Determining the spacing

I opted to do the final two pieces (the left and right book end) separately so for now, I’m focused on getting the spacing for the inverted-Ts. To figure this out, I cut some scraps to a bunch of sizes so I could play around with it. I marked the size on the scrap directly to avoid mixing them up. I made spacers for both the gap between the T’s (red) and the overange (purple). I also switched to millimeters here to make it less frustrating.

View fullsize
Gapping the T's View fullsize
Setting the gap for the top View fullsize
Test fitting View fullsize
Plenty of room

There is not a lot of wiggle room with the gapping (for the handle) and the position of the lockout selector can cause the tool to tilt one way or another. This doesn’t damage anything (and my previous rack did the same thing) but this… bothers me, so this led me to play with the spacing a whole bunch to get it so things hung straight

 

 

In the end, for my specific tools, I went with the following sizes:

[Back to Top]

Step 3 - Batching out the T's

Now that I’ve figured out the sizes, I can rip the plywood down to the required sizes and get the pieces ready for assembly.

View fullsize
Sample routing View fullsize
Making a reference mark View fullsize
Quick sand before assembling View fullsize
Assembly jig View fullsize
Story stick for marking screws View fullsize
Pre-drilled View fullsize
Final assembly View fullsize
Batched out View fullsize
Made the end pieces

[Back to Top]

Step 4 - Attaching the top

I have a top plate and a back plate to connect all of these inverted-Ts together. Due to the place I was planning on placing this, I opted to have a bit of an overhang with the top-plate whereas the back-plate would be sized to fit. Since I was planning on having overhang, I split this assembly into two phases: the first phase, I would get everything gapped correctly and use the top-plate to roughly tie everything together. Once everything is tied together, I would cut the back plate to exact size. Before assembling, I stained all of the surfaces that would be a pain in the butt to get to after assembling.

View fullsize
Batching out spacers View fullsize
First dry assembly View fullsize
Splitting a piece of plywood View fullsize
Quick stain all of the surfaces that would be a pain to access View fullsize
Staining the top plate View fullsize
Dry fitting #2 View fullsize
Squaring everything up View fullsize
Clamping as I go View fullsize
Marking the screw locations

[Back to Top]

Step 5 - Attaching the back

Now that everything was tied together, I cut the back plate to exact size and screwed it on along with the French cleat and standoff.

View fullsize
Diligently pre-marking the screw locations View fullsize
French cleat added to the top View fullsize
Standoff added to the bottom View fullsize
Adding some tape to the bottom standoff

After assembly and some test mounting, I gave the entire thing a coat of stain:

[Back to Top]

Step 6 - Finishing touches

And we’re essentially done!

Front-heavy!

Sometimes, due to bits being front-heavy and the pivoting of the lockout-selector, the drills can tip forward

Cut some hardboard

I cut them to 70mm wide so I could easily slide them in. The length, I just eyeballed

No more tipping
Test loading
Plenty of room for storage up top
Added a magnetic tool bar on the overhang

[Back to Top]

Milestone

This was a quick little weekend project that I could use up some leftover plywood kicking the shop: I thought about using glue but am glad I didn’t — getting all of the T’s to line up straight, parallel and perpendicular was quite a challenge and required a ton of finagling. I probably could have made do with two screws top and bottom to attach the T’s, but it was easier to just roll with three screws and not think about it.

In my zealous endeavour to use up plywood, one thing I didn’t take into consideration was that the two types of plywood I have (I call them ‘rough’ and ‘nice’) differ in their thickness by a not-insignificant amount. I tried to account for this by placing the T’s in a symmetrical pattern based on the use of the ‘rough’ plywood.

[Back to Top]

---

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Drill Rack Mk IIDIYDPMon, 30 Oct 2023 13:10:00 +0000https://www.minmaxgeek.com/blog/spotlight-reel6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:653f055947a8947d6175ffe0There are a handful of tools I use in the shop where the power cord is never quite long enough: my sander, tracksaw and pressure washer to name a few. Obviously, extension cords are a thing, but it’s always feels like a huge pain to have to coil up the cord after I’m done (first world problems much?).

The obvious solution is to get a retractable cord reel but I have an additional complication: the majority of cord reels are geared at being mounted high up on the wall or ceiling and my shop has a fairly low ceiling make this less than ideal.

---

Quick Links

• Placement

• Additional criteria

• Woods retractable reel

• Shelf mounting

  • Prototyping

  • Final mount

• Fixing the catch

• Closing thoughts

---

Placement

Since my shop has a relatively low ceiling and I have the added ceiling-complication of a garage-screen-door and random bulkheads, I specifically didn’t want to mount any cord reel high up (ceiling or wall). When I initially laid things out in the shop, I mounted my compressor hose-reel close to the floor — ideally, I would run this cord reel in that same general area.

By opting not to wall/ceiling mount my cord reel, and specifically, mounting it in this compact space, this introduces a requirement that the reel not flop around. My air reel has a fixed-frame, no-flopping-around design, so this was roughly what I was hoping to replicate, but for power.

This criteria ends up being the most difficult one to meet as the majority of reels seem to be designed to dangle from the ceiling. The design of having the cord reel dangle is so that you can simply ‘pull it and go’; by having a design that doesn’t dangle, it’s on me to extend the cord and make sure there is slack before walking off in random directions. I’m the type of user that

[Back to Top]

Additional criteria

In addition to the no-flopping design, I wanted two other features:

1. Single outlet: the tools I plan on using the extension cord for (power planer, track saw, etc.) are “single-tasker” tools - I’m going to be purely focused on using that tool and only that tool, so I only need the singular power cable to support that.

2. The power cord: in an ideal world, I’d want a (roughly) 50-foot reel of 12AWG but with the no-flopping design, this would be cost-prohibitive. Something more realistic would be in the 25-50 foot range, using 14AWG. This would let me comfortably power things like my track saw without worrying about amperage. As for length, for tool I’d be using in the shop, 10-15 feet would be plenty, the additional reach would allow me to have a bit more flexibility when using my pressure washer.

[Back to Top]

Woods retractable reel

If you’re in the market for a pro-grade reel (read: you have a much bigger budget), it’s quite possible to find a reel that hits all of my objectives from brands like Reelcraft or Coxreels. I was able to find a retractable reel made by Woods/Southwire that seems to hit all my needs.

Looking at the mounting plate, this looked super-promising: could there be through holes for running bolts? perhaps a slot/dovetail style bracket that we could drop the entire hose reel into? Even the box and manual suggest that this can mounted flat on a shelf/floor:

View fullsize
From the box View fullsize
From the instructions  

Big. Fat. Nope.

 

I’ll skip to the end: this reel was designed to be ceiling/wall mounted in a floppy manner. Anything suggesting that this is designed to be floor or shelf mounted is, at best, an afterthought. Also, seriously? Freestanding floor operation — has anyone ever tried to use a retractable (anything) in ‘freestanding mode’? Apparently the documentation folks don’t realize if you pull on the extension cord, uh, the entire reel will just come with you for the ride.

The reel comes with a swivelling wall/ceiling plate that doesn’t have any way of locking in place (to keep from flopping).

View fullsize
Swivelling wall/ceiling plate View fullsize
Shimmed to take out the wobble

That being said, this reel is the closest I’ve seen that could be flat mounted.

[Back to Top]

Shelf-mounting

In hindsight, you could just drill a few holes through the bottom plate and run bolts/screws through to anchor it down and be done. At the time, I wasn’t sure if

1. I was going to keep this reel and

2. The plastic on the base gave me a vibe that it wouldn’t drill nicely and might chip and break

So I wanted to find a different way to non-destructively mount this.

 

---

 

Prototyping

I had a bunch of 3/4 x 3/4 furniture braces that I could use: I use the existing mounting pins to tie into the furniture braces and in turn, the furniture brace would be tied down to some wood which I could then mount.

View fullsize
Drilling a 3/4x3/4 furniture brace View fullsize
Hole enlarged to 5/16 View fullsize
Testing with shims View fullsize
Comparing with 3/8 ply

[Back to Top]

Final mount

For the actual final mount, I opted to remove the embedded nut on the underside of the baseplate - this turns the L-bolt into a quick-release pin. I mounted this buffer-base onto a scrap of wood which now lets me mount the reel to a flat surface.

Buffer pad of 3/8 ply
Secured to a scrap
Fixed end installed
Removable pin installed
Final shimming
Mounted down
View from behind

[Back to Top]

Fixing the catch

So on the very first pull after finally mounting the reel, the latching mechanism failed. I wasn’t terribly upset as I kind of expected this to happen — this is a cheap reel and it’s designed like other cheap reels I’ve used before. Thankfully fixing this is fairly straightforward: pop the housing apart, play with the latching mechanism until it engages (shown below is after the fix - the teeth are engaging the latch).

 

 

[Back to Top]

Closing thoughts

I imagine nobody at Woods/Southwire actually tried to floor/shelf mount this prior to suggesting that it could totally be usable in such configurations but thankfully, this was an easy thing to retrofit. As an extension cord, it’s … fine? I was able to use it with my track saw without any issues.

One thing to note is that, the leader-wire is in fixed position (which you would expect), but unlike my air-reel, you can’t rotate the housing of the reel which means the leader-wire will always have to come out the right side — something to consider depending on the location of your outlet (thinking about this a bit more, you may be able to reorient the reel by disassembling it and flipping the entire assembling within the chassis - lol, I’m not desperate enough for that hassle).

In summary:

• As a retractable reel? There are better (12AWG) options if you don’t need it to be fixed in place

• It is possible to mount this such that is becomes fixed in place (you could probably also just drill mounting holes through the plastic base)

• This is a cheapo-style retractable mechanism — so you may need to pop it apart to fix it once in a while

• If you need a fixed-in-place design, this is likely one of the only affordable options there are

[Back to Top]

---

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Product Spotlight: Retractable Power ReelDIYDPThu, 14 Sep 2023 11:00:00 +0000https://www.minmaxgeek.com/blog/dog-bed6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:6501e5a154ac6c3061a5d67cWhenever Nicole leaves the house (even if it’s back and forth to unload/load the car), our (incredibly needy) dog goes into a life-altering crisis mode - he has to watch the window like his life depends on it. I get it — he’s velcro’d to her (better her than me), but he exudes so much angst that it drives me bonkers (I’m usually in the kitchen, cooking). So we have a little rule: you wanna stare? That’s fine — you gotta lay down.

As is though, he’s too short to see over the window while laying down which then makes him more anxious and it’s a never ending cycle. So here, I’m going to make a raised platform for his bed so everyone wins: he can watch her in utter panic and he can lay down so his panic doesn’t bother me.

---

Quick Links

• Goals and Objective

• Step 1 - Aaaaaaand done

• Milestone

---

Goals and objectives

I wanted to make a slightly elevated platform for the dog bed so that pooch can still see out the window when he’s laying down. He just exudes way too much anxious energy if he’s not laying down - very much a “OMG! WHERE YOU GO? WAI YOU LEAVE” vibe. By having a bit of a raised platform, it solves two problems:

1. He can still see out the window

2. For the person who’s still in the house (usually me — cooking), I don’t have to watch his dumb ass be super anxious

There’s a bonus that I can use the elevated platform to store some of his less-used toys.

I had previously threw together a very quick prototype using scraps in about 5 minutes and while it works, it was an eyesore for me, so now that I have some time, I can knock out a prettier one. Coming off the tail off the Mega Planter build, this should be a super easy project to knock out with scrap wood in about an hour.

[Back to Top]

Step 1 - Aaaaaand done

Not much to this:

1. Find some scrap plywood that fits the bed (with a bump-out so that we can clear the base-molding), glue, pin and screw it together

2. Make a simple drawer with glue and pocket holes

3. Attach drawer slides

4. Sand, stain and let dry

View fullsize
Original prototype View fullsize
Off cuts View fullsize
Marking for glue View fullsize
Scraps for a drawer View fullsize
Marking for pins View fullsize
Drawer box made View fullsize
Installing drawer slides View fullsize
Installing the mating piece View fullsize
Making a drawer pull View fullsize
And a drawer full of goodies

[Back to Top]

Milestone

To help give the bed a bit more ‘grip’ we threw a silicone baking mat under there.

Quick and easy project - just the way I like it.

[Back to Top]

---

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Dog Bed StandDIYDPTue, 12 Sep 2023 23:15:00 +0000https://www.minmaxgeek.com/blog/mega-planters6020ac0b21a9b86f0577e29e:60535320d9a09304764a163f:64ff2f33efb77071723688edWe had some reasonably good luck with the low-effort self-watering planters I made but I think we were running into a ‘volume of dirt’ (and thus, available nutrients) limiter for growing things. Nicole also wanted to have a bit more (dedicated) plant space so this was a good opportunity to make some biggie-sized planters. This was supposed to be a reasonably quick project — I was thinking a few weeks, a month at the most — and I had three months before my next planned project. What could possibly go wrong, right?

---

Quick Links

• Starting point

• Goals and objectives

• Rough design

• Step 1 - Used wooden frames

• Step 2 - Extracting the wood

• Step 3 - Surfacing

• Step 4 - Excavation

• Step 5 - Cutting the boards to length

• Step 6 - Painting

• Step 7 - E-track

• Step 8 - Joist tape

• Step 9 - Rough construction

• Step 10 - Sheathing, bulk heads

• Step 11 - Getting dirt in the planters

• Step 12 - Worm composters

• Step 13 - Top plate, trim, lights

• Step 14 - Plumbing

• Step 15 - Winter prep

• Step 16 - Misc. accessories

• Milestone

• Mix-ups, surprises and lessons learned<

• Appendix - Costs

• Appendix - By the numbers

---

Starting Point

Until recently, we’ve been getting by using DIY Self Watering Planters but once you have more than a few planters, they tend to be big, bulky and they feel like a lot of work (even though the self-watering aspect tends to be something of a time saver). Then there’s the end-of-season maintenance to tarp things up and get all the planters out of the way. It would be nice to have a more-permanent setup that is even easier to work with.

For us, we have a relatively large section of our yard that we can allocate for a permanent planter setup.

I had originally thought that this project would be relatively quick - that I could get to the ‘start loading dirt’ phase in “a few weeks” (a month on the long side)… #foreshadowing.

[Back to Top]

Goals and objectives

The main theme of this build was that I wanted to ‘do it once’ — that I would put in a boatload of effort up front and for the benefit of very little input down the road. And I wanted to look at this project on a 10 - no, 15 year scale. There were two functional areas I wanted to focus on:

1. Endurance: things like rot, water penetration, UV fade and rust

2. Ease of gardening: ease of watering, composting, [physically] support plants, ability to counter pests (birds and insects)

Aesthetics comes into play more from a maintenance angle — it would be nice to maintain the completed appears with very minimal effort. I have a theme of “black on black on black on black” — everything can be touched up with black spray paint.

[Back to Top]

Rough design

 

I went down a wild rabbit hole of design ideas.

I went down a wild rabbit hole of design ideas.

 

When I started this project, I balked at the price of bricks so I decided I would use wood — it was sheer great luck that we happened to get the drying kiln frames for cheap. We had talked about making three (or maybe four) large planters to span an area. The scale involved meant that I could look at (roughly) plywood-sheet sized constructs (so, 4x8 footprint) and I initially hoped to make the planter into a mega-sized self watering unit.

View fullsize
Overall idea View fullsize
Implementation A View fullsize
Implementation B  

---

 

The risk of dealing with random-BS with self-watering planters wasn’t worth it to me, so I eventually abandoned the idea and started to focus on ‘just a basic rectangle box’. My big concern was rot — no matter how thick and beefy the wood is, once it gets waterlogged, it’s just a matter of time before rot comes along and there’s nothing for screws/nails to hold onto anymore. In particular, I was focused on how to keep the boards coplanar with each other (since they would all be warped/used construction grade lumber).

I went back to the idea of using wood (the green bar) to brace the planter wall and how to make it pretty and functional. I really liked the idea of using the [green] 2x4s as studs essentially. With the grey bar at the top (in conjunction with the existing blue bar), you now have a double-wide support for a top plate - enough that you could just friction fit a top in place. This would make the top entirely replaceable; or at the least, more easily serviceable.

View fullsize
Looking at one of the faces of the planter View fullsize
Leverage the green supports to make a 'wall' View fullsize
Option of fixed or removable

I started to think about how, if this was an aquarium, the sheer weight of the water would cause the tank to burst without some form of cross brace. I specifically didn’t want a massive, chunky wooden brace (which would also be something that would potentially rot).

View fullsize
Starting point View fullsize
Install an eye-bolt on the inside View fullsize
Turnbuckle to brace View fullsize
Using e-track instead of wood  

---

 

I was happy knowing that I had some options on how to brace and support the planter when the time came - I felt like I would have a better feel for how much bracing I might need once the planter was further along the construction process. I wanted to provide a permanent watering solution for the planter: I would run the garden hose over to where the planter is and from then I needed to find a way to get this water into the planter. For aesthetic reasons, I specifically wanted to avoid flopping a hose over the top; the computer builder in me thinks that’s bad cable management, so it’s a no go.

I looked at ‘misusing’ a whole bunch of garden hose accessories (i.e., using a garden tap reversed lol). In the end, the path of least resistance was just to drill a hole through the planter wall and feed the hose through.

 

---

 

The last aspect I wanted to consider was finding a way to cage the planters with the goal of keeping birds and squirrels off the goodies. I knew that I was going to have some steel grid/mesh so I had a thought of erecting two T-pillars, one at each end, this would let me lean the steel mesh against it. While I initially thought of just using two pillars (one at each end), there might be some benefit to having additional pillars:

• less sag

• individual panels are smaller — which makes them much easier to remove

View fullsize
Erect T-pillar at each end View fullsize
Drape/lead steel grid/mesh against the pillar.

The tricky thing about this is “how high do you make the T-pillar”? I don’t have very much experience successfully growing anything so I have no grasp of the scale that plants can grow. Sure, some plants will certainly grow through the mesh, it then becomes a bit tricky to remove some of the panels to get into the planter without damaging the plants. For the time being, I opted to table the idea of caging the planters, I’ll revisit this if things manage to grow enough to matter.

 

---

 

One thing I specifically did not come up with s solid plan for is how to connect the corners. Truthfully, I didn’t like anything that didn’t eventually lead to the phrase “custom metal fabrication”. When taking apart the frames, I had a lot of extra Simpson Angles (A34, A35, L) — but all of them make me super wary for application in a planter:

• No amount of galvanized anything is going to survive direct-burial

• A lot of the angles I did have were already showing rust (from being in a kiln) - so it’s a matter of time before they rust-fail

• Stainless steel options did exist, but there were very expensive ($8/corner — I have 60 corners)

• Even ignoring the cost, every single brace that I could find has you putting fasteners right into the corner — what will become the weakest part of the wood when it becomes waterlogged (because it’s holding wet-dirt after all)

There are some limited ‘ornamental style’ where you might be able to get your screw points further away from the end-grain but those are even more expensive. I briefly considering rabbeting out the wood to support something like an L-strap, turned sideways, on the end, but that’s a ton of work which might not work out well in the long run (like, how can I replace a failed piece??)

I ultimately went with no special hardware (since all of it requires putting screws right near the end-grain) and if something failed, then I would go down the road of a custom fabricated brace — one that I would attach from the outside (so there would be an aesthetic element to it).

I planned to have the boards laid out in a ‘chasing pattern’

My planter is going to built as a “stack of rectangles” — so when stacking them, I would alternate between the two patterns with the hope that this would reduce the chances of corner failure. Since I was planning on adding braces (essentially vertically connecting boards, even if one (or possibly more) corner(s) failed (the greens), it should be able to be held together via the connective force between the rectangles. I dunno. It’s a hope.

One of the important things we made sure to do was be very generous with the spacing and placement of the entire space: I want to be able to [easily] walk a mower behind the planters (next to the fence-line) and I wanted to be able to [easily] fit our dump cart/wheelbarrow between the rows of planters.

[Back to Top]

Step 1 - Used wooden frames

There was a ginseng farm about 30-minutes away that was moving out of the region; as part of their wind-down operations, they were getting rid of thousands of kiln-drying racks. These are essentially glorified 2x6 frames and to make it a tray, they had some heavy-duty (to me) 1x1 steel mesh stapled on the bottom side.

They were initially selling for $10/tray although when we got there, they were much more interested in getting rid of them more than anything (as their moving date was coming up fast) so they let us take as many as we felt like taking. Factoring in the ‘bonus’ frames we took, we ended up paying $6.50 per frame — a bargain from a materials perspective.

View fullsize
Loading the frames View fullsize
A whole bunch of trips View fullsize
Metal frame

Since the kiln trays were of varying age, their individual condition was hit and miss - we knew that there would be some boards that would just be entirely unusable due to cracks, twist, warp or sheer nail holes. The plan was to disassemble the frames, get the wood surfaced up so it was a nice clean starting point.

[Back to Top]

Step 2 - Extracting the wood

I planned on running the boards through a planer to get to fresh wood and so it was imperative that I thoroughly remove all the metal from the frames. We could tell that the frames were assembled by different people as they had a bunch of different ways of being put together:

• Large nails holding the framing members together

• Sometimes brackets with screws holding the corners together (and sometimes with nails)

• Fencing staples holding the steel grid down

• Sometimes long staples were used instead

To break down a frame into its component pieces, there were a four steps needed:

1. Remove the staples holding the metal grid to the frame

2. Disassemble the corners

3. Smash the boards apart

4. Back out all of the nails

Nicole came to help me pull nails and this made it go by so much quicker; all in all still took us a week to do the rough frame-brakedown.

View fullsize
Removing the metal-grid View fullsize
Furring strips View fullsize
Digging out a staple View fullsize
Enter vise grips View fullsize
Sigh View fullsize
Finally disassembling the corner View fullsize
Backing out the nails View fullsize
Backed out View fullsize
Nail puller and prybar View fullsize
Drop a cloth on it View fullsize
Not fun View fullsize
Also not fun View fullsize
Fencing staples are much nicer View fullsize
Narrow crown staples View fullsize
One frame's worth of fasteners View fullsize
Sigh View fullsize
No nice way to remove this View fullsize
Such an annoyance View fullsize
Back side of the furring strips View fullsize
Cutting them down to size View fullsize
Corner connector View fullsize
Covered nails View fullsize
So much holding power View fullsize
Can try hammering from the back View fullsize
Put whole body into levering it View fullsize
Steel grid sheets View fullsize
Just a bit under half done... View fullsize
Just under half

I ended up with almost 10 pounds of metal fasteners — not counting the bucketful of corner braces and random mending plates.

View fullsize
10lbs of fasteners View fullsize
Broken cats paw

[Back to Top]

Step 3 - Surfacing

I thought that I could just through the boards through the planer and be done — it would get me my fresh wood and I’d be done as fast as I could feed the boards through. I knew that unless I ran it through the jointer as well, that it wouldn’t be perfect and I was okay with any imperfections. I was not ready for the amount of warp and twist on some of these boards though.

I did end up fully milling the shorter boards to be perfectly square; the long boards were not feasible to run through my baby bench-top jointer (safely). Instead, I bought a power planer and roughly smoothed one face before switching to the bigger bench-top planer to do the other side.

Before any of this can happen, it’s imperative to truly ensure that all of the metal fasteners are removed so I bought a metal-scanner and thoroughly checked every board (on all surfaces) — this scanner saved me from a small jar’s worth of small nails and staples I missed. In some cases (like the end of a board), I opted to just cut the end off but some cases, it was impossible to remove the fasteners so I had to skip that board entirely.

View fullsize
Setting up to surface the short boards View fullsize
Can't see it View fullsize
Right there! View fullsize
Broken staple View fullsize
Another stealthy one  

---

  View fullsize
For the smaller boards View fullsize
Running everything through the planer View fullsize
Halfway done (the long boards) View fullsize
For the longboards View fullsize
Another angle View fullsize
Smooth edge View fullsize
Ton of sawdust  

---

  View fullsize
Setting up a jointing station View fullsize
Another angle View fullsize
Zip! View fullsize
One flat edge on all of the long boards View fullsize
Cutting the short boards to final width View fullsize
Cutting longer boards to width View fullsize
Cleaned up View fullsize
Filling in some of the cracks View fullsize
Some nail holes are a total lost cause though

[Back to Top]

Step 4 - Excavation

I thought this would be the hardest part but it was surprisingly easy all things considered — getting the dump bin was the game changer. For the excavation, I only went down as far as the edger went — since the planters will float on top of the gravel bed, I don’t really have any frost-heave considerations. From start to finish, this was about 3-days from marking the perimter to having the 5-yards of dirt hauled away.

View fullsize
Cutting the perimeter View fullsize
Harder than it looks View fullsize
Power edger assist View fullsize
Edger cuts View fullsize
It's a slow slog getting started View fullsize
Some surprises View fullsize
We marked the lines for our sake View fullsize
Having a dump cart was helpful View fullsize
Dump bin View fullsize
They didn't care how we dumped it

After we got the dirt out, we put down some weed barrier (we were on the clock because it was threatening to rain). I picked up a truckload of bricks and we got a quick delivery of rock to get the base layer down.

View fullsize
All the dirt, out View fullsize
Landscape fabric down View fullsize
Staples with grommets View fullsize
A whole bunch of bricks View fullsize
Neatly loaded in my truck View fullsize
Figuring out orientation View fullsize
All the bricks laid down View fullsize
Steel mesh View fullsize
Rock delivery View fullsize
Threw some rocks on there just to keep it from moving View fullsize
About a day of hauling rocks View fullsize
Very roughly leveling this out (with rake) View fullsize
Packing with screen View fullsize
Packed in View fullsize
Checkpoint done!

[Back to Top]

Step 5 - Cutting the boards to length

I waited until now to cut the final boards because I wanted to excavate first and size the planters to fit. Another considering is that I knew I would definitely be cutting off material from both ends (because the nail holes and end-rot were pretty bad) so I needed to find a length that could fit within the boards I had.

View fullsize
Long boards done View fullsize
A whole bunch of offcuts View fullsize
Trimming off the end of the board View fullsize
Microfractures View fullsize
Cutting real close View fullsize
Short boards cleaned up View fullsize
Cut to final length View fullsize
Short boards cut to length View fullsize
Figuring out the length of the long boards View fullsize
Makeshift stop View fullsize
Some boards didn't make the cut View fullsize
A close up of some a crack View fullsize
Filling cracks

[Back to Top]

Step 6 - Painting

My goal was to keep the planters maintainable - so an emphasis on not only building it to be robust but also being able to keep it super easy to keep looking fresh. As such I went with the obvious choice of black. This way touch-ups are just a spray-paint can away. Based on the longevity of my picnic table, I went with two coats of primer and two coats of exterior paint (originally wanted three coats but just couldn’t get the coats in between the weather)

This was, by a long shot, the most tedious part of the build — and it was almost entirely due to the horrendous weather we’ve had this summer. I want stupid-hot, stupid-dry weather so that I can knock out multiple coats of paint but instead we got a rain-season…

Starting with the short boards

I used my cheapo HVLP to knock these out pretty quick.

Slowly chipping away at the backlog
Starting a second set of boards while the first dry
First batch of boards, drying
I got a bigger paint sprayer

Nothing wrong with the one I was using except for the fact it was a huge pain in the ass to constantly refill the cup. And each time I refilled the cup, I’d have to thin it down and I would inevitably spill 6% of it….

Blasting through all of the short boards
This is just the first coat
Weather

The sky was constantly threatening to rain on me

Paint for the long boards
Letting the first batch dry while I bring the second batch
Brought indoors to dry
Starting to prime the long boards

[Back to Top]

Step 7 - E-track

While my short-boards are all nice and square (for now!), the long boards are all slightly wave since I only ran them through the planer (on both sides). To keep the planter panels coplanar with each other, I wanted to find a way to attach them to each other. Most builds I’ve seen have a stake of sort at regular intervals where the planter boards screw into the stake, connecting them all together. From an aesthetic reason, I was against this — I wanted a clean box (or as clean as I could get it).

I opted to use e-track as it’s got the strength to forcibly hold the boards all together and is reasonably good at not rusting. I covered the e-track with plasti-dip just to be sure.

View fullsize
Bought 4-ft e-track View fullsize
Prepping for cutting View fullsize
Cut guide View fullsize
Securing the wood in place View fullsize
Clean cut View fullsize
Got cleaner as I did more View fullsize
Smoothing down the cut edges View fullsize
Plasti-dip View fullsize
Hit the back side too

[Back to Top]

Step 8 - Joist Tape

So far the planter is built fairly tough but it’s still a bunch of wood filled with [wet] dirt and sitting on [wet] ground. Water is gonna sit on surfaces and then I’m hosed. So I had a crazy idea of using some joist tape to absolutely cover the surfaces which might have water pooling on them. Joist tape also is a wonderful solution to my problem of what to do about some of the massive nail holes left in the wood and it mitigates my concern about water seeping through the end grain. A friend of mine may have made a comment that the 2x6s that make up the planter are just a structural support for joist tape…

The joist tape will be visible so to a degree, I do need to worry a bit about how I line the tape up so that it’s not hideous. I’m going to screw it up but at least I can try to get it lined up right.

Figuring out the overhang
This way the tape is centered in the edge
Super clean application
Making sure to iron down the tape
Cutting off any overhang
Wrapping the end grain
Setting up a tape station

Since all of the short boards were milled to the same dimensions, I setup a quick guide to get the tape to the right depth

Setting up a repeatable stop
Getting one side down

The clamp applies just enough force to allow me to hold the tape but the tape can still be repositioned if needed.

Super clean
For the longer boards

For the longer boards I pulled out the laser because I didn’t want to muck with stops (also, the boards are not the exact same dimension)

Line the tape up with the laser

[Back to Top]

Step 9 - Rough construction

Finally: I can start to build the planters! This phase is just assembly of the long and short boards into a bunch of stacking rectangles.

View fullsize
Build day View fullsize
Joist tape 'patches' View fullsize
Patched View fullsize
Batching out the nail-hole patching View fullsize
Figuring out the orientation View fullsize
First box is crucial View fullsize
Clamped top and bottom View fullsize
Drill guide View fullsize
Getting ready to drill View fullsize
Pre-drilling View fullsize
Supporting the box View fullsize
Repeat View fullsize
Repeat View fullsize
Changed the height

The tower of rectangles is complete but none of the rectangles are connected to each other and they aren’t aligned and distributed in any way. The next day, I take everything off the stone and figure out the layout and put it back together. I figured everything out (in theory) on paper and set out to see if reality would match what I

View fullsize
Story sticks View fullsize
Starting to align the boxes View fullsize
Making sure the rectangles are level View fullsize
Stacking game View fullsize
Sliders View fullsize
Sliding spacers in use View fullsize
Repeat View fullsize
Story stick to get the support e-track located View fullsize
Aligning the e-track View fullsize
Secured View fullsize
Done for now

[Back to Top]

Step 10 - Sheathing, bulk heads

So the thing about planters is that they typically hold [wet] dirt — so I had the thought of using some form of plastic to further protect the inner walls from moisture damage (in theory, the outer walls would be handled by gravity). I had a bunch of vapor barrier left over from my basement Mega-Pantry project so I opted to use that. I ran out so close to the finish line but a friend of mine had some extra so I didn’t have to buy 500 million square feet just to finish this.

View fullsize
Cut station View fullsize
Stitched together View fullsize
Stapled into place View fullsize
Intentionally left the ends View fullsize
Brace added

So now that I’ve spent all this time making this a protected as I can, the next thing to do is to punch a massive hole into it right? Of course.

In my vision, I have a means of getting a water hose into the planter without draping it over the top edge. By using an aquarium bulkhead, I can get a watertight seal that also ensures that the wriggling in and out of the garden hose won’t physically damage the opening over time either. When I was done, I gave everything conspicuous a quick splash of spray paint.

View fullsize
2" bulkhead View fullsize
Making a drill guide View fullsize
Clamping the guide in place View fullsize
YOLO hole View fullsize
Splash of spraypaint View fullsize
Painted View fullsize
Finishing my sheathing View fullsize
Cut a hole in the sheathing View fullsize
Neatly taping it down View fullsize
Testing View fullsize
Emptied everything out View fullsize
Setting the paint depth View fullsize
Blends right in

[Back to Top]

Step 11 - Getting dirt in the planters

With the basic construction done, it’s finally time to get dirt in the planters! We put in a call for some dirt and got quite lucky — it was delivered within an hour. We spent about six hours getting all the dirt moved into the planters. I didn’t make any effort to compact the dirt — it was due to rain so there was going to be natural compaction of the soil soon enough.

View fullsize
A much bigger truck View fullsize
We ordered five yards

I don’t have a need, for it, but I wanted to lay down some landscape fabric and steel mesh to keep the dirt roughly contained should anything happen to the planters - aka mostly because I had it and wanted to find a use for it. We moved the dirt by using two garbage cans - so about 70 gallons a time.

View fullsize
Measuring the steel grid View fullsize
Marking the cut line View fullsize
Allocating landscape fabric View fullsize
Landscape fabric down View fullsize
Steel grid on top of it. View fullsize
5-gal bucket as a placeholder for a composter View fullsize
Using garbage bins for transfer View fullsize
3hours of moving dirt later View fullsize
A ton of dirt later..

[Back to Top]

Step 12 - Worm composters

We’ve had a worm composter in the house for a couple of years now with reasonable success — so I wanted to try embedding a worm composter directly into the planter. This way the worms can do what they do naturally and we have additional composting capacity and we don’t have do disperse the compost — it’s already in the planter. You can totally buy a premade worm composter that can be direct buried and if I was just making one planter, I totally would just save the time and buy it. Since I was making three, it made more sense to just try and DIY all three it for somewhere between 1/3 and 1/2 the cost of buying a single one.

View fullsize
Basic components View fullsize
Mark out holes for worms to move freely laterally View fullsize
Drilled the holes out View fullsize
Drilled a hole in the base View fullsize
Use a template bit to cut out the base View fullsize
Getting the lid on View fullsize
Remove the handle bar View fullsize
Excavate the planter a bit View fullsize
I put duct tape on the side holes View fullsize
Leveled the bucket flat too View fullsize
Plan is to bury the composter up to the rim View fullsize
Rinse and repeat for all planters View fullsize
Dinner plate on top

[Back to Top]

Step 13 - Top plate, trim, lights

Making good progress and I can feel the end is very close. The top plate has been something that’s concerned me since the start — it’s a flat wooden surface that is difficult to get water to trickle off. I debated adding a shim to angle it and I also debated ripping a bevel into the top using my bandsaw but I ultimately went with the EZ-mode solution: joist tape.

I went with the boards ‘chasing each other around the perimeter’ configuration because that’s work worked and I just eyeballed the placement until it was close and then I went around with a combination square to finalize the inset. Once I got the top plate initially secured (using as few screws as I felt comfortable using), I went back with some angle brackets — not so much for full on strength but just to keep the top from pivoting too much (as it’s balancing on the top edge of the planter). A splash of spray paint makes it disappear from casual glance.

Some of the long board, if you’ll recall, had twist in them — and I didn’t joint it so the warp is still here. I bought some mending plates to try and correct this a bit. It’s not perfect, but it brings the gaps together a bit better.

View fullsize
Joist tape View fullsize
Roughly getting the top plate in position View fullsize
Stainless angle brackets View fullsize
View fullsize
Securing the top plate View fullsize
Covering the screws View fullsize
Support braces installed View fullsize
Adding corner trim View fullsize
Splash of spray paint View fullsize
Installed the solar lights View fullsize
Warped wood View fullsize
Attached to the underside View fullsize
After bracing  

My original goal

During my original design process, I had though about making the top plate sacrificial — every few years you pop it off and replace it with a new top plate. This way I could have a bit more aesthetic appeal but still keep to my original goal of being relatively low maintenance. I had hoped to have three colors: black for the body, white for the corner trim and a rich shade of brown for the top plate.

When I started pricing out the cost of materials to do this (compared to using the existing boards I had on hand that were already primed, painted, taped and ready to go), it wasn’t really a viable option.

 

[Back to Top]

Step 14 - Plumbing

We’re not great for watering our plants with any amount of consistency so anything I can do to help that will go a long way. For the initial iteration of this planter, I opted to go with soaker hoses and a mechanical timer. Watering the planters is as easy as popping outside, clicking a timer and we’re done. Everything uses quick disconnects so when we need to mow behind the planters, it’s just a single disconnect to get the plumbing out of the way.

View fullsize
Custom length hoses View fullsize
Using a 4-way manifold View fullsize
Right angle connection View fullsize
Nice and tidy View fullsize
Quick water test View fullsize
Controlled by timer

One annoying thing about soaker hoses is that the soaker bits start “right away” i.e., right at the connection point of the hose - that’s where the water starts to come out. So, if you need to go through a material first (like the side-wall of a planter), you’ll need to insert a piece of straight hose so that the soaker-bits are entirely inside the planter. For me, it was cheaper to buy a small pack of anti-kink extensions than it was to make my own since I had run out of hoses and repair kits by that time.

Having all of the 90-degree connections and the manifold probably does reduce the flow rate of water through the hoses but that’s not been an issue. In fact, I’ve found that I needed to dial back the water flow to a couple of the planters so that they all water evenly. Having

[Back to Top]

Step 15 - Winter prep

After going through this much effort, I wanted to find a way to make it easy to winter prep these planters: I bought some nicer tarps and found a way to make it easy to pack up the planter for the winter. The goal is to be able to consistently pack up the planters in about 5 minutes or so and by using D-rings, I can keep the entire setup fairly neat and tidy — they don’t particularly stand out unless you are specifically looking for them.

View fullsize
Centered the anchor point View fullsize
Placed just above the ground View fullsize
Front View fullsize
Rear

[Back to Top]

Step 16 - Misc. accessories

Off the bat, I got a few things to make the planter a bit easier to work with: some short and long fiberglass stakes and some fiberglass + PVC electric-fence-posts — perfect for running string lines across the planter. I bought a bulk pack of plastic plant support hoops but we’ll try those out next spring.

To make things easier for Nicole, I picked up a plastic (water sealed) .50 cal ammo box — this way, we can keep the commonly needed gardening tools right there, outside in the weather, by the planters.

[Back to Top]

Milestone

Over the course of the last week or so (funny enough, as we near the end of the season), this project has finally started to wind down and Nicole spent a couple days moving everything from our puny small planters to these comparatively massive garden beds (remembering to remove the duct tape from the worm composters!). Not sure if this will help us grow things — neither of us are green thumbs and we’re definitely not good at keeping plants alive but at least this will give us a shot next year.

Stepping back though, I’m pretty happy with how the planters turned out part of me thinks that we maybe should have pushed for four planters — it didn’t take very long to fill up the existing ones!

A few strong points:

• Having a generous buffer of space (the rocks) around the planters and making sure to allocate a ton of space between the planters makes it pretty nice to use

• Hooking all three planters up to a common (timed!) water source makes it so much easier to ensure the plants get water if-when they need it

• Adding tie down anchor points now, during nicer weather and test fitting everything before cold weather gets here I think will make the winterization process a lot nicer

• The MVP of this entire project was the gorilla dump cart! Without that cart, we used it for everything - without it, it would be miserable to move all the dirt, stone and fill around

 

---

 

I set out with the initial goal of building a set of planter boxes from untreated lumber that I want to last me the better part of a decade. No — pushing on two decades. I think there’s more than a fighting chance that I’ve met that objective. Let’s look at the failure vectors:

• Rust: every single fastener is stainless; there’s a small chance that the fasteners for the D-ring are not stainless but at the first sign of rust I’m happy to swap that out. Everything else is plastic, fiberglass, PVC, etc. I suppose the e-track is not stainless, but I couldn’t really find 2-foot e-track in a stainless variety - I’m hoping the extra two coats of plastidip (and the fact it’s wrapped behind vapor barrier) should make this a non-issue

• UV exposure: what exposure - every single visible (and underside) surface is covered in primer, paint and possibly up to two layers of joist tape. I’m not saying that these things can’t fail but that’s nothing that a quick (albeit recurring) splash of spray paint can’t fix over the decades I hope to have this

• Rot/Moisture: this was the big one, how to make an outdoor box that is meant to hold wet dirt… not get wet. Short of epoxying or rubber-coating everything (and that was my original plan to be honest), I think I’ve given everything the best chance it has. There are some holes in the vapor barrier from stapling it on but staple holes I imagine will be minor. The only substantial source of moisture contact is the fact the planters are literally sitting on the ground - although the ground contact surface is wrapped in joist-tape. I really hope this is good.

Only time will tell, but I think we’re in good shape.

[Back to Top]

Mix-ups, surprises and lessons learned

• If I had to nail down a single lesson from this entire project, it’s “holy crap, just use brick”. For the purposes of being a planter, brick would have been indestructible: no rust, no rot, not so much UV fade, relatively waterproof, no warping, negligible heat expansion. Full stop. Done.

• This project was definitely one of the longest projects I’ve ever done. I initially thought the hardest part would have been the landscaping/excavating elements and it probably would have made it into the top three if we hadn’t done material-delivery and a dump-bin - would have been countless trips back and forth in the pickup - and doubly exhausting because you have to unload (and sometimes load) each trip.

• The most pain in the butt process was definitely painting. To that end, splurging for a paint sprayer where you can just dip the intake hose into the can directly? Totally worth it because it means you don’t have to stop (what feels like) every 35 seconds to refill a small cup and then spend time thinning the paint down. It’s just spray-spray-spray. The only frustrating element is that it’s really dependant on weather: for the only time in history, I wanted heatstroke conditions (35C+, all-day-sun) but for a month (or maybe more), all we got were gloomy cloudy or rainy days with a high of 24C. The dry time (for basic touching) just goes through the roof.

• One surprising discovery were the plastic and fiberglass electric fence posts — they come with nice, hanging slots for running string across. The only downside is that I find they aren’t nearly as sturdy as I’d like (they are dependent on how hard you pack the dirt) so I might revisit this with some concrete-in-pepsi-bottle footings that we can just pop the fence-post into. We’ll see how they hold up next year and tackle it if needed.

[Back to Top]

---

Appendix - Costs

Trying to summarize the cost of the planters is tricky because of two factors

1. For stuff I already have, do I include that? When I started the project, I already had a bunch of tools, supplies, equipment etc.; for example, I already had some leftover vapor barrier from a previous project, so I didn’t have to buy much to complete the project.

2. For new purchases, how much do I include? I bought a bunch of tools that will definitely see future use (i.e., paint sprayer) and I have a bunch of consumables that are leftover, so the full cost isn’t entirely captured.

Best I can do is inventory costs by rough category

• Construction: this is primarily the costs of raw materials (ginseng kiln frames, stainless-steel screws, staples and brad-nails, braces, mending-plates, screw-crap covers, e-track, PVC trim) but also some small hand tools (replacement nail puller, prybar and vise grips for the ones that broke) and some misc. purchases (like precision glue bottle) - $730

• Gardening: this includes material purchases for the DIY worm composter (5-gallon buckets, ready-seal lids, plastic plate) as well as any gardening specific accessories (fiberglass plant stakes, electric fence posts, clips, string line) - $265

• Landscaping: this includes the major expenses related to the excavation and filling of the base layer (bin rental, 5-yards of triple-mix, 2-yards of 1” clear stone, a bit of screen, weed barrier, landscape staples, an extra transfer shovel and the Gorilla Cart) - $1300

• Misc.: a bit of catch all for ‘decor’ related things (ammo-crate, solar lights) - $50

• Paint: this covers just the consumables (primer, paint, spray-paint, Plastidip) - $615

  • Paint color: Grand piano (aka hella black)

  • Primer: tinted with ~1oz black

  • Spray paint: Canyon black, Satin

• Plumbing: this covers all of the bits to get the planters to be easy-watering (soaker hoses, anti-kink adapter, repair kit, quick-connects, 4-way manifold, 90-degree adapter, irrigation timer, aquarium bulkhead). I didn’t include the preexisting stuff I already had (hoses, flow valves etc.) - $325

  • Bonus: from quick testing, it appears that the quick-connect kit I bought on Amazon is compatible with the set that I’ve been using from Yardworks so I’ll likely buy another batch of the cheapo ones just to have extras.

• Tools: this includes power tools mostly that I bought for the build (metal detector, paint sprayer, power planer) but doesn’t include the tools that I already had. I didn’t include smaller things like the hole-saw I needed to buy for drilling into the planter. - $750

• Waterproofing: this encompasses the consumables I needed to protect the interior of the planters (joist tape, vapor barrier and sheathing tape) - $400

• Winterizing: this includes all the bits needed to get the planters ready for winter (tarps, d-ring, rubber strap and bungee cords) - $185

 

---

 

Which brings the rough total to an eye watering $4620. It is neat, however to look into how the expenses break down, both by category (roughly itemized above) and by ‘type of expense’.

View fullsize
Breakdown by category View fullsize
Breakdown by 'type' of expense

[Back to Top]

Appendix - By the numbers

It’s kind of fun to go through some of the numbers involved

• Ginseng frames: 28

  • Broken down into

    • 2x6x10 boards: 56

    • 2x6x4 boards: 72

    • 2x4x4 boards: 36

    • 1” x 1” steel grid sheets: 25

    • 0.75” x 0.75” steel grid sheets: 2

    • Unusable steel grid sheets: ~1

  • Trips to bring home: 7

  • Metal fasteners to remove: ~10lbs

  • Wood shavings from planing: ~250 gal

• Landscaping

  • Landscaping staples: 42

  • Earth excavated: ~5 yards

  • 1” clear stone: ~2 yards

  • Triple mix: order 5-yards, got 6-yards

  • Screen: ~10-gallons

  • Landscape fabric (3’ x 100’ rolls): 3+

  • Bricks: 52 (used Oldcastle Townsend wall blocks)

• Painting

  • Primer: ~3 gallons

  • Paint: ~4 gallons

  • Spray paint: ~2 cans

  • Plastidip: ~2 cans

  • Number of times I cursed at the small paint sprayer and painting in general: 500+

• Planters: 3

  • 2x6 (42” and 96”) boards: 36 (stacked 5 high)

  • Stainless steel angle brackets: 24

  • Stainless steel mending plates: 12

  • 2-ft e-track: 18

  • 1” stainless steel pan-head screws: ~300 (attaching e-track to panels), + 6 (hanging the solar lights)

  • 2.5” and 3” stainless wood screws: ~180 (perimeter) + 36 (top plate)

    • Why have two different sizes? I originally targeted 2.5” but occasionally the countersink bit would slip and plunge a bit too far (I predrilled every single hole), so for those scenarios, I would switch to 3”

  • 1/2” stainless staples: ~250

  • 1-1/2” stainless brad nails: 72

  • Solar lights: 6

  • Number of non-stainless fasteners used: 0

• Waterproofing

  • Rolls of joist tape: ~14

  • Rolls of sheathing tape: ~2

• Winterizing

  • D-ring shackles: 6

  • 18” bungee cords: 15

  • 30” rubber strap: 3

[Back to Top]

---

 

 

Product links may be affiliate links: MinMaxGeek may earn a commission on any purchases made via said links without any additional cost to you.

]]>Build Log: Mega Planters