ecobee4 with Amazon Alexa Voice Service

What’s in the Box?

ecobee4 Smart Thermostat unboxed

The first thing I noticed is that ecobee continues to demonstrate their understanding of the importance of a well laid-out box opening “experience.” It feels downright iPhone-ish (that’s a good thing). DIY geeks like me get warm fuzzies due to the attention to detail, while those who might otherwise feel a bit intimidated about installing a smart thermostat for the first time get a “Well this isn’t so bad!” vibe.

The next thing I noticed was Amazon’s Alexa circular logo proudly displayed on the front of the box. The official phrasing from ecobee is that the new ecobee4 features “built-in Amazon Alexa Voice Service.” But that’s just a fancy way of saying that for almost all intents and purposes, the ecobee4 is an Amazon Echo Dot and an ecobee smart thermostat mashed up into one device that sits on your wall — and in addition to SmartThings and IFTTT, also supports Apple HomeKit. As far as I know, it’s the only device currently on the market that combines Amazon Alexa and Apple HomeKit technologies in the same unit. I shudder when I think of the machinations ecobee CEO Stuart Lombard and his executive team had to go through to negotiate that type of licensing… while still keeping the ecobee4’s price point competitive with other industry players. But that’s not my job, it’s theirs! Mine is to play with stuff and see if I can break it talk about the experience.

Compared to its older sister, the ecobee4 (shown on the right below) is slightly larger and somewhat “rounder.” I didn’t measure them, but the touchscreen looks about the same size… or maybe even a touch bigger on the ecobee4:

ecobee3 (left) and ecobee4 (right)

The ecobee4 is also slightly thicker than the ecobee3, probably to make room for all that Amazon juiciness:

The ecobee4 is slightly thicker than the ecobee3

The rear of the ecobee4 reveals a number of vents, which I’m guessing are for ventilation (up top) as well as microphone (lower right) and speakers (bottom center):

Rear view of the ecobee4 smart thermostat

The ecobee4 mounts to the wall in a similar fashion as did the ecobee3, but the two wiring harnesses aren’t identical, so while it’s not quite as simple as unplugging the old unit and snapping in the new one, migration from an ecobee3 to an ecobee4 is as simple as migration from pretty much any other thermostat: disconnect the old wires, mount and connect the new wiring harness, plug in the ecobee4. Yes, friends, it’s really that simple. Here’s the ecobee4 with its wiring harness and optional cosmetic wall plate (top) compared to the ecobee3 (bottom):

ecobee4 and wiring harness compared to the ecobee3

You don’t need to install the trim ring, and the install looks much prettier if you don’t. But I have holes in my wall where an old-school thermostat was mounted and I haven’t gotten around to repairing those holes, so I opted to use the trim ring for this install.

Installing the ecobee4

As I mentioned above, whether you’re coming from an ecobee3, another type of smart thermostat, or even an old-school “dumb” thermostat, installation of the ecobee4 is a snap. The instruction guide provided by ecobee will walk you through everything, and their tech support will help in the unlikely event you get stuck. For me, installation took about 5 minutes.

First, I turned off the power switch at my furnace.

Next, I took a picture of how my previous thermostat was wired. My wires’ colors happen to correspond to the letters on the old wiring harness, but don’t assume that’s the case for your thermostat. The ecobee install guide will explain how to identify the proper wires on your HVAC unit to make sure you’re wiring things up properly. However, if your colors do match up like mine, you can proceed with ease. Here’s my old harness:

Old ecobee3 wiring harness

Note that on the ecobee3, I connected the single red wire to the Rh terminal. That’s different on the ecobee4, however, and if you have a single red wire, it must be connected to the Rc terminal. There’s a small blue insert in the ecobee4’s box that makes this clear.

Third, I removed the old thermostat and its harness:

Wall ready for install

Because I have screw holes in the wall from a previously installed Honeywell, I opted to use the trim ring that comes with the ecobee4 to hide the holes. It’s a cleaner look to mount the ecobee4 without the ring, so if I ever get around to filling in the holes and repainting, I’ll remove the ring. For now, the ring works fine. I’ve also been told that the trim ring takes paint well, so that’s another option worth considering.

New ecobee4 wiring harness and optional trim ring

All that was left was to slide the correct wires into the quick-connect terminals:

ecobee4 wiring harness connected

Finally, I clicked the ecobee4 into place and turned on the power at my furnace. My new ecobee4 sprung to life!

My ecobee4 is alive!

ecobee4 Setup and Configuration

An automated setup wizard walked me through getting my ecobee4 up and running. It correctly sensed the terminals I had connected, which gives the ecobee4 a pretty good (well, perfect actually) guess at my setup, which is a standard single-stage furnace with an external A/C unit:

Wiring terminals auto-detected

After a couple standard questions, the setup wizard then asked how I wanted to configure my WiFi network settings:

ecobee4 setup wizard asking who I want to configure my WiFi settings

Because I have an iPhone, I chose the “Use iPhone, iPad, or iPod” option, which utilizes the HomeKit setup wizard on iOS devices for painless WiFi configuration. I could have also opted for the traditional “select the network, type in the WiFi password” option. Both will result in the ecobee4 displaying a happy screen like this one:

ecobee4 WiFi connected

The next step was to link my ecobee4 with my ecobee account using the ecobee app on my iPhone (naturally, there’s also one for Android devices). If you don’t have an ecobee account yet, the app lets you set one up. Once the account is set up, the ecobee4 generated a registration code which I typed into my app:

Registration of ecobee4 via iPhone app

Once the ecobee4 was linked to my ecobee account, the mobile app prompted me to set up the Amazon Alexa Voice Control:

ecobee4 Amazon Alexa Voice Control setup

All I needed was my standard Amazon.com username and password. An Amazon screen popped up in the app to confirm I wanted to link the ecobee4 with my account:

Amazon Alexa link confirmation

Once I agreed to the terms, the ecobee app confirmed that the Alexa Voice Control service was now linked to my ecobee4:

Alexa link to ecobee4 confirmed

Testing Alexa Voice Control with the ecobee4

With the ecobee4 installed, configured, and linked to my ecobee and Amazon accounts, I was ready to test things out. I started with a simple “Alexa, hello!” The ecobee4 lit up the blue indicator on top and replied “Hello!” Success!

The ecobee4 lights up in blue when Alexa listens or speaks

I then tried “Alexa, good morning,” to which she responded with an interesting fact about the day, and having a “case of the Mondays.” The reply took about 30 seconds, which allowed me to tinker with the volume settings to my liking. Because of the speaker location against the wall, or perhaps because I used the optional trim ring, Alexa’s voice quality isn’t quite as clear as she sounds when coming from a full-sized Echo or the smaller Echo dot. Still, it was easy enough to understand and interact with her.

I logged into my Alexa mobile app and saw that my ecobee4, which I’d named “Maple Kitchen” (we live in Maple Valley, WA and this unit is in the kitchen) appeared there alongside all my other Alexa devices:

ecobee4 appears as an Alexa device in the Amazon Alexa mobile app

It was at this point of my testing that I bumped into my first minor snag. By default, the wake word for every Amazon Alexa-driven device is “Alexa.” It’s the word she listens for, and then “wakes” to interpret the next words you say. For most users, the default is fine. But for users with family members whose names sound very similar to “Alexa,” it makes sense that you’d want to change that. On Amazon Echo devices, you can select between “Alexa,” “Amazon,” “Echo,” or “Computer” as possible wake words. Star Trek fans will, naturally, select the latter.

We chose “Echo,” because we have a teenager at home whose name sounds a lot like “Alexa,” and any time anyone said her name, our Echo would wake and complain about not understanding the command. I also have a friend who works on Amazon’s Echo team who has a daughter named “Alyssa,” so he also changes his wake word from the default. So when I selected “Maple Kitchen” in my Alexa app and tried to change its default wake word, there was no “Wake Word” option for the ecobee4. I’m certain this is a fix that ecobee will eventually push via firmware, and I suspect that there’s a natural lag between when Amazon releases an Alexa feature directly to its own devices vs. when it releases them to third-party partners like ecobee. So for now, we’re forced to call our ecobee4 “Alexa” when we want to talk to her, but she’s already been waking up multiple times a day when we’re not actually trying to talk to her, and I’m sure that families with an Alexis, Melissa, Alyssa, Alex, or similar names will also bump into this issue. I’ll update this review if/when that gets addressed. Update: ecobee has confirmed that this issue will be addressed via firmware update.

Next, I tried to control the ecobee4 temperature settings via voice command. I said “Alexa, set Maple Kitchen to 73.” But Alexa complained that there was no device named “Maple Kitchen.” I realized that was likely being caused by the fact that I hadn’t re-run Alexa’s discovery of smart devices, which you can do with a button press in the Alexa app or via voice command. I chose voice command and said “Alexa, discover devices.” The ecobee4 responded “Starting discovery. This can take up to 20 seconds.” A few seconds later, she responded “Discovery is complete.” I tried the original command again: “Alexa, set Maple Kitchen to 73.” But again, she complained that there was no device named “Maple Kitchen,” and while I could see the ecobee4 in the list of Alexa devices in my Alexa app, it did not appear on my list of Smart Home devices (which is in a different section of the app). I verified that the ecobee skill was enabled in the Alexa app, but decided to also try disabling and then re-enabling the ecobee skill in the Alexa app, then re-running Alexa discovery. That did the trick! The “Maple Kitchen” device showed up as an “ecobee thermostat” device in my list of smart devices:

ecobee4 shows up as a smart home device in the Alexa app

When I once again said “Alexa, set Maple Kitchen to 73,” she replied “Setting Maple Kitchen thermostat heat to 73.” Success!

For those adding their first ecobee thermostat, I doubt this will be an issue. But if you’re adding an ecobee4 to a setup where you already have the Alexa ecobee skill enabled in the Alexa app, I recommend disabling and re-enabling the skill (it’s literally two button presses) to make sure Alexa’s Smart Home discovery can “see” your new ecobee4.

For the most part, the Alexa Voice Service seems to be able to do most of what an Amazon Echo can do, with a few exceptions. Messaging doesn’t work yet, but that’s a very new Alexa feature that was released by Amazon last week, and I have no doubt it will eventually show up on the ecobee4. The ecobee4 did let me stream Amazon music and my Audible playlist to the ecobee4’s internal speaker, but didn’t support (yet?) Pandora, Spotify, or some of the other third-party services music I attempted from my Alexa app.

The ecobee4 gladly streamed my Audible books and Amazon music via the Alexa app, but doesn’t (yet?) support Pandora or Spotify

I was also able to freak my wife out when I launched American Gods from my Audible playlist to the Maple Kitchen ecobee4 while I was away from the house:

Alexa app lets you stream to the ecobee4 even when you’re not home

Remember, the ecobee4’s internal speaker isn’t really designed for audiophiles, so your music is going to sound as if it’s streaming from a small speaker in a thermostat on your wall. The real power of the ecobee4’s Alexa integration is as a home automation interface rather than a streaming audio device.

ecobee4 Integration with Smart Home Technologies

While the ecobee4’s Amazon Alexa integration is touted on the front of the box, the ecobee4 also works with the most popular home automation ecosystems such as Samsung SmartThings, Apple HomeKit, and IFTTT. It took me less than 3 minutes to configure the ecobee4 to work with my SmartThings Hub, which allowed me to I was able to control the ecobee4’s functions through my SmartThings mobile app as well as via voice control through Alexa to SmartThings (which works very much like Alexa’s direct control of an ecobee thermostat via the Alexa ecobee skill).

The ecobee4 integrates easily with Samsung SmartThings

I was also able to use the ecobee4’s Alexa service to trigger other SmartThings events throughout the house that were completely unrelated to the thermostat. I also connected my ecobee4 to my IFTTT account, and configured a geo-fence so that my ecobee4 would automatically kick into “Home” mode as I approach the house, rather than wait for one of the motion sensors to determine I was in the house:

ecobee4 integrates easily with IFTTT

In each case, integrating the ecobee4 with these third-party automation protocols was as simple as a few clicks. The ecobee4 is without question the “most open” smart thermostat on the market, supporting multiple methods of monitoring and control — whether via Amazon Alexa, Apple HomeKit, Samsung SmartThings, IFTTT, or other home automation protocols still on the horizon.

ecobee4 Thermostat Functions

Of course, this wouldn’t be a thermostat review if I didn’t stop geeking out long enough to test the actual thermostat functions of the ecobee4. Quite simply, these work exactly as they do in the ecobee3. The included remote sensor is the same as those that ship with the ecobee3. The interface is the same. The programming is the same. The reports are the same. The ecobee3’s smart thermostat functions are excellent, so I appreciate ecobee’s “if it ain’t broke, don’t fix it” approach. Those familiar with the ecobee3 functions and interface will feel right at home with the ecobee4.

Like its predecessor, the ecobee4 supports a wide array of heating and cooling setups, and also supports one auxiliary device, such as a humidifier or dehumidifier. Again, none of this was broken on the ecobee3, so it was nice to see ecobee not trying to “fix” it for the latest model. There are a lot of ecobee3 users out there whose input drives development, so I predict that anything that does pop up on the ecobee3 that requires fixing will be pushed out via firmware to both the ecobee3 and ecobee4. Basically, the the ecobee4 does everything the ecobee3 can do, plus HomeKit support (which showed up part-way through the ecobee3’s run) plus the ecobee4-exclusive Amazon Alexa service.

Remote Sensors with the ecobee4

The remote sensors work with the ecobee4 exactly as they did with the ecobee3. Again, not broken, so no need to fix. The remote sensors are, for me, one of the biggest advantages the ecobee line of thermostats has over competitors like Nest and Honeywell. They help save money while keeping the parts of the house currently occupied more comfortable.

I was using two remote sensors with my ecobee3, but I forgot to remove them via the touch-screen interface before replacing my ecobee3 with the ecobee4. I had to remove the battery from the sensors for 5 minutes to reset them, then pair them with the ecobee4. The next time I upgrade one of my ecobee3 units to an ecobee4, I’ll try to remember to remove the sensors via the ecobee3’s touch-screen interface prior to unplugging it. But even if I forget again, it was a simple enough process to re-pair the sensors with the ecobee4.

Room for Minor Improvements

This is the part of the review where I try to be as brutally honest as I can and point out the things I’d like to see changed on the ecobee4. The most pressing, for me at least, is the ability to change the wake word to something other than “Alexa.” I haven’t received any official word from ecobee, but my guess is that this is a temporary limitation driven by Amazon, and that ecobee will push a firmware fix for this soon. Update: My initial guess was right, and ecobee has confirmed to me that this is on the way. I’ve also mentioned that Alexa messaging doesn’t work as of this review, but again, I’m confident that’s an Amazon issue and that a future firmware push will support it. The lack of full compatibility with music streaming services via the Alexa app isn’t that big of a deal for me, especially since the ecobee4’s speaker is too small to sound great with streaming music. Alexa’s voice quality from the ecobee4’s speaker is just “OK.” It works fine, but sounds a bit garbled if you turn the volume up too high, and doesn’t match the voice quality of a standalone Echo or Echo Dot unit. I doubt that’s something that can be addressed via firmware, but if it is, that’s something else worth a few moments of brainstorming at ecobee headquarters.

Oh, also… can we please, please, pretty please get the clock back on the main display? You know, the one that appeared on the original first-gen ecobee Smart? The one I’ve been begging for since the ecobee3 first launched? Thanks, guys. That would be awesome.

Final Thoughts

So is the ecobee4 right for you? If you don’t already have an ecobee and have been considering a smart thermostat, it’s a no-brainer. The ecobee4 keeps all the functionality of the ecobee3, adds the majority of the functionality of an Amazon Echo Dot (and I suspect most of what’s missing isn’t far behind), and still comes in at the same MSRP as the previous model of $249.I still prefer ecobee’s schedule-based approach to heating and cooling over the Nest’s method of overly optimizing, and the remote sensors extend that advantage. The addition of Alexa Voice Control to what was arguably already the smartest smart thermostat on the market makes calling the ecobee4 a “smart thermostat” feel insufficient. There are plenty of “smart thermostats” out there. The ecobee4 is the world’s first “genius thermostat.”

Even if you have another brand of smart thermostat, I’d still suggest considering a switch to the ecobee4. No other thermostat has its features or supports as many methods of home automation integration. If you were an early adopter of smart thermostats and your first-gen T-stat is getting a bit long in the tooth, the ecobee4 is a perfect upgrade.

I could only think of two groups who might want to think twice about whether or not the ecobee4 is the right choice. The first group is ecobee3 owners who already have an Amazon Echo device. They already have the same functionality, albeit shared between two devices, but if they really want the latest and greatest and have a good “hand-me-down” destination for their ecobee3 (or can sell it on eBay), they may want to upgrade anyway.

The other group where it might not make sense is homeowners whose thermostat location wouldn’t work well as an Amazon Alexa microphone location. For them, the ecobee3 (which now has a lower price) might be a better choice. Our kitchen is a perfect place for an ecobee4, because it’s an area where we spend a lot of time, and where we already use an Amazon Echo (which I plan to move downstairs to the gym now that the ecobee4 can handle its workload). But our upstairs level where three bedrooms are located (and which has its own HVAC system) has an ecobee3 mounted in the hallway. Anyone wanting to interact with Alexa would have to yell into that hallway. We’ve got an Echo Dot in the bedroom, where we don’t have to shout at it to get weather updates, control lights, or play music. But if your thermostat is located in a high-traffic area of your house, the ecobee4 is the perfect device for you.

If the ecobee4 is right for you, you can pick one up right now on Amazon.com. The ecobee4 ships with one remote sensor, but you can purchase additional sensors for maximum coverage and flexibility.

ecobee4 Launch Giveaway

To celebrate the launch of the ecobee4, ecobee was generous enough to provide four units to give away to my blog readers! If you’d like a chance at winning one, just do the following three things:

1. Subscribe to my blog by putting your email address here (don’t worry, I won’t spam you).

[blog_subscription_form]

2. Follow me on Twitter: @sjjenkins.

3. Tweet a link to this review and and tell me why you should win a free ecobee4 using the hashtag #ecobee4giveaway. Be sure to include @sjjenkins and @ecobee at the end of your tweet with the hashtag.

Winners will be chosen at random and must live in the continental US to be eligible (yes, I’ll check to make sure you met all three conditions). I’ll announce the winners on my Twitter feed when I make the random selection at the end of the month. Good luck!

Join the Conversation

As always, I welcome your questions, comments, and feedback via the comments below, on my Twitter feed, or share your own rating of the ecobee4 Smart Thermostat below!

The post ecobee4 Smart Thermostat Review appeared first on Steve Jenkins.

With a street price of under $90, the ER-Lite was a tough act to beat for the money. But later, when Ubiquiti Networks released the newer EdgeRouter X with a street price right around $50, many (including me) wondered if the even more budget-priced ER-X could produce similar routing performance on high speed networks like Google Fiber.

The UBNT EdgeMax forums and Ubiquiti Subreddit are rife with questions, discussions, and comparisons between the ER-L and ER-X, asking which is “better.” Many insist that based on the hardware and specs, one will be theoretically faster than the other, while others argue they’ll theoretically be the same.

What I didn’t see, however, were any posts from anyone with a high speed Internet connection who actually tested the two routers back-to-back and posted actual results. So I bought an EdgeRouter X to compared, and performed some speed tests last night and this morning vs. the EdgeRouter Lite.

Speed Test EdgeRouter Configurations

First, I configured the ER-X as closely as possible to my ER-L’s working configuration. The two major differences in their configurations were:

1. The ER-X had ports eth1 – eth4 configured as switch0 (because the ER-L has no hardware switched ports).
2. The ER-L has 12 port forwards and 32 DHCP reservations configured (because it’s my working router) while the ER-X had 2 port forwards configured and no DHCP reservations. I didn’t (and still don’t) know how whether that would affect the test, but I don’t think it mattered much in the final results.

Other than those two differences, the two routers’ configurations were as close to identical as I could get. Both configs were essentially the config.boot.erl and config.boot.erx example configuration files in the Google Fiber directory of my UBNT EdgeRouter Configuration repo. DPI, which is considered one of the more CPU-intensive features, was enabled on both routers.

• I ran two sets of tests on two days. On Friday night, I ran two sets of three “grouped” tests: three tests in a row on the ER-X followed by three tests in a row on the ER-Lite. On Saturday morning, I ran a total of 10 “alternating” tests: one test on the ER-Lite, then one on the ER-X, switching after each test until both routers had completed five tests each.
• The tests were conducted using Google’s own test site at Provo.Speedtest.GoogleFiber.net.
• The group tests were performed between 10:55 AM and 11:01 AM Mountain Time on Friday night (2/17/17).
• The alternating tests were performed between 8:26 AM and 8:34 AM Mountain Time on Saturday morning (2/18/17).
• All tests were performed using Google Chrome on a MacBook Air.
• Each router’s WAN port was connected directly to the Google Fiber Network Jack.
• Each router’s LAN port was connected directly to the MacBook Air via Ethernet.
• No other network clients or cables were connected to the routers during testing.

EdgeRouter X vs. EdgeRouter Lite Speed Test Results

To make analyzing the data easier, I put all the speed test results into a publicly shared Google Sheet.

The sheet shows the timestamp of each test, so you can see that not much time passed between each speed test.

I calculated overall average download and upload speeds using both routers across the test periods and showed the MIN and MAX results for comparison.

I used green and red backgrounds to flag the fastest and slowest test results. I also used a green background in the MIN and MAX comparisons to identify the highest minimum and maximum down and up speeds.

All results are shown in megabits per second (Mbps).

Google Fiber speed test of EdgeRouter Lite

Friday Night Grouped Tests

• Highest tested download speed. ER-X: 827 / ER-Lite: 890
• Highest tested upload speed. ER-X: 945 / ER-Lite: 938
• Slowest tested download speed. ER-X: 755 / ER-Lite: 861
• Slowest tested upload speed. ER-X: 861 / ER-Lite: 931
• Combined average down / up speeds over the test period: 829.5 down / 923 up
• ER-X average: 780 down / 912 up
• ER-Lite average: 879 down / 934 up.
• On average, the ER-X was 11.93% slower down and 2.38% slower up.

Saturday Morning Alternating Tests

• Highest tested download speed. ER-X: 880 / ER-Lite: 916
• Highest tested upload speed. ER-X: 933 / ER-Lite: 938
• Slowest tested download speed. ER-X: 740 / ER-Lite: 858
• Slowest tested upload speed. ER-X: 863 / ER-Lite: 930
• Combined average down / up speeds over the test period: 862.3 down / 920.1 up
• ER-X average: 835.2 down / 907 up
• ER-Lite average: 889.4 down / 933.2 up.
• On average, the ER-X was 6.29% slower down and 2.85% slower up.

Interpreting the Results

Based on this quick and dirty testing, it looks like the EdgeRouter Lite barely wins on both downloads and uploads, though it’s a closer race on the uploads. I weigh the Saturday morning results heavier than I do the Friday night results, as alternating the routers seems more likely to help minimize the effect of external factors.

That said, I’m sure purists will be quick to point out that these weren’t laboratory conditions. But residential network users don’t live in laboratory conditions, real world testing seems appropriate for network hardware that intends to live in the real world.

Across all the tests, upload speeds were consistently faster than download speeds. All the ER-Lite’s upload tests in both test periods were in the 900s, while three of the ER-X’s upload tests (one on Friday, two on Saturday) dipped into the 800s. Overall, however, the upload speeds between the two devices were very close.

On Friday night, all three of the ER-Lite’s download tests were faster than the fastest ER-X download. On Saturday morning, four out of five of the ER-Lite’s download tests were faster than the fastest ER-X download test. Only one of the ER-X’s download tests (880) was faster than the slowest ER-L download test (858).

The EdgeRouter Lite consistently “edged out” the EdgeRouter X in uploads and downloads during both test periods, with the exception of one upload test where the ER-X’s max was 7 Mbps faster than the ER-Lite’s max. Apart from that, even though the ER-Lite’s results were only slightly higher, they were higher nonetheless. Whether or not the differences would have been greater with identical port forwards and DHCP reservation configurations potentially eating a few more CPU cycles is hard to say. Probably not, but I suppose it’s possible.

Final Thoughts – EdgeRouter X, EdgeRouter Lite, or EdgeRouter POE?

OK, so the EdgeRouter Lite barely won the real world head-to-head speed battle. But will you actually notice the difference between 916 vs 880 down or 938 vs 930 up during normal everyday Internet use? Probably not. The only time it will likely matter is when you’re screen capping speed tests to taunt your geeky friends on social media.

Beyond that, the speed differences themselves are close enough to not really matter that much. What should matter more are the features you care about in a router. If you have a small number of clients and want to get away without an additional switch, choose the EdgeRouter X and configure four of its ports as a hardware switch. Looking to power multiple PoE access points like the UBNT’s UAPs? Choose the EdgeRouter PoE, which lets you configure three ports as a hardware switch, or an EdgeRouter X-SFP which supports passive PoE over all its ports at half the cost of the ER-POE . Looking for the absolute lowest cost? Get the ER-X. Care about squeezing out the absolute highest speed test result? Get the EdgeRouter Lite. Working with UniFi wireless gear and want all the pretty dashboard features in your UniFi Controller to light up? Forget about the EdgeRouters and pick up a UniFi Security Gateway (USG) — which also runs EdgeOS (though is slightly trickier to customize).

Also, it’s worth noting that I didn’t compare performance of additional features of both routers. I only tested speeds using an online speed test. You’ll have to look elsewhere for comparisons of VPN performance, or IPSec performance, or any other number of things that make head-to-head comparisons more complicated than simply “choose this one.” I knew this going in, and understand it as a limitation of this testing method. I was mainly curious about reported Internet speeds, so that’s what I tested. Take it for what it’s worth.

Regardless of which one you pick, all of Ubiquiti’s router options will perform fine on a high speed Internet connection. What should drive your decision is the needs of your network.

As always, I welcome your comments, questions, and feedback below.

The post EdgeRouter X vs. EdgeRouter Lite Google Fiber Speed Tests appeared first on Steve Jenkins.

Unfortunately, there’s no obvious way to disable the “UPS: Passed a self-test” emails via the Smart-UPS web interface. If you go to Administration -> Notification -> Power Events/Diagnostics, you only see the following:

APC UPS Diagnostic Events interface

While you can enable and disable email notifications for when your UPS starts or fails a self test, a link to “UPS: Passed a self-test” is nowhere to be found. But it does exist… if you know where to look for it.

To find it, click on the link for “UPS: Failed a self-test.” You’ll see the standard window for enabling/disabling notifications for that event:

APC Smart-UPS notification page for UPS: Failed a self-test

Look at the URL of that web page in your browser’s address bar, and note that it ends in the hex value 0x0106. Change the last digit 6 to a 5 in the address bar so that it ends with “code=0x0105” then hit ENTER so your web browser loads the new page. You’ll see the hidden page for setting the “UPS: Passed a self-test” notifications:

The hidden APC Smart-UPS notification page for UPS: Passed a self-test

Now you can uncheck the email address, hit Apply, and the notifications will stop!

Did you find this tip helpful? Let me know in the comments below!

The post How to Turn Off “UPS: Passed a Self-Test” Emails on APC Smart-UPS appeared first on Steve Jenkins.

The ecobee3 lite is very similar to the ecobee3, but does not include support for Wireless Room Sensors, Occupancy Detection, Smart Home/Away features, and Accessories (HRVs, Ventilators, Humidifiers, Dehumidifiers). But you still get all the other smart functions offered by the ecobee3; Smart Recovery, Schedule and Vacation settings, Home IQ, connection to Homekit/Amazon Echo/IFTTT and SmartThings, Alerts and Reminders and our Web Portal and Mobile App.

The majority of smart thermostat users probably won’t need to hook up an an accessory like a humidifier to their system, so the loss of that feature doesn’t feel like that big a deal. The wireless room sensors, however, are what really set the flagship ecobee3 apart from the competition, so that’s where the $80 in savings should really be measured for the average homeowner. If the additional features that come by virtue of the built-in and remote occupancy sensors (Smart Home/Away, Follow Me, and occupancy detection) aren’t worth $80 to you, then the ecobee3 lite is the ideal way for the budget conscious buyer to make their home a little smarter while making their energy bill a whole lot smaller.

I was an early adopter of the original ecobee Smart Thermostat since its launch, and then I upgraded to multiple ecobee3 units at our main home in Seattle, Washington in September 2014. In November 2014, I documented how to upgrade from an original ecobee Smart to an ecobee3 at our place in Utah. The only place I was still using an original ecobee Smart was at our Cabin near Wenatchee, Washington. I was primarily using it for its remote management capabilities, so we could turn on the heat or AC before the 2.5 hour drive and the place would be (mostly) heated or cooled by the time we walked in the door. I didn’t see much of a need to upgrade the original ecobee because the electricity in Wenatchee is the cheapest in the country at 2.5 cents per Kwh, compared to 10 cents per Kwh in Seattle. But when the ecobee3 Lite came out, I decided it was the perfect opportunity to try it out. I have to shamefully admit that when I replaced the old-school Honeywell thermostat that used to be at the cabin, it left a large mismatched paint spot on the wall. I always swore I’d fix it… but never got around to it. So here’s what I was starting with at the cabin, prior to upgrading to the ecobee3 Lite:

Original ecobee Smart Thermostat at the Cabin

Yes, I know, hideous.

What’s in the Box?

I’d previously made an ecobee3 unboxing video, and the unboxing of the ecobee3 Lite was pretty similar, though it’s clear ecobee has made an effort to make things even more sleek and modern. At first glance, the ecobee3 Lite looks a like like its older brother:

ecobee 3 Lite unboxing

There was nothing wrong with the ecobee3 manual, but the ecobee3 Lite manual received a total redesign and is very easy to follow:

ecobee3 Lite manual

Underneath the manual was the mounting plate / wiring harness, mounting hardware, and the updated Power Extender Kit (PEK), which allows you to install the ecobee3 Lite even if you only have 4 thermostat wires available at the thermostat location on your wall.

ecobee3 Lite mounting plate, hardware, and PEK

The new PEK works just like the original PEK, in that it lets you connect to the most common 5 terminals on your furnace’s control board: Yellow (AC), White (Heat), Green (Fan), Common (Ground), and Red (Power).

Five wires coming out of the ecobee3 Lite PEK

The big improvements in the PEK are the push-button quick connectors for the wire coming from the wall location:

ecobee3 Lite PEK push-button connectors

…and a magnetic pad on the rear of the new PEK, which makes it easy to mount to nearby metal surfaces such as a furnace panel or sheet metal ducting:

Rear of the ecobee3 lite PEK has a magnetic mounting pad

Fortunately, I had plenty of wires available at the cabin, so I didn’t need to use the PEK. I just thought there were enough improvements in the new PEK that I still felt it worthwhile to share them.

The ecobee3 Lite’s mounting plate looks very much that of the ecobee3, just a bit smaller and with fewer connectors, but still comes with the bubble level to make OCD wall-mounters happy:

ecobee3 Lite wall mount

The box also contains a larger plate to cover up any holes or paint mismatch from previous installs:

Larger plate for covering up old thermostat locations

Out with the Old, In with the New

Because I was upgrading from an original ecobee Smart, I went through all of my settings screens and took photos. This makes configuring the replacement a lot easier. Taking a photo of the old thermostat’s wiring diagram is always helpful:

ecobee Smart’s wiring diagram

But even if you’re starting from scratch and not replacing an older ecobee with an ecobee3 Lite, the easy-to-follow instructions in the manual will get you up and running quickly.

After snapping photos of my settings, I shut off the power to the furnace (very important) then disconnected my old thermostat (ignore the additional wires to the left, which are from a Honeywell air filter).

Removing the old thermostat

Then I went downstairs where the old ecobee’s equipment interface was located, disconnected it, and connected the wires from the upstairs thermostat location directly to my furnace’s wiring terminal. If you’re replacing a traditional thermostat, you won’t need to do that step. The newer ecobee3 and ecobee3 Lite no longer require an equipment interface, so most homeowners will only have to re-wire things at the thermostat wall location (but read the manual for your particular setup).

My old thermostat was not centered on the wall, and it didn’t have enough slack for me to center my ecobee3 lite, so I decided to wire-nut a couple feet of thermostat wire to the existing wires so I had more flexibility in my install. I also decided to cover up that ghastly unpainted spot once and for all, so I hit the local Hobby Lobby and spent $20 on a rustic wood panel wall-hanging. It was a bit too wide, so I used a saw to chop down the sides to the width I needed, wiped some wood stain on cut ends with some stain, drilled a hole in the middle for my extended thermostat wire, mounted the ecobee3 lite mounting plate directly to the wood, then screwed the wood panel into a wall stud (thank goodness it was centered):

ecobee3 wall plate mounted to wood panel

With plenty of slack, I was able to easily connect the appropriately-colored wires to the ecobee3 Lite mounting plate, then tuck the slack behind the wood panel:

ecobee3 Lite mounting plate installed

Finally, I gently pressed the ecobee3 Lite into the mounting bracket and turned on my furnace. After a few seconds, the unit sprang to life!

Getting Set Up

Registering the ecobee3 Lite was easy with an iPhone. When I opened my iPhone’s WiFi screen, the ecobee3 Lite as there waiting to be set up:

iPhone setup of ecobee3 Lite

It was as simple as typing in a code that appeared on the ecobee3 Lite’s screen to sync it with my ecobee account so I could manage it remotely:

Connecting the ecobee3 Lite to an ecobee account

Once everything was set up, my ecobee3 Lite was happy, and went into its automatic temperature calibration mode:

ecobee3 Lite calibration mode

While it calibrated, I used my app to check the differences between the ecobee3 and ecobee3 Lite’s app menu options. As expected, they looked very similar. The ecobee3 at my Seattle house menu looks like this:

ecobee3 mobile app main menu

And the ecobee3 Lite looks the same, without the Sensors menu item:

ecobee3 Lite mobile app main menu

Finishing Touches

After setting up my heating and cooling preferences (some via the ecobee3 Lite’s screen and some via the app), I decided to dress up the install a bit. I took the antlers that had been on the wall above the old thermostat and mounted them to the wood panel:

Antlers added to punch up the rustic vibe

When everything was done, the finished product looked like the perfect mix of modern and rustic:

Install complete!

Making Your Smart Home Smarter

If you’re into more extensive home automation, I run a SmartThings hub at the cabin, as well as an Amazon Echo Dot, both of which integrate wonderfully with the ecobee3 and ecobee3 Lite. I can control the ecobee3 Lite using the ecobee app, or I can trigger thermostat events based on any SmartThings sensor or input (like automatically lowering the temperature when my SmartThings hub senses my phone has left the cabin for the afternoon), or I can say “Alexa, turn the Cabin thermostat up 2 degrees.” With an MSRP of $169, the ecobee3 Lite is an unbeatable choice for those looking to connect a smart thermostat to a more widely integrated smart home.

Final Thoughts

So is the ecobee3 Lite right for you? If the price of the flagship ecobee3 is out of your budget, then the less expensive ecobee3 Lite is a great way to get the basic features of the more expensive model without breaking the bank. If you’ve got a home that wouldn’t benefit from remote sensors, the ecobee3 Lite is a better choice than anything else on the market. It’s just as easy to install, set up, and use as its slightly older brother. And it looks just as good on your wall, too… just make sure it’s centered.

If you’re looking for a more in-depth review of the ecobee3 Lite, check out my original ecobee3 review, my ecobee3 two month follow-up, and my HomeKit-enabled ecobee3 review. The ecobee3 Lite shares the majority of the ecobee3’s features, is HomeKit-enabled, and uses the exact same app and web interface as the ecobee3. The only difference is the absence of remote sensors and occupancy detection (meaning Smart Home/Away and Follow Me features aren’t available), and the ability to add accessories like a humidifier. Apart from that, the ecobee3 Lite looks and acts exactly like a HomeKit-enabled ecobee3.

I welcome your questions, comments, and feedback below! But if you’ve got an ecobee3 or ecobee3 Lite technical question, you’re better off contacting ecobee support or joining the conversation at SmartHomeHub.

The post Review: ecobee3 Lite appeared first on Steve Jenkins.

The answer is, as is usually the case, “it depends.” On more modern router hardware (higher speed processor and more available RAM), I suppose it’s possible. But I wanted to see if it was true for the Linksys E4200 router. As of 2017, it’s a bit long in the tooth in hardware terms, but it’s still a very popular device for third-party firmware fans. My testing was limited only to the E4200 v1, so while it’s probably safe to extrapolate my results to other Linksys E-series router of approximately the same vintage, any firm conclusions beyond that will require your own testing.

Kernel Version is Different than Firmware Version

Before I get to the testing, I’ve seen some confusion regarding DD-WRT version numbers and the Linux kernel version on which they are based. If you look on your DD-WRT router’s “Status / Router” tab, you’ll see a “Kernel Version” and a “Firmware Version:”

Kernel Version vs Firmware Version

Don’t confuse a K3.x build (meaning based on a Linux 3.x kernel) with the v3.0 version number. They are not related at all, and it’s just coincidence that they kind of match right now. All the DD-WRT builds I tested were DD-WRT v3.0.

Testing Method

I ran all these tests on January 1, 2017, while enjoying some vacation time at our cabin in East Wenatchee, Washington. Because it sits on the Columbia River, the location is ideal for speed tests; it’s connected to the high-speed fiber network installed by the public utility district to link the hydro-electric dams along the river. Localtel, my local ISP, hosts a test server on Speedtest.net, which I used for all tests.

My cabin sits in a sparsely populated area, and the only two neighbors close enough to create channel interference use WiFi channels 1 and 11, leaving channel 6 wide open for me on the 2.4Ghz band. Neither of my neighbors use the 5Ghz band.

Wired testing was done using a MacBook Air 7,2 (early 2015 vintage) and a Cable Matters SuperSpeed USB 3.0 to RJ45 Gigabit Ethernet Network Adapter with a 3′ Cat6 patch cable. WiFi testing was done using the same MacBook Air’s internal WiFi adapter, as well as an iPhone 6. For both types of tests, no other network clients were attached to the router.

Wireless speed testing was performed separately on the E4200’s 2.4Ghz and 5Ghz antennae by disabling one of the antennae at a time. A custom SSID was used to ensure no other wireless clients were connected besides the two test clients.

The router’s nvram was erased and the router rebooted before and after each firmware flash. All firmware default settings were used, with the exception of the following:

• Router was overclocked to 533Mhz.
• WiFi channels were set to 6 on the 2.4Ghz band and 36 on 5Ghz band, with no interference on either band. Only one band was active at a time.
• 2.4Ghz band was set to “N-Only” and 5Ghz was set to “NA-Mixed.”
• Wide HT40 (40Mhz) channel width was set on both bands.
• WPA2 Personal encryption was enabled.
• SSHd and UPnP were enabled (as both likely would be in a real-world setting).

I used a custom SSID of “E4200” and the password “testing123” for the WiFi tests to ensure no rogue clients could connect during testing.

After each flash, I ran a total of 15 speed tests:

• Three wired speed tests.
• Six 2.4Ghz band tests alternating between the MacBook Air and the iPhone until each had completed three tests.
• Six 5Ghz band tests alternating between the MacBook Air and the iPhone until each had completed three tests.

After completing all 15 tests on all firmwares, I noted there was no statistically significant difference between the wireless speeds of the MacBook Air compared to those of the iPhone. In general, the iPhone’s WiFi tests were slower than the MacBook Air’s, but they always correlated in a comparison of the head to head tests, i.e. if the MacBook Air was faster on one firmware vs another, so was the iPhone. For this reason, and to make the reports below a bit easier, I’ll only include the MacBook Air’s WiFi results in my findings. But know that the iPhone’s speed tests did nothing to skew the results. They match up perfectly with the MacBook Air’s.

I’ve also included the highest download and upload speeds in each batch of results below, and will use each firmware’s highest value to compare. Taking an average of speeds isn’t that useful or statistically accurate, since I’m trying to test to maximum potential speed in a real-world setting.

It’s also interesting to note that based on the test results, my ISP is clearly attempting to cap download speeds, and doesn’t appear to do so on the upload speeds. So for the majority of tests, the upload speed was faster than the download speed.

E4200 DD-WRT and Tomato Firmwares Tested

I started with a fresh flash of the latest stock Linksys firmware, Version 1.0.06 (build 3): FW_E4200_1.0.06.003_US_20140520_code.bin.

Next, I tried testing with the latest-available build of DD-WRT, which was 30949, released on December 15, 2016. But it proved too unstable on the E4200 to test properly and would lock up the router requiring a power-cycle, so I used build 30880, which (as of the writing of this article) is the stable build I recommend in my E4200 Max Speed settings article.

I flashed and tested DD-WRT builds in this order:

• K2.6 Mini (trailed build): dd-wrt.v24-30880_NEWD-2_K2.6_mini-e4200.bin
• K2.6 Mega: dd-wrt.v24-30880_NEWD-2_K2.6_mega-nv60k.bin*
• K3.x Mega (trailed build): dd-wrt.v24-30880_NEWD-2_K3.x_mega-e4200.bin
• K2.6 Kong 22000++ build: kingkong-nv60k-broadcom.bin**

I included the K2.6 Kong 22000++ build in my tests because it’s been a solid and stable choice for the E4200 since 2014. During testing of the above builds, there was no measurable wired or wireless speed differences between any of the DD-WRT K2.6 builds, so only the DD-WRT K2.6 Mini results are included below.

I also decided to include two of the most popular Tomato K2.6 firmwares for the E4200: the Shibby build tomato-E4200USB-NVRAM60K-1.28.RT-N5x-MIPSR2-138-Mega-VPN.bin and the Tomato RAF build tomato-E4200USB-NVRAM60K-1.28.9014MIPSR2-RAF-v1.3g.bin.

Finally, just for fun, I tested the hardware I used to replace the E4200 router at my cabin: a Ubiquiti EdgeRouter Lite and UAP-PRO access point (though I now recommend the newer UAP-AC-PRO for anyone shopping for an access point). Those results are also included below.

All the above third-party firmwares for the Linksys E4200 (including the stock firmware) are available in my DD-WRT firmware repository at http://ddwrt.stevejenkins.com/.

E4200 Router Speed Test Results

Stock Linksys Router Firmware – Version 1.0.06 (Build 3)

Stock Wired Speeds

Stock 2.4Ghz WiFi Speeds

Stock 5Ghz WiFi Speeds

DD-WRT K2.6 Mini Results – Build 30880

DD-WRT K2.6 Mini Wired Speeds

DD-WRT K2.6 2.4Ghz WiFi Speeds

DD-WRT K2.6 5Ghz WiFi Speeds

DD-WRT K3.x Mega Results – Build 30880

DD-WRT K3.x Wired Speeds

DD-WRT K3.x 2.4Ghz WiFi Speeds

DD-WRT K3.x 5Ghz WiFi Speeds

Tomato RAF K2.6 Results – Ver 1.28.9014

Tomato RAF K2.6 Wired Speeds

Tomato RAF K2.6 2.4Ghz WiFi Speeds

Tomato RAF K2.6 5Ghz WiFi Speeds

Tomato by Shibby K2.6 Results – Ver 1.38

Tomato by Shibby K2.6 Wired Speeds

Tomato by Shibby K2.6 2.4Ghz WiFi Speeds

Tomato by Shibby K2.6 5Ghz WiFi Speeds

Ubiquiti (UBNT) EdgeRouter Lite Router + UAP-PRO Access Point

UBNT ERL + UAP-PRO Wired Speeds

UBNT ERL + UAP-PRO 2.4Ghz WiFi Speeds

UBNT ERL + UAP-PRO 5Ghz WiFi Speeds

Results Comparison

The following table shows the fastest tested download and upload results for each of the firmwares tested (again, there was no difference between any of the DD-WRT K2.6 build variants tested, so only the K2.6 Mini is shown). You can sort the table by clicking on the header or filter using the search box.

Final Analysis

For hardwired connections on an E4200, the stock Linksys firmware can’t be beat at 417.17 / 305.83 Mb/s. The K2.6 kernel DD-WRT couldn’t even reach half those wired speeds with 52.68 / 69.03 Mb/s. Even worse, the K3.x build was almost half again as slow as the K2.6 with results of 81.04 / 72.57 Mb/s.

For 2.4Ghz WiFi, the stock firmware is again significantly faster than the K2.6 DD-WRT and Tomato builds. And while the K3.x DD-WRT speeds were only slightly slower than the K2.6 speeds, they were still slower: 52.68 vs 43.32 down and 69.03 vs 61.63 up.

For 5Ghz WiFi, the stock firmware finished with a clean sweep. At 172.20 / 201.75 Mb/s, it clobbered the K2.6 results of 93.18 / 65.87 Mb/s. The K3.x was, once again, the slowest of the bunch at 68.57 / 63.90 Mb/s.

Bottom Line: DD-WRT K2.6 vs K3.x

If you have a Linksys E4200 and care only about speed, keep the stock firmware. If you’re willing to give up some speed in return for DD-WRT’s features, always choose a K2.6 build over a K3.x. The Tomato results back this up, as they were similar to the DD-WRT K2.6 speeds and also faster than the DD-WRT K3.x speeds.

Of course, on newer hardware with a faster processor and/or more RAM, it’s possible a K3.x build could be just as fast (or maybe faster?) than the same K2.6 build. But if you have an older Linksys E-Series router, the bottom line is that a K2.6 build will always be faster.

I should also note that I no longer use the E4200 as my primary router. I’ve switched to UBNT routers like the UniFi USG (which I run at our main house) and the EdgeRouter (which I run at two of our secondary residences). Both require a separate access point such as the UAP-AC-PRO, but the performance and features made it an easy choice for me. I still keep old ones around as client bridges, and as back-up WiFi routers if I have an equipment failure.

Do you still run a Linksys E4200 or other E-Series router? Do you still want to run a K3.x build instead of a K2.6 one? Have you tested K2.6 vs K3.x on different hardware with different results? Tell me about it in the comments below!

The post DD-WRT K2.6 vs K3.x vs Tomato Router Firmware Head to Head on Linksys E4200 appeared first on Steve Jenkins.

You have three choices:

1. Pay $789 (I wish I was kidding) for replacement Ferrari part #61917100. On one parts supplier’s website, the parts description actually says “Yes, that’s the correct price.  These things have become frightfully expensive as stocks dwindle.”
2. Pay $50-ish on Amazon for BMW part #61311356193, the hazard switch on a 1971 BMW 2002, which is the identical part — except for the red plastic cover piece, which quickly screws off and can be swapped with your original one.
3. Disassemble the original switch and spend 5 minutes fixing it yourself… for free.

Initially, I did step #2. I figured for just under $50, it was worth having a new switch in place. I ordered the part so that it would arrive at my Utah house (where I keep my 328) while I was there for two weeks… but when I got the Amazon notification yesterday that the part had arrived, but didn’t see it in my mailbox, I double-checked and realized that I’d accidentally sent the part to Seattle. I considered having another one shipped, and then returning the first one when I got back to Seattle, but I decided instead to at least take a peek inside the switch to see if I could figure out what was going on first.

Removing the Ferrari 328 Center Tunnel Console Panel

Here’s a look at the center tunnel console panel in the Ferrari 328. The hazard light switch is right in the middle:

Ferrari 328 Hazard Switch location

I first used my trusty Benchmade Triage 915SBK pocket-knife to gently pry off the black plastic cover that surrounds the switch, which showed me that I’d actually need to access the switch from the underside.

The flat console panel in the center tunnel is held down by four screws — the first two are easy, and the last two are slightly more complicated, but still not difficult. Start by removing the two screws that secure the ash-tray in place, which in turn secure the front portion of the panel:

Ferrari 328 ashtray screws

Using a pocket-knife or thin screwdriver, gently pry the middle of the front and rear edges of either the driver’s or passenger’s side temperature adjustment switch. I say “gently” because a replacement switch will cost you $962. The middle portion of the switch is the strongest, and also where metal tabs hold it into the console, so as long as you pry gently there’s little chance of breaking anything. Once the switch is out, reach down into the console and push out from underneath the two middle switches in the rear-most row of four: the antenna switch and the fog-lamp switch.  Gently remove the switches from their wiring harnesses. In this photo, I’d removed both the driver and passenger side temperature switches while I was exploring, but you really only need to remove one:

Ferrari 328 console switches removed to access the last two screws that secure it

Directly below those two center switches are the remaining two screws holding the console panel in place:

Two rear screws holding the panel into place

It helps to have a magnetic screwdriver, or a magnetic pick-up tool (like this one that I often use in gunsmithing), so that you don’t drop the screws and washers down into the tunnel.

Removing the Ferrari 328 Hazard Switch

Once those screws are removed, gently lift the console up and turn it to one side to reveal all the wiring under the console switches. I also unplugged both temperature switches to make things a little less crowded while working:

Center console wiring exposed

Remove the round plastic wiring harness from the underside of the hazard switch. Don’t worry about remembering which tab fits into which harness slot, it will only line up and connect one way. There’s a separate single harness that connects to the center tab of the switch, so gently pull that off, too.

Now unscrew the black plastic ring that secures the switch to the console (it should only be hand-tight). It should come off easily, as well as a slim silver washer with it.

Your hazard switch should now easily lift out of the top of the console:

Ferrari 328 hazard switch removed

If you ordered a replacement switch (and if you did, I hope you ordered the BMW one), simply unscrew the red plastic button cap with the triangle on it, swap it with the BMW one that has the word “HAZARD” printed on it, and put everything back together.

BTW, if you see a Ferrari 328 with the word “HAZARD” on the switch, that means someone switched the switches… without switching the cap. As a judge, I’d deduct at least half a point for that at a Ferrari concours.

Repairing the Ferrari 328 Hazard Switch

The hardest part (for me) of this entire operation was this next bit: cracking open the outer body of the hazard switch. Examine the switch closely, and you’ll see a number of plastic tabs on the bottom portion that click into openings on the top portion. You’re going to bend and mar the plastic a bit, but you’ll never see it, so that’s OK. I ended up using a combination of a utility knife:

Use a utility knife to pry apart the switch

And a small flat-head screw-driver:

A small flat-head screwdriver can help pop the switch apart

Just work around the outside of the housing until you “pop” the top and bottom apart. The switch is comprised of the top plastic housing, bottom housing with metal tabs, a center spring, and the button assembly:

Ferrari 328 hazard button components

Examine the bottom portion of the switch and you’ll find a silver tab with a small pin near the top:

Silver tab with pin

Here’s an overhead shot, where you can see the pin near the top of the tab:

Top view of the silver tab with pin

Partially re-assemble the switch with the spring and the button assembly to get an idea of the way the switch is designed. It’s rather elegant, actually. The pin slides through a channel, which catches and holds the button down. When pressed again, the pin travels through the channels and returns to its lowest point, which allows the button to pop “up.” Line it up and play with it a bit, and you’ll see what I mean:

Hazard switch pin and channel system

What has likely happened to your switch is that this silver tab has been bent outwards, so that it doesn’t catch the channel properly.

This angle clearly shows that my switch’s tab was bent:

A bent silver tab is the likely cause of your switch’s problem

Use a pair of pliers to bend the tab inwards so that it engages the channel as much as possible without impinging the movement of the button. This will ensure that the button will “catch” in the off position once again. Here’s mine after being bent back into a straight position:

Silver tab bent back into position

Re-Assembling the Ferrari 328 Hazard Switch

Once you’ve bent the silver tab back into place and tested to make sure the button can catch properly, it’s time to re-assemble the hazard switch. The first couple of times I tried to do this, the red plastic cap got caught on the top section of the switch housing, and made it difficult to push things back together. Temporarily unscrewing the red plastic portion made things much easier:

Removing the red cap helps when re-assembling the switch

Insert the spring and button assembly into the bottom portion of the switch housing, then line up the top and bottom portions of the housing by finding the flat section on both potions — see the flat section in the above photo where the threaded area is missing. With the flat sections lined up, make sure all the tabs on bottom line up with a cutout section on top, the quickly snap the two sections back together. This may take a couple tries to get right.

Now test your switch to make sure the button stays down:

Ferrari 328 hazard button in the down position

Also test the up position:

Ferrari 328 hazard button in the up position

Re-Installing the Ferrari 328 Hazard Switch

Re-install the hazard switch by doing everything in the reverse order. First, however, I recommend temporarily plugging-in the inner (single) and outer wiring harnesses onto the bottom of the switch and testing your work… before you go through the hassle of buttoning everything back up. Hopefully, your test is successful and your hazard lights can be turned on and off without any issue. If not, crack open the switch again and check to make sure all the gold-colored tabs move freely, and that everything looks like it’s making contact at the right times. If you can’t get it working, go ahead and order the $50 BMW switch. It’s a bargain compared to the price you’ll pay to get a switch in a yellow box.

If your test is successful, unplug the wiring harnesses, re-attach the washer and plastic nut to the underside, secure everything back in place, plug in all the switches, insert the console panel, and tighten the four screws holding it in place.

Give the hazard switch one final test. You may have to adjust the height of the button by unscrewing the red cap a couple turns, so that it sits flush in the down/off position:

Final test of a working Ferrari 328 hazard switch

Congratulations! You just fixed an $800 part yourself. Of course, after this experience, you’ll probably be less inclined to use your hazards unless absolutely necessary (but please do use them when necessary). But even if your switch ever fails in the “on” position by the side of the road again, you might be able to use items from your on-board tool pouch along with your pocket-knife or Leatherman Wave (you do carry a knife and/or Leatherman, right?) to attempt this fix roadside.

I’m almost glad I shipped the replacement BMW hazard switch to the wrong house, so that I had a chance to learn and attempt this DIY fix. If this ever happens again in my Ferrari 328, I’ll probably use that BMW part and swap it out. But since I also have an F40 and 512TR — both of which use the same hazard switch — this seems like a handy repair procedure to have in ones mechanical bag o’ tricks. If not for your own car, but perhaps so you can help a fellow Ferrari Club member while on a club drive.

As always, I welcome your questions, comments, and feedback below!

The post How to Fix a Broken Ferrari 328 Hazard Switch appeared first on Steve Jenkins.

The audio/video cabinet at the Utah house

To help increase airflow to all that equipment, I’d sometimes open the cabinet doors, but then I’d be stuck staring at the equipment, its lights, and a massive bundle of cables (even though I use zip-ties to try and tidy things up). But that’s not a good long-term solution, so I decided to install a 6″ AC Infinity AIRPLATE S3 fan.

AC Infinity S3 fan

The S3 fan has a street price of under $25, and uses a 6″ x 6″ fan. The entire unit is 6.3″ x 6.3″ with the brushed aluminum panel, which looks just as good in person as it does in the unit’s glamor shots, and would be at home in even the fanciest of cabinets. The fan moves 52 cubic feet of air per minute, with a noise rating of 23 dBa. Dual ball bearings help the fan run so quietly.

The unit came with clearly written instructions (surprisingly in grammatically perfect English), a plastic template to help visualize and cut the necessary hole for the fan, a USB power cable, and a separate USB power adapter. You can power the fan with the adapter, or by plugging it into any USB outlet of any device in your cabinet that delivers least 500mA of power (which will be pretty much any USB outlet). The newest version of this fan (shown above) has a multi-position power switch to choose fan speed, but I have one of the earlier units with a simple on/off switch. The unit also has a USB pass-thru cable, meaning you can daisy-chain up to four fans from a single power supply. Whoever designed this fan really thought things through, and it shows.

Using the provided plastic template, I marked the spot on top of my cabinet where I wanted to install the fan. I had previously removed all the equipment from the cabinet, as I planned to use this project as an excuse to tidy up all the cabling. But if you plan on leaving all the equipment in place, I recommend putting a towel on top of the equipment to protect it from falling sawdust.

Using a hole saw attachment on my drill, I cut a hole in the middle of the marked square, then cut a few more holes inside the markings, so I’d have less hand-cutting with the mini-hacksaw. Had I been in Seattle, I’d have used my reciprocating saw, but a drill is the only power tool I keep at the Utah house, so I did my best cleaning up the hole saw cuts with a mini-hacksaw. Using the fan to check my work, I finally had a hole large enough to flush-mount the fan:

Fan ready to be placed in the newly cut hole

Note that because the flanges on the fan are wide, having perfect edges on your cabinet isn’t necessary. Once the fan drops into the hole, everything looks neat and tidy… though I accidentally scratched the cabinet with the hole saw near the lower right corner of the fan:

Fan flush mounted into the top of the cabinet

The view from inside the cabinet shows my sloppy cuts, but this is a view nobody else will see:

Bottom view of fan

The plastic template also has markings for screw holes, but I prefer to drill them using the installed fan as a template. I used a 5/32″ bit to drill four holes, then dropped in the supplied machine screws. I also used a wood stain pen to touch up the scratch I’d made earlier:

Top view with machine screws installed

I also decided that while I had the hole saw handy, I’d create another hole in the cabinet to feed the cable for my Amazon Echo Dot… and future-proof for anything else I might want on top of the cabinet later:

Fan installed and additional hole cut

Using the provided nuts, I secured the machine screws from underneath. Tightening the fan into place should help eliminate vibration and noise:

Bottom view with machine screws and nuts installed

I plugged the USB power cord into the provided power adapter, plugged it into the battery backup unit I keep in the cabinet, then hit the on switch. I had to look at the fan to make sure it was running. It’s quiet… and I mean really quiet. I held my hand above the top of the fan, and could feel air flowing out of the cabinet (it’s better to “pull” hot air from a cabinet than it is to “push” cooler air in).

I replaced the decor items on top of the cabinet, making sure not to block the fan. The fan is so quiet, and so completely hidden by the decor, that you’d never know it’s there. The gap in the cabinet doors (you can see blue light from the DirecTV DVR through it) is wide enough to allow cool air to flow in as the fan pushes hot air out:

With everything put back in place, you can’t even see the fan

Had I wanted to get really fancy, I’d have purchased the same fan with the digital thermostat control unit, which could turn the fan on and off based on the temperature inside the cabinet. It’s about $44 on Amazon (compared to $25 for the fan alone):

AC Infinity fan with thermostat controller

But I was going for a “stealth” install on this project, so I might use the fancy controller on a different future project. It can control up to four fans.

AC Infinity also has smaller fan units, as well as units with two, three, and four fans, so you can probably find one that’s right for your particular project.

Bottom line? If you’ve got your electronics hiding inside a hot & stuffy cabinet, you can dramatically extend their service life by circulating some cooler air around them with an AC Infinity fan. Like a number of other items I’ve reviewed, this item falls in the “cheap insurance” category, and can easily pay for itself by preventing the failure of your expensive A/V or computer equipment. I’m a fan of this fan.

As always, I welcome your questions, comments, and feedback below.

The post Review: AC Infinity Airplate Cabinet Fan appeared first on Steve Jenkins.

A standard GFCI outlet

GFCI stands for Ground Fault Circuit Interrupter, and that’s precisely what it does: it stops the flow of electricity to the outlet by opening (interrupting) the circuit if it senses any abnormal current flow to the circuit’s ground, thereby preventing a possible electric shock to anyone using a device that’s plugged into the GFCI outlet.

You probably have a GFCI outlet in your kitchen, as well as in each bathroom in your house, as they’ve been required in potentially damp locations for many years. And if you have any outlets outside your house, it’s been a US requirement since 1973 that they be connected to a GFCI-protected circuit.

You’ve also probably had to “reset” at least one of your GFCI outlets at some point by pressing its RESET button (shown in red in the above photo) after it was tripped by a power surge, storm, or an overloaded circuit. But other than when you discover something’s not working and have to reset it, you probably don’t give your home’s GFCI outlets a second thought… which could be a dangerous mistake.

GFCI Outlet Service Life

GFCI outlets save thousands of lives each year by preventing electrical shocks due to dampness, improper use of extension cords, etc. But depending on the vintage of your home and/or its GFCI outlets, they may no longer be protecting you like you believe they are. GFCI outlets have an average service life of only 10 years, so if your home is older than that, there’s a good chance your GFCI outlets aren’t fully protecting you. And if you live in an area more prone to storms or power surges, it’s not uncommon for a GFCI outlet to wear out in 5 years or even less.

Testing your GFCI Outlet

The best way to see if your GFCI outlet is still doing its job is to test it. Keep in mind that the outlet must be powered (meaning its breaker in your electrical panel must be on) in order to be properly tested. An unpowered GFCI outlet can’t be tested, and it won’t allow you to reset it until it’s powered.

The simplest way to test your GFCI outlet is to press its TEST button, which interrupts the flow of electricity and pops out the spring-loaded RESET button. You can then press the RESET button reconnect the circuit and resume normal operation. The most common GFCI outlets have a red TEST button and a black RESET button as shown in the first photo, though newer outlets have gone away from those colors and the buttons tend to not stand out as much.

If you press the TEST button and the GFCI outlet doesn’t trip (and you’re certain it’s powered), replace it as soon as possible. If your test causes the outlet to trip but the RESET button won’t reset it, that means it’s reached the end of its service life and needs to be replaced (it probably wasn’t protecting you anyway). Again, verify that the breaker isn’t tripped before determining that the GFCI outlet is faulty, since a GFCI outlet can’t be reset if it’s not getting “upstream” power.

As crazy as it might sound, you should test your GFCI outlets monthly. If that’s too often for you, you should test them at least quarterly at a bare minimum. Every six months or once a year is nowhere near often enough. I know it’s extra work, but it really could save your life.

Using a Test Device

While pressing its test button is better than nothing, the only true way to test a GFCI outlet is with a test device that actually creates a ground fault, such as this Triplett Plug-Bug:

A basic GFCI outlet tester can literally save your life

At under $7 on Amazon, it’s extremely cheap insurance to protect you and your family from hazardous electrical shocks. Just plug it in the outlet, press the button on the tester (not the outlet), and see whether your GFCI outlet trips. Indicator lights on testers like this one can also be used to show whether any outlet (it works on both GFCI and standard) is wired properly by identifying common problems such as an open ground, open neutral, open hot, or reversed wires. I recently used my tester on a kitchen GFCI outlet at our cabin, and discovered that it had been improperly wired for many years. The outlet would trip when I tested it with its own test button, but additional “downstream” outlets were still receiving current. I was able to re-wire the outlet properly, and I now use a tester to check my GFCI outlets instead of pressing the outlet’s test button. Again, for under $7, I recommend you do the same.

Replacement GFCI Outlets

Because they’re a wear item that will need replacement, I save money by buying GFCI outlets in 3-packs. I use (and recommend) these 15A GFCI outlets by Eaton. Eaton has a great reputation for quality, but the feature I really like is that these GFCI outlets run a self-test periodically. A blinking red LED indicator means they’ve reached the end of their service life and need to be replaced. So if you don’t think you’ll remember to manually test your GFCI outlets monthly, using these self-testing versions are a good option. They’re available in a number of colors to match your existing outlets and switches.

Safety First

Testing your own GFCI outlets is definitely something any homeowner can (and should) do. Replacing a GFCI outlet that has reached the end of its service live is something most homeowners can do… but it’s important to follow basic electrical safety procedures such as shutting off the circuit at the breaker and testing all wires in an outlet box before touching anything. This is particularly important in multi-gang boxes (boxes that contain more than one outlet and/or switch). Never assume that a wire is dead just because the one next to it is (ask me how I know). Always use an electrical tester before touching anything. If you don’t have one, you can pick up a circuit tester in a set along with a GFCI outlet tester for less than $15. Again, it’s cheap insurance that every homeowner should have.

Final Thoughts

It probably comes as a surprise that the GFCI outlets in your home are designed to wear out after a few years (much like your home’s smoke and CO detectors), and that there’s a very good chance that some or all of your GFCIs are way past their intended service life. Using inexpensive tools, you can regularly and safely test your home’s GFCI outlets to identify potential problems before you encounter a serious injury (or worse). If replacing your own GFCI outlets is outside your comfort zone, hire an electrician… or better yet, invite a handy friend over for some barbecue.

Just like testing your expansion tank or flushing your water heater, testing your GFCI outlets is something all homeowners should do on a regular basis… though most don’t even know they’re supposed to. Add a monthly GFCI outlet check to your household maintenance calendar, and sleep better knowing they’re doing their job.

As always, I welcome your questions, comments, and feedback below.

The post You Should Test and Replace your GFCI Outlets More Often Than You Think appeared first on Steve Jenkins.

Pink Power Wheels Cadillac Escalade

Not so fast…

Remember this scene from Princess Bride? “There’s a big difference between mostly dead, and all dead.”

It’s possible that your Power Wheels battery is only mostly dead, which means there’s a chance you could revive it with this trick.

Forget the cheap-o OEM Power Wheels wall-plug charger. I took a pair of wire cutters to ours long ago. It’s a “trickle” charger, not a “float” charger, meaning it delivers a constant trickle of 12 volts at a low amperage to slowly charge the battery. But because the charger has no circuitry to measure the charge level of the battery, it never stops charging… which is why the manufacturer warns you not to charge it for more than 24 hours at a time. But nobody remembers to do that, so we all end up with over-charged (dead) batteries, or batteries that we forget to charge at all… which also end up dead.

A “float” (or “smart”) charger also delivers a low-amperage charge to the battery, but features circuitry that senses when the battery is fully charged. When that happens, it stops charging and only monitors the voltage. When the batteries voltage drops enough, the charger re-activates and fills it up again.

Like I mentioned above, I’d snipped off the “wall wart” adapter portion of our Power Wheels charger long ago and crimped a female spade connector to each lead, giving me something metal to which I could attach charging clamps from my trusty CTEK 7002 smart charger, which kept the 12V Power Wheels battery at the ready without risk of over-charging. But I use my charger for other stuff, too… meaning I sometimes let the Power Wheels battery sit for months without charging. So when I went to try and charge it yesterday, my smart charger sensed too little voltage to even register the existence of a battery on the other end, so the charger lit up its red “ERROR!” light.

That’s the main drawback of “smart” chargers… sometimes they’re too smart for their own good. But because I had those female spade connectors crimped onto my Power Wheels charging plug, I was able to “piggy-back” charge it by wiring it in parallel with a separate 12V battery, like this:

“Piggy-back” charging a mostly-dead Power Wheels 12V battery with another 12V battery wired in parallel

Prior to hooking everything up, I made sure the black 12V Werker battery shown above was fully charged to the point that that my charger lit its green “full” LED.

Then I connected everything as shown in the above photo: the positive spade connector from the Power Wheels plug is connected to the positive terminal of the fully-charged 12V battery, and the negative spade connector is on the full battery’s negative terminal. The “thirsty” Power Wheels battery immediately began drawing a small amount of charge from the fully charged battery, so that the combined voltage of the two batteries was well below 12V on my multi-meter. When I attached my charger clamps to the battery terminals (positive to positive, negative to negative), the charger sensed the circuit’s voltage, and decided it was was low enough to require charging, so it started charging. You can see in the first photo that the charger’s status LED is showing “more than 50% charged.” After a few hours in that state, the light turned green:

Batteries wired in parallel are now fully charged

At this point, I disconnected the black 12V battery from the circuit completely, and connected the charger to only the leads on the Power Wheels charging plug:

Previously, the Power Wheels battery had been so dead that the CTEK charger gave up and turned on the error light. But after being wired in parallel and “piggy-back” charged, the Power Wheels battery had recovered enough to allow the smart charger to “see” it and start charging, so I let it charge like this for another hour or so before I checked it again. Eventually, the green LED indicated that the Power Wheels battery was fully charged. I disconnected the charger and confirmed with my multi-meter: a victorious 13.2 volts!

I put the battery back in the pink Escalade, and it ran like a champ. Because of its age, it’s hard to say whether this battery will still provide runtimes as long as it did when it was new. It probably won’t. But the good news is that it’s not all dead, and it’s not even mostly dead any more. I’ve dodged the $56 new Power Wheels battery bullet for at least another day.

Keep in mind this trick only works with batteries and chargers that use the same voltage — 12V in this case. If you have a 6V Power Wheels battery, you’ll need a 6V charger and another 6V battery to piggy-back. Do NOT try to charge a 6V battery with a 12V charger or “host” battery. Also remember that when you wire two batteries together, it really makes a huge difference whether you wire them in parallel, or in series. If you don’t know the difference or why it’s important, do some web searching on batteries in series vs. parallel before you start tinkering.

The post Dead Power Wheels Battery? Maybe You Can Trick it Back to Life! appeared first on Steve Jenkins.

The Ubiquiti UniFi Controller comes with its own self-signed SSL certificate, which means that every time you try to access the web interface, your browser complains about the security of the self-signed certificate, then makes you jump through hoops to get there (I’m looking at you, Google Chrome). I run a hosted version of the UBNT UniFi Controller on a CentOS Linux web server that already has its own a valid SSL certificate. And because the UniFi Controller’s GUI runs on a different port than the standard HTTPS connection on that web server, I can use the same existing SSL certificate for that hostname for my UniFi Controller. This post will show you a quick and easy way to use your existing SSL certificate (including a Let’s Encrypt SSL certificate) on your Linux-based UniFi Controller using my unifi_ssl_import.sh script.

In order to use an existing SSL certificate with a UniFi Controller, you need the following:

• A valid SSL certificate (.crt)
• The private file (.key) generated with the SSL certificate’s signing request
• The complete certificate authority chain from your signer (usually a .pem or .crt file). This is the part that normally causes the most frustration when trying to import an SSL into a UniFi Controller, but I’ll explain and walk you through it below.

If you Don’t Already Have an SSL Certificate

If you don’t already have a valid SSL certificate… go get one. There are plenty of free options, so there’s no excuse not to have one. In fact, I’ve written this easy-to-follow guide that walks you through all the steps necessary to generate the the private key and download a free one-year SSL certificate from StartSSL to use with your UniFi Controller (renewals are free, too). Another popular option is Let’s Encrypt, which I’ll address separately below.

Using Let’s Encrypt SSL Certificates with UniFi Controller

A recent popular option for free SSL certificates is Let’s Encrypt. Compared with traditional SSL certificates, they are easier to retrieve, but they expire every 90 days — meaning you’ll need to renew them before they expire. Let’s Encrypt recommends using the certbot client (and so do I) to automate the renewal process. You can run a twice-daily certbot cron job that checks the expiration date on your SSL certificate, and if it expires within 30 days, it will renew it automatically.

I should inform you that there are scripts out there that combine the functions of checking, renewing, and importing a Let’s Encrypt SSL certificate into the UniFi Controller — but I don’t like any of them, which is why I added Let’s Encrypt support to my dedicated UniFi SSL import script. The “all-in-one” scripts are never as flexible, and they usually force you into a single method of certificate renewal, which may not work for your server. I wrote my script to do only one thing, and do it well. Because the recommended certbot client is so easy to configure and use, I recommend setting up two cron jobs that run twice per day: one that uses certbot (or any of the the other Let’s Encrypt clients) to check for and download a renewed Let’s Encrypt certificate when needed, and a separate cron job that runs my unifi_ssl_import.sh script. The script is lightweight and smart enough not to do anything unless the downloaded Let’s Encrypt SSL certificate is newer than the one currently in the UniFi Controller.

The UniFi Controller Keystore File

SSL certificate information for the UniFi Controller’s web GUI is stored in a file named keystore. On Ubuntu and other Debian-based systems, the file is located in /var/lib/unifi/data/keystore. On CentOS, Fedora, and other non-Debian systems, it’s stored in /opt/UniFi/data/keystore.

The script’s method for using your own SSL certificate with a UniFi Controller involves modifying the keystore file, so it’s extremely important to make a backup of your original keystore before doing anything. My method doesn’t touch any other files associated with your UniFi Controller’s installation, so if your UniFi Controller stops working after using my script, you can simply restore your backup keystore file to its original location, restart the UniFi Controller, and you’ll be back online.

What the unifi_ssl_import.sh Script Does

The script is heavily commented, so you can read through it to see exactly what it does. But in basic terms, the script:

1. Makes sure it has everything it needs to attempt an import.
2. Backs up your keystore file.
3. Uses the openssl command to export data from your private key, SSL certificate, and SSL certificate chain to a temporary PKCS12 file, along with a password that the UniFi Controller is expecting.
4. Uses the keytool command to import the temporary PKCS12 file into the keystore file.
5. Uses the UniFi Controller’s java-based import_cert function to update the controller’s SSL certificate chain.
6. Restarts the UniFi Controller to apply the changes.

Building an SSL Certificate Chain

An SSL certificate chain is a list of certificates that ensures a trusted relationship all the way from the “root” certificate of the signing authority, through any “intermediate” certificates from other signing authorities, and eventually to the “end user” certificate on a web server. To see how they connect, take a look at Ubiquiti’s certificate chain from ubnt.com:

UBNT.com’s SSL Certificate Chain

The root certificate (there’s always only one) belongs to Starfield Services. There are two intermediate certificates belong to Amazon (Amazon Root CA 1 and Amazon), and the client certificate (also always only one) belongs to ubnt.com. All four certificates make up the SSL certificate chain — indicating that Amazon trusts ubnt.com (client), Amazon trusts Amazon (intermediate), Starfield trusts Amazon (intermediate), and Starfield is a a root level authority that deserves to be trusted by everybody.

The UniFi Controller won’t let you connect to the GUI unless a chain of trust can be validated all the way to the root certificate. To ensure that the UniFi Controller can display the “green lock” when you connect to it, it needs to know about and trust every certificate in the certificate chain. The easiest way to ensure that is to download a copy each of the certificate authority files (also referred to as CA files) in the chain. Don’t worry, they’re easy to get. And once you have them, you can easily combine them into a single chain file, which you can then import into the UniFi Controller along with your client certificate, so it knows about the entire chain.

If you’re using Let’s Encrypt, the good news is that it automatically downloads a chain file named chain.pem file into the same directory as your private key and signed certificate. The bad news is that because Let’s Encrypt is relatively new, one of the certificates referenced in that chain.pem is cross-signed by an IdenTrust root (this IdenTrust root, to be exact) so that IdentTrust cross-signed certificate needs to be downloaded and imported separately into the UniFi Controller’s keystore. But the other good news is that the unifi_ssh_import.sh script handles all that for you, so Let’s Encrypt users don’t need to worry about chain files at all.

If you’re using a SSL certificate from another provider like StartSSL, you can build your own certificate chain by downloading and combining the root and intermediate CA files together into a single chain file that the unifi_ssh_import.sh script can import. All certificate authorities are required to make their certificates publicly available, so no matter which signing authority you use, their CA files are easy to find with a web search. StartSSL’s CA files are are located here. If you followed the instructions in my Create a Free SSL Certificate with StartSSL article, then you’ll have a root_bundle.crt file located in your server’s certificates folder (we’ll assume it’s /etc/ssl/certs) which you unzipped with your signed certificate. Combine that root_bundle.crt CA file with StartSSL’s Root 1 CA file and you’ll have the whole chain. On your server, do:

# cd /etc/ssl/certs
# wget https://www.startssl.com/certs/ca.crt

That will download the Root 1 CA file into your certificates folder. Now combine it with your root_bundle.crt CA file into a chain file with:

# cat ca.crt root_bundle.crt > startssl-chain.crt

Use the full path to that startssl-chain.crt file as the CHAIN_FILE configuration option in the unifi_ssh_import.sh script, and you’ll be all set!

Using the unifi_ssh_import.sh Script

First, download my unifi_ssl_import.sh script to your Linux box (I recommend putting it in /usr/local/bin) and make it executable with chmod a + x. You’ll need to edit few configuration options before you run it for the first time:

• UNIFI_HOSTNAME: the full name of the server where your UniFi controller runs, such as unifi.example.com
• UNIFI_DIR: on Fedora, CentOS, and other RedHat systems, this will be /opt/UniFi. On Ubuntu and other Debian-based systems, this should be /usr/lib/unifi. Both options are in the file, so just uncomment the one you need and comment the other one out.
• UNIFI_SERVICE_NAME: the name of the sysvinit or systemd service that runs the controller. It’s the word you type between “service” and “start” or “stop” from your command line to start and stop the controller service. If you’re using one of my UniFi Controller Linux start-up scripts (try one… you might like it!), this should be UniFi. If you’re using someone else’s script (sniff, sniff), it might be unifi.
• LE_MODE: If you’re using Let’s Encrypt, then set this to yes, which will user the UNIFI_HOSTNAME to automatically locate the certs.pem, privkey.pem, and  the script find all the right file paths and automatically build a complete chain file that the UniFi controller won’t complain about.
• LE_LIVE_DIR: This is set to the Let’s Encrypt default of /etc/letsencrypt/live, so you probably won’t need to change it, but you can if you need to.
• PRIV_KEY: the full path to your private key (default is /etc/ssl/private/hostname.example.com.key). Set this to the correct name and location of your key. This option is ignored if LE_MODE is enabled.
• SIGNED_CRT: the full path to your signed SSL certificate (default is /etc/ssl/certs/hostname.example.com.crt). Set this to the correct name and location of your cert. This option is ignored if LE_MODE is enabled.
• CHAIN_FILE: the full path to the CA chain file you created above (default is /etc/ssl/certs/startssl-chain.crt). Set this to the correct name and location of your chain file. This option is ignored if LE_MODE is enabled.

There are a few other configuration options such as ALIAS, KEYSTORE, and PASSWORD, but I don’t recommend changing them, as these are set to what the UniFi Controller is expecting, and changing them will almost certainly cause something to barf.

If you’ve already got a Let’s Encrypt certificate on your server, simply set the UNIFI_HOSTNAME, verify the correct UNIFI_DIR for your system, set LE_MODE to yes, and run the script.

For any type of user, if the script is unable to find any of the required input files, it will warn you and exit. Fix the problem and run it again.

The first time the script runs, it will backup your original keystore file as keystore.orig. If you’ve never modified this file, it still contains the UBNT self-signed certificate. If you run the script a second time (or multiple times, as with a cron job) it recognizes the existence of keystore.orig, assumes that means the current keystore file has been modified, and backs up the modified keystore file as keystore.bak (instead of overwriting the original backup). That way, if things ever get really hosed, you can copy keystore.orig to keystore, restart the controller, and regain access to the GUI.

If you’re using a standard SSL certificate provider like StartSSL, GoDaddy, Comodo, DigiCert, Verisign, etc., you’ll only need to run unifi_ssl_import.sh every 1-6 years, depending on when your certificate expires and you have to install a new one. So it’s probably pointless to set it up as a cron job.

If you’re using Let’s Encrypt, however, your certificate will expire every 90 days. Let’s Encrypt recommends running a twice-daily certbot cron job to check the status of your certificate (which can also renew it when necessary). I recommend that Let’s Encrypt users run a cron job that also executes unifi_ssl_import.sh twice daily, about 2-3 minutes after the certbot job. If the certificate is new, it will get automatically imported into the UniFi Controller. If not, the script will exit quietly.

The script is unifi_ssl_import.sh, and is part of my UniFi Linux Utils GitHub repo here:

Once you run the script, give the controller a few moments to restart (don’t panic if you can’t connect for up to 30 seconds), then open the URL of your UniFi Controller in a web browser (UBNT currently seems to favor Chrome). If the warnings don’t appear and you see a happy green https://, you’ve successfully imported a signed SSL certificate into your UniFi Controller!

Final Thoughts

Hopefully, this article and script will help you make short work of the often frustrating task of using a signed SSL certificate with the Ubiquiti UniFi Controller. Your comments, suggestions, and feedback are welcome in the comments below, but I don’t offer technical support on my blog. If you’re having trouble with this script, please post in this thread in the UniFi Wireless section of the UBNT Forums:

https://community.ubnt.com/t5/UniFi-Wireless/Automated-Script-and-Walk-Thru-for-Existing-SSL-Certificates-on/td-p/1603237

I monitor that thread, and am happy to provide help there.

The post Use an Existing SSL Certificate on a Linux UniFi Controller appeared first on Steve Jenkins.