Attoparsec

Skinner Box

2011-09-29T23:49:00.000-07:00

My big new project for Burning Man this year was the Skinner Box, originally brainstormed between me and my dad two years ago on the way back from the 2009 Burn. It is, simply put, a very cruel joke: an exercise bike in the middle of the desert which randomly dispenses hard alcohol. Before taking it down I made this demo video:

Not a very complicated project, but it was still a frantic couple of weeks getting it finished in time. I used a cute little 6 volt submersible food-grade pump which was being sold on Amazon for sous-vide cooking purposes. The bike was $3 from Goodwill, and I rigged up a simple tachometer using a $5 used bike computer. The non-contact magnetic sensor was just a reed switch, so it was trivial to detect it closing when hooked to an Arduino.

Photo by Jen Kale'a

It had flashing lights, a progress needle hooked to a servo, and a buzzer playing happy and sad noises to tap as many addictive gamer impulses as possible. Plus, of course, random positive feedback. The progress needle has to increment 15 times in order for a unit of alcohol to be dispensed. As long as you're pedaling, it will keep incrementing about every other second. Except that each time it does, there is a 4% chance of failure, which sends it back to the beginning. 0.96¹⁵ = 0.54, so you have slightly better than even chances of success each time you start. Each win provided about a third of a shot of the cheapest vodka I could find. Warm, cheap, slightly dusty vodka.

Photo by Jen Kale'a

It was fairly well received on the playa. We were camped farther out on 5:00 than we'd been hoping for, though, so there wasn't a huge amount of foot traffic. It mostly just confused people, I think. Most of the people who actually used it were the ones who knew what a Skinner Box was and got the joke from that level. I did catch one or two people engaging in classical Skinner Box behavior, though, developing complex theories as to when it would and would not dispense alcohol. We didn't go through nearly as much of the cheap vodka as I had planned for, so I spent some time on Saturday going around giving the surplus out to needful bars.

The el-wire sign I made for it turned out to be the best thing about it. It was a very convenient for finding camp at night. I was very pleased with its crisp, clean lines. I probably should only have used two colors of wire, though. I was going for a classic 50s neon look and I think the color scheme just looks kind of... toothpasty. Oh well.

Photo by Jen Kale'a

In the end, it was a fun project, but I won't be taking it back. I don't think I like doing static installations like this in front of camp. It encouraged me to just sit there waiting for people to come by. I want to be out and about, wearing and/or riding my art. I do have some other installation ideas, but they will be 1) bigger 2) more conceptually simple and 3) on fire. As they should be.

Kalamazoo: Out and About

2011-09-27T22:49:00.000-07:00

It's been a busy month or so for the Kalamazoo, which is among the reasons why I haven't made a post about any of it until now. While I hope to take it out to events in the future, the project itself is now done as far as I'm concerned. Any further improvements would require radical changes to the drive train, and that just isn't going to happen. Too many other things to work on!

Scientific Expedition and Picnic
In early August a group of friends and I dressed up in our silly steampunk best and drove the Kalamazoo down to a nearby park for a picnic. We sipped lemonade from a mismatch of thrift store tea cups, nibbled on cucumber sandwiches, and generally had a grand time.

Many pictures were taken...

Photo by Gabriel Cain

In stereo, even...

Photo by Espressobuzz

And I've since taken some liberties with them:

A large amount of footage was also shot. Having recently been turned on the music of Caravan Palace, I couldn't resist making a full music video for their song Dragons. I think it nicely captures the feeling of the expedition:

ThingOut!
A week later, Ignite Seattle teamed up with Dorkbot to put together ThingOut!, a showcase of local Maker projects. We took the Kalamazoo down, which turned out to be a really good choice. It was awesome! It turns out that kids love the thing. There were 3-6 kids climbing all over it pretty much the entire time we were there. They tended to pick up the mechanical skill of driving it a lot faster than adults, too. Maybe because they don't have other driving experience to unlearn first? The whole thing left me very eager to take the device to other events, despite the logistical hassle of having to rent a trailer every time.

Burning Man 2011
And a week after that I left for the playa, taking the Kalamazoo down for the second time. This made me a bit nervous, if only because last year had been such a big failure. But despite it still be very, very slow, it worked much better this year. Hooray!

On Wednesday night I took it all the around Esplanade. That was about 3.6 miles and it took over 7 hours (with many rest breaks, but still). My arms nearly fell off! But I did it. The Kalamazoo finally proved itself under full playa conditions. The addition of the trunk made an expedition of this length quite easy to pack for. We had 2 gallons of water, food, tools and bike pump for an emergency, and spare lamp oil for the lanterns. Very civilized

Photo by Jen Kale'a

People would occasionally jump on for a ride or to help pump, but we were going so slowly that was fairly rare. Got a lot of good comments, and some annoying/stupid ones. But the people who liked it really liked it. One person, upon finding us slowly creaking along out by the Temple, said we had made his Burn. Which, in turn, totally made mine.

Photo by Jen Kale'a

Flywheel weight casting

2011-07-17T23:34:00.000-07:00

After finishing the new flywheel, I couldn't leave the lead weights so clunky and ungainly. So I set about casting them with nice conformal curves to match the circle of the octagonal flywheel.

I needed a shape to make the mold around, so I took a block of beeswax I had laying around and started working it into shape. First I roughly shaped it on the flywheel itself, to get the size right. The curve was compared against a paper template I made up using a yardstick compass.

Then I heated up a spare section of angle iron with my blow torch, and used that to smooth down the sides. With the blank ready, I printed out an "ATTOPARSEC" label and transferred it across, carving the letters into the wax. Might as well keep up my branding, right?

This was affixed (using a bit of melted wax) to the bottom of a small cardboard box I made, and RTV silicone was poured over it. Silicone can take the heat of lead casting fairly well, but it isn't cheap. If I did this again, I'd probably play with sand casting first.

With the positive removed, the silicone mold shows how well it picked up all the details.

While fairly low temperature, lead casting should be approached with caution. Like all metal casting, it shouldn't be done over concrete. Because lead is, of course, toxic, only do it outside in a well ventilated space. Don't breath the fumes if you can help it! Wash all your clothes and your self immediately after finishing. But other than those worries, it's a pretty easy process. A little propane burner is more than enough to melt lead, and any steel can with fairly thick wall can be used as a pot. I drilled a #19 (0.166") hole for a pouring spout, which also helps strain out the dross. Later I drilled it out to an H (0.266") to get it to flow better.

Wearing full safety gear (all natural fibers, long pants and thick coat, leather boots, safety glasses and full face shield, filter mask, casting gloves, hair tucked away) I grabbed the pot using a pair of blacksmithing tongs and poured in the molten lead.

It came out a bit chunky, and not very fast. It was a bit too cool and the hole wasn't big enough, but the result wasn't too bad.

The mold suffered a bit of degradation, but that was worst on the first cast. I used it a total of 6 times, and it could certainly handle a couple dozen more still.

The end result is pretty good, I think. I'm not entirely sure why the first one was such a different color. It was poured at a lower temperature, I think. The exact alloy would have been a bit different, as the later ones had some lead from a scuba diving weight mixed in.

I've only attached 4 of the weights so far, giving the flywheel roughly the same mass as before, though they're slightly more optimally placed. It works quite well and looks very slick. It's a real flywheel finally, not some weird kludge.

This is probably the last major change to the Kalamazoo. It's been fun, but it has already gone on a year longer than I originally planned. There are too many other crazy things waiting to be built!

Flywheel mania

2011-07-11T00:09:00.000-07:00

The most persistent problem with the Kalamazoo has been the temperamental drive. Every stroke of the pump lever would result in a terrible clank, and it was very hard to keep the thing moving. You had to time your pumps exactly right, while not applying any force at the top and bottom of the stroke. That is why one of the first things I did this year was cut the decking to allow for much longer arms on the simple flywheel hack. That helped, but a greater flywheel effects was needed.

This is what I came up with, a welded hexagonal ring. It's made out of two sections which bolt together, which allows it to be mounted without having to remove the crankshaft. Lead weights are then bolted to the ring to increase its rotational inertia.

About a week after finishing it, we finally got around to running a full roadtest using it, with very positive results.

The Kalamazoo is finally handling as I had always imagined, though it's still very slow. Carrying passengers worked well and made the motion even nicer. We've found that having a secondary driver in the forward position doesn't help very much. Synchronizing pumping between people turns out to be a difficult thing. I'm thinking of adding an unpowered dummy in the front position, but I probably won't get to that.

I'm now working on casting the lead weights into a more conformal shape which will allow them to be mounted on the outside of the wheel. This should be both more efficient and more aesthetically pleasing.

Kalamazoo: Flywheel and steering

2011-06-17T07:56:00.000-07:00

First Kalamazoo update of 2011! This covers the recent upgrades to the flywheel and steering mechanism.

Also, after some last minute failures, the Kalamazoo will not be appearing in the Fremont Solstice Parade, which is disappointing. But a lot of good progress has been made, so I'm confident I'll have it working properly in time for Burning Man this year.

EL Wire Interface Kit

2011-06-08T21:25:00.000-07:00

As I've vaguely mentioned in a couple of posts, I've been working on productizing the the EL wire interface cards I used in the Lightsuit. This involved getting a custom circuit board printed, which was a first for me. Designing some simple packaging and documentation was also fun. It isn't anything fancy, but it is entirely functional.

The kit was finished just in time for the Kitsap Mini Maker Faire last week. That makes it's past time to make the official announcement online: the Attoparsec Eight Channel EL Wire Interface Kit is now available!

What can this kit do for you? Control up to 8 strands of EL wire from a microcontroller. Why would you want to do that? Because it's awesome -- and can be lot more interesting than just having it blink.

Setting up that little demo took a couple of hours. For an example of what can be done with a little bit more effort...

The kit is selling for $20 in person, $22.50 online with shipping included. Plenty of time left before That Event In The Desert to do something really cool. Just saying.

Kitsap Mini Maker Faire -- The Day After

2011-06-06T21:59:00.000-07:00

I spent yesterday at the Kitsap Mini Maker Faire showing off Attoparsec projects and generally having a delightful time. It was small -- but very well attended. We took the Lightsuit, the new EL wire interface kits, the etched brass fan, the Fireprop, and the full-scale hip joint from the hexapod project. I also took the RC hexapod model, thinking it would be a hit with kids. I was very, very right! I felt bad showing off something I just made from a kit, but my excuse was that it made a nice explanation for the large hip joint prototype.

I was talking pretty much the entire 6 hours of the Faire. It was exhausting, but also a lot of fun. We also sold 3 kits! (Which I still haven't posted about properly, I know.) Our display was pretty random, so I'm glad people seemed to really get into it all the same. I think I have a better idea what makes a good booth presentation now, which will be useful when we go down to the big SF Maker Faire in force next year. But if there is a second year for the Kitsap Mini Maker Faire, I'll definitely go again.

TARDIS hat

2011-05-31T23:27:00.000-07:00

Lauren was making a TARDIS dress. It obviously wouldn't be complete without a hat that lit up and made VWORP VWORP noises. If that doesn't sound like a job for Attoparsec, then I haven't been branding hard enough!

The idea was pretty simple. Laser cut a lightweight plywood frame. Paint blue and insert a plastic diffuser. Add LEDs. Connect to a control package which can make them fade in and out while playing an audio file. Add a switch so that it can be easily and subtly triggered as part of a costume. So that's what we did.

First we prototyped in cardboard, just to get the overall size and aspect ratio down. Using that, I drew up the design in Inkscape. A quick stop by the local makerspace and it was done.

Along with some 1/8" dowel and a fancy little finial knob, that was the frame of the hat. Stained it blue, added the LEDs, sewed on a combination head band and elastic comb pair to keep it tightly (but not uncomfortably!) in place, and it was done.

The electronics were pretty simple. An Arduino with WaveShield to control the lights and play the music.

I used an old Radio Shack mini amplifier to play the sound file. All this was hidden in the dress, with just a thin section of 2-lead ribbon cable leading up to the hat itself. Sharpied black and tucked into the hair, it was invisible to all but the closest of inspections.

The end result was quite pleasing, if I do say so myself. Have I mentioned I love laser cutting as a construction technique? I really love it.

Here's a quick demo video of the hardware:

Some friends also put together a ... slighty more fanciful demonstration video.

Kitsap Mini Maker Faire

2011-05-31T00:27:00.000-07:00

Attoparsec will be a last minute addition to the Kitsap Mini Maker Faire on June 5 in Poulsbo, WA. We will be displaying various projects, giving introduction tutorials on the use of jeweler's saws, and hopefully be selling the new EL wire interface circuit kits. Be there and be square!

April Tools Results

2011-05-04T21:20:00.000-07:00

Last Saturday we took part in the April Tools Wooden Boat Challenge. Basically, we were given an unknown set of materials, 1 hour to plan and 3 hours in which to build a boat. Hand tools only (except for driving screws). It was a lot of fun!

This is what we had to work with. 1 and 1/8 sheets of plywood. 8 1x2s. 1 2x2. A roll of duct tape. 3 tubes of caulk. Some wood screws. And a meter of rope.

We got a nice curve for the hull by draping a string and tracing it out. This let us precisely determine the length of the curve ahead of time, very important when working with limited materials!

Then the sawing started!

Predictably (if you know any of the team) we brought far too many tools. Better too many than too few!

The side walls were screwed to cleats screwed to the bottom plate. A single thwart helped brace them during construction. We had a lot of problems with the cleats splitting until we started pre-drilling the holes using the hand brace. Of course, its chuck didn't go small enough for the tiny drill we were using. Luckily, we had a pin-vise on hand! And that's exactly why we brought far too many tools with us.

Getting the ends bent into place took a lot of effort. The final joints were okay, but we made heavy use of the caulk to get it actually water-tight.

Lots and lots of caulking.

It was only at this point that I could test if I would actually fit. I was chosen as the primary pilot, as I have by far the most small boat experience thanks to my 1500 km trip down the Mackenzie in 2007. Not that I've done any since then...

At the last minute, we decided it needed a dragon's head prow. At the last second, we decided it needed an eye hole bored out using an antique ship auger we had with us. We finished exactly as the tools-down buzzer sounded.

Our mighty craft, the Knorr! It wasn't the most elegant build, but we were all pretty happy with it. We accomplished exactly what we had planned 3 hours earlier.

As it turns out, though, what we planned had some serious problems. Namely, it had absolutely no primary stability, as the following video plainly demonstrates.

So, we didn't get very far in the race. But oh well! It was fun, we got to spend the day outdoors on a very lovely day on the very lovely Sunshine Coast. If we do it again next year, I expect we'll make a much better showing.

April Tools

2011-04-12T20:51:00.000-07:00

As I've posted elsewhere, I'll be taking part in this years April Tools competition on April 30 in Madeira Park, BC. Basically, we're given a pile of unknown materials and 3 hours to build a small boat, which we then race around the harbo(u)r.

I've wanted to do this for years, and I'm getting quite excited about it. My team is registered under the very appropriate name The Archimedes Screwballs, should you be in the area (ha!) and want to stop by to cheer. As you almost certainly will not be in the area, I'll be posting pictures afterwards. And very possibly a video of one of us (me, I'm thinking) getting quite wet.

Lightbox

2011-04-06T20:12:00.000-07:00

(This was actually a December project, but it was only recently permanently mounted, so you're hearing about it now.)

So, Lauren was living in a basement room. No windows. This made getting up in the morning quite difficult, as it was always pitch black in there. As I have a predilection for elaborate technological solutions, I suggested we should make a lightbox disguised as a window, lit with daylight bulbs. I've always enjoyed how psychologically powerful the trick of daylight-colored light coming through fake windows is. So that's what we did.

We found a nice antique stained glass window at an architectural salvage store. Originally I planned on adding drapes, but that was before finding such a gorgeous window to work with. It's mounted on a simple wooden box I made, with hinges at the top and a latch at the bottom.

The electronics are hidden inside. I probably should have gone with 4 banks of tubes, but even the really cheap units were a bit pricey. The box was finished with some ebony stain, which nicely matched the weathered, mildewed window frame.

I'm quite happy with the result, for such a simple build. It's hooked up to a power strip with a built-in programmable timer, so it turns on 15 minutes before the alarm goes off.

One caveat: It was recently discovered that napping in the evening with the lightbox on can really, really mess up your Circadian rhythms. You have been warned.

Full-scale hydraulics

2011-04-05T22:24:00.000-07:00

As I've mentioned elsewhere, I'm giving up on the hexapod as a project for this year. It's just too complicated and expensive to pull off in that time. Instead I'm going to spend the summer working on the Kalamazoo, hopefully getting it into much better shape for the Burn this year.

That doesn't mean I've stopped working on the hexapod, however, just that I've changed the schedule. Prototyping of the hip joint continues with decent, if slow progress. Most recently I've switched out the teeny tiny hydraulic test platform for something a little bit beefier.

It was sold for powering dump truck bed lifts, with basically no documentation. I'm not sure exactly how many amps it wants, but it's a lot. It works well enough, though. Should be capable of about 1.3 gallons per minute, which is conveniently close to 1/6 the number I've calculated as being required for the entire system. So if a single leg can move at speed being run off this pump, that will be very promising.

The hip joint is really the key to the entire project, as far as I can see. The mechanical design is the trickiest, as unlike the other two leg joints the forces are all parallel to the pivot, making it want to bind. And this is where the top speed of the final device is determined. During a normal stride, the other two joints will only move through maybe 10 degrees, while the horizontal hip joint will travel through upwards of 90. It needs to be a low friction and fast moving.

When dealing with hydraulics, speed is a matter of flow rates. A pump can only move so much hydraulic fluid at pressure per unit time, and that determines how fast a given cylinder will move. The volume a cylinder needs to actuate is simply the area of the bore times the length of travel. The faster the pump can provide that volume, the faster the cylinder moves. Of course, the smaller the bore is, the less force it will apply. And the shorter the travel, the worse the lever moment will be for pivoting something. No free lunch, here or anywhere. Luckily, however, the force being applied by even a very narrow cylinder is much more than I should be needing!

So for this joint I wanted a narrow cylinder, with as short a travel as possible. After poking around online for awhile, I finally found one that was 1" bore by 4" travel. Tiny, but it should still be big enough to do the job. I don't really know, of course, but that's what empirical testing is for! And last night I started on that, hooking up the new pump to the old control valve and electronics and the new cylinder, mounting it for the first time on the prototype hip joint, and giving it a spin. And I managed to do it without either losing a finger or spraying gallons of hydraulic fluid around the garage!

In short, a very real success. That's a full-scale joint running with full-scale hydraulics at a reasonable speed. (Would result in a hexapod walking speed of 3+ mph, depending on the leg length. Though without the other two cylinders running, it's not a very fair test.) I need to get it running using the Arduino PID controller next, to get a sense of how bad the momentum effects are going to be. Assuming that looks good, it's on to prototyping the rest of the leg. Which probably means switching to cartridge valves, which means making a manifold, all kinds of fun stuff. But there are several smaller projects clamoring for my attention this month, so most of that will have to wait.

Horizontal Hip Joint Prototype

2011-03-14T22:14:00.001-07:00

I've been (slowly) working on the prototype design for the horizontal hip joint. This is by far the trickiest joint of each leg, as the ~500 pounds it needs to support will be torquing the bearing instead of being a clean radial load. It's also the limiting factor in the overall speed of the hexapod: in a normal walking gait, the other two joints will only move very slightly, while the horizontal hip joint will swing through 90 degrees of travel. The smoother the action is here, the faster the top speed is likely to be.

I decided to use car wheel hub bearing assemblies, as they are cheap and robust. After playing with one under very moderate load, I decided that I just didn't like the torque it was being put under. So I went with two, stacked one top of the other. Now all that torque is converted into a much nicer radial load within each bearing. You can get an idea of the arrangement in this rough sketch. (Except I grew the bearing support frames upside down, oh well.)

This is the completed prototype. (The large shelf is only there for testing purposes.) The action is very smooth, and while a bit tedious, making 6 of these would not be a problem. I'm quite happy with it. Unless some major problems are discovered in later testing, I think this design will work nicely. Making just one of them is giving me a greater appreciation of just how big this thing is going to be.

Here's a closeup of one of the bearing assemblies. They're simply bolted to plates which are welded to the support frames. Everything is under compression.

And here it is with about 175 pounds of static load on it, which it handled quite nicely. Once I get a better test stand for it, I'm hoping to do some tests with it loaded closer to the design goal of 500 pounds. I'm running out of large chunks of steel to hang off of it, though.

Hydraulics

2011-02-19T22:15:00.000-08:00

As part of the investigation in the feasibility of the hexapod project, I've spent the last month playing with a small hydraulics test system. The final system will need 18 actuators, all moving simultaneously and quickly, controlled fairly precisely. Is this possible? (Or, more the point, affordable on my budget?) It's an open question.

The first step was the hook up a small electric motor to a small cylinder and control if from an Arduino. This was done with a solenoid valve interfaced through a basic solenoid driver circuit.

Control of the valve is only one side of the equation, however. To move the joint to a specific position, you need to know where it is, so you know when to stop moving it! This requires an encoder. Out of curiosity I tried using a flex sensor -- I don't recommend it. They're super noisy. Much more promising was the AS5040 chip, which is a non-contact magnetic system. A friend had a test board which I could borrow. It worked much, much, much better. With the addition of a PID library, it was all more or less working.

So, I'm sure I can control the hydraulics as needed. At a much higher cost than I would like -- the valves are going to be at least $100 each! The next step is to rigorously spec out the entire system using real components. Specifically, the horizontal hip joint needs to be able to cycle fast enough, using a real cylinder, driven by a hydraulic pump supplying a realistic volume. The big question is determining just how much force will be needed there. I'll be prototyping the joint and performing empirical measurements with full loading over the next few weeks. And once I have that answer, I'll have a pretty solid guess of just how much this is all going to cost -- and thus if I can do it this year or if it will have to stretch into 2012.

Etched brass fan

2011-01-09T20:00:00.000-08:00

As part of vixy's costume for SteamCon this year, she needed a steampunk Asian-style fan. After some brainstorming, we decided to make one out of brass sheet, with a design etched on the front. This wouldn't be radically different from several other projects, though with 28 individual blades it would still be a significant amount of work.

The design itself was also a challenge, as it needed to be properly tiled across the blades. After some consultation on planetary gear systems, vixy drew it up in Photoshop and then cut it up into different layers using a template we worked out. In the end I was given a series of etching masks and set to work.

As previously mentioned, I first attempted the etching using a new (to me) galvanic process. The results were slow and uneven, and I was blowing through fuses on my power supply at an alarming rate. I finally gave up roughly halfway through and bought a large quantity of ferric chloride from VetCo. The rest of the etching was finished in a single night, thanks to the power of parallelism. Each blade was placed in a ziplock bag with a healthy does of ferric chloride and sealed up. All of these were then placed in a large plastic storage box. This contained the inevitable leaks, and also allowed me to gently agitate all of them at once through the simple expedient of holding it in my lap and bouncing my knees while watching a movie.

The results, once cut out, were pretty nice. This was the version shown at SteamCon, in fact, due to simply running out of time. But the etched pattern didn't really pop, and the folding/unfolding action was very stiff. The blades are threaded together with monofilament (see below), this is what lets you open the blades to the correct position. Unfortunately, the monofilament I had on hand turned out to be much thicker than I had realized. Not only did this make the action very clumsy, it also added a lot of fatigue on the monofilament and it started breaking within hours.

To fix the breaking problem, I bought much lighter, 4# monofilament. I also drilled a second line of holes in the blades, to add redundancy. Now it can break once in any place and still work.

To fix the pattern visibility, I needed to add a patina. I had attempted this originally using the same ammonia vapor technique that had worked well in the past. Those had all been on much smaller surface areas, however. Trying to do it on even half of the blades at once was painfully slow. After a pause of several weeks, I returned to the problem and decided to try using salt water, which I had read about as a way to improve the reaction. And it did! I made up a batch of saline (1 cup water, 2 tablespoons salt) and dabbed it onto the blades with a paper towel. These were suspended over an ammonia bath inside the same cheap plastic storage container used earlier. It only took about 30 minutes to get a very good patina on each blade. I had to come out every 15 minutes or so to reapply the salt water. Some blades too a lot longer, but as I was running them through in batches it still only took a couple of hours to get them all done.

The blades were then lightly sanded to remove the patine on the raised (unetched) sections of the design. They were bolted back onto the central pivot, and I faced the task of threading them all together. This was very difficult when threading the single line of thicker monofilament. Now that I was running two lines of much thinner monofilament, it was almost impossible to get the tensioning correct, leading to inappropriate spacing between blades when the fan was opened up. After some frustrating failures, I finally tried clamping the blades in place while working on it. Success! This prevented the tensioning on already threaded sections from coming loose, and it also made the entire this much easier to manage while working on it. It still took about two hours, but the result is quite satisfactory.

The original plans for this fan included wood slats on either end and a spring-loaded release mechanism. Those are now indefinitely on hold, as it has already taken far longer than planned, and the shear mass of that much brass is greater than expected. It's a fairly hefty artifact already, and adding to that seems unwise. Also, I'm very unsure making it spring-loaded is a good idea. The last thing I want to do is to break the monofilament and have to rethread it again. So for now, the etched brass fan is finished. Look for it next time vixy has an excuse to get dressed up!

Helpful holiday tip from your friends at Attoparsec

2010-12-29T12:51:00.001-08:00

Don't leave ferric chloride sitting in an enameled steel container for months on end.

Cause, uh, it eventually will go somewhere else.

Galvanic Etching

2010-11-14T14:27:00.000-08:00

I've been working on a spring-loaded brass fan, which requires a lot of etching to be done. Normally I would do it chemically with ferric chloride, but that would get very expensive very quickly. So I decided to finally take the plunge and try out galvanic etching, generally following the guidelines here.

It has worked well enough, so far, but it's a bit of a messy process. With a bit under an inch between the work piece and the cathode, it takes 30-45 minutes to get a decent etch. Given that I have 28 pieces to do for the fan, this is only barely acceptable.

Gunk builds up on the cathode fairly quickly, which eventually bridges to the work piece and generates a short. I have to sit there with it and clean things off every 5 minutes. Possibly with a pump this could be improved. My power supply eventually blew its fuse, thanks to all the shorts.

The results are decent, but surprisingly rough in places. A ferric chloride etch would be much cleaner. Some of this seems to be the exact laser printer/glossy photo paper resist transfer process that I'm using. It starts to peel away towards the end of the etching process. The quality of the non-resist areas is really nice, though, giving a very smooth surface.

Finished Lightsuit

2010-11-04T22:45:00.000-07:00

As some of my cryptic Twitter updates have hinted at, I've been working on a el-wire suit. Only 10 months early for Burning Man 2011!

At its core is an Arduino Mega, a Memsic 2125 dual axis accelerometer, and 24 channels of independent el-wire control. All this is mounted on a faux-military jumpsuit. It has multiple modes of operation, including a motion-reactive setting which turns the entire thing into a bar graph showing the local acceleration.

The el-wire control is based on the circuit described here. I etched 3 circuit boards, each with 8 channels of control. I've posted the SVG file for the boards here, should you want to use it. It uses the same components as in the link above.

And what does it look like? Something like this:

The battery held up for 5+ hours without any problems. A couple strands went dead, but I kind of expected that, given how delicate el-wire joints are. I'm hoping the failure rate will drop with time as the weak ones are replaced. It was reasonably comfortable to wear, though the el-wire attachment has to be improved. They were too loose and one on the ankle broke its stitching. I need to improve the programming on the accelerometer mode, adding some smoothing and non-linear scaling to the values. But not bad!

Difference Engine

2010-11-03T20:22:00.000-07:00

A couple of weeks ago I was lucky enough to be able to visit the Difference Engine at the Computer History Museum. It would be safe to say that I considered it something of a pilgrimage.

I will assume anyone reading this knows the general history of the Difference Engine. In brief, a 19th century mechanical calculator that was never built at the time, but has since been proved to be quite practical. Mostly notable because Babbage went on to design the Analytical Engine, also never built, which would have been a general purpose computer in every modern sense of the term. This is as close to real steampunk as the world ever got.

The principle behind the Difference Engine is actually pretty simple. If you take any expression f(), you can create a list of output values. f(0) = a, f(1) = b, etc. You can then look at the differences between the output values, and the differences between the differences, and so on. At some point, if f() is a polynomial, these differences will become constant. For an example, let's look at f(x) = x^2 - 5x + 10.

x	f(x)	f(x)'	f(x)''
0	10
1	6	4
2	4	2	2
3	4	0	2
4	6	2	2
5	10	4	2

Once we're looking at the difference of the difference (or the second derivative) we're looking at a constant value of 2 (which is what we'd expect, as f(x) is a second order polynomial). Neat, but so what? Well, now that we've worked this out, we can calculate further values of f(x) just using addition. We can now work right to left, adding f(5)'' to f(5)', and then adding that to f(5) and we get f(6)!

x	f(x)	f(x)'	f(x)''
0	10
1	6	4
2	4	2	2
3	4	0	2
4	6	2	2
5	10	4	2
6	16	6	2

So that's pretty cool. And that's exactly what the Difference Engine does.

Each of those columns is a 31 digit number. There are 9 columns, allowing 9th order polynomials to be calculated. Adding is a very simple operation mechanically -- gears turning gears at a 1:1 ratio. The complication comes from carry operations, which is what the helices on the columns in this picture do. (This is the rear of the machine.)

With the mechanism so far, you can calculate some fairly complicated expressions with a high degree of reliability. But this is the 19th century. You're not calculating them for use in a ray tracing or something, you want to generate a big book of trig tables. Which means that all these values have to be typeset. By hand. Backwords. Needless to say, this can introduce a lot of errors. If you're obsessive enough to build a Difference Engine in order to generate accurate values, this situation simply will not do. Obviously you need to design a printer as well, which is exactly what Babbage did.

It isn't enough just to print onto paper (though it does do that), because then the values would still need to be transfered to type. So it also printed into a soft plaster sheet, creating a stereotype mold from which an entire page can be cast. No chance for human error.

Now, this is all stuff that anyone slightly obsessed with the history of the Difference Engine would know. What is particularly exciting about there being real working Engines in the world is that one can talk to the people who really use them and have real working knowledge of the things. Which is absolutely what I did.

This is the base of the printer. That crank wasn't on the original plans -- it turns out that the printer jams fairly easily, at which point you need to run it backwards separate from the main Engine. So they added a crank at that end, and a little clutch to disengage the printer.

Also, notice the little cable at the bottom? That gets pulled when the printer hits the end of the page, at which point a lot of stuff needs to be reset. The cable runs the length of the machine...

...finally coming out at the main crank, where it attaches to a clutch on the bevel gear. Thus, when the printer ends a page, the crank is disengaged and the person turning it knows to stop. (This was in the original design!) What wasn't in the original design was the 4:1 reduction gear on the crank itself. It just takes too much effort to turn otherwise, not something Babbage would have been able to tell in the design phase.

Speaking of the main crank, it turns out that turning it requires a lot of training, because the Engine requires very different amounts of force at different phases of a calculation cycle. And it jams fairly easily, so you need to know when to apply a lot of force, and when it shouldn't need much force so stop immediately.

Back at the printer, it has these escapement wheels and adjustment cranks. The left escapement defines how many columns are on a page, and the one of the right sets spacing between lines -- you can just see the short-short-short-short-long spacing in the wheels, which puts a larger space between every fifth line, which is what you'd want in a table of numbers to make lookup easier. The crank adjusts the overall spacing between lines.

Other cool stuff lacking in specific pictures:

Most of the bolts worked loose while it was on the cargo plane being delivered. It was built only with mid-19th century technology, so no lock-washers or Loctite was used. Because of this they have been fighting jams and problems ever since. They recently made a series of adjustments to the angular position of the take-off gears from the main drive shaft, which determines the relative timing of various operations in a computational cycle. They managed to do this because one of the museum employees happened to be visiting London and could stop by the other Difference Engine there and take some measurements.

The printer works, but they do not normally have it set up with ink or plaster. This is mostly because period ink is water based, and would have to be cleaned off immediately to prevent corrosion. The one thing that is not rigorously period is the lubricants used. Original ones would be horrible animal-based ones and just not worth it. Originally this Engine used petroleum lubricants, with an associated machine smell. Upon encountering it, the wife of the owner (who will eventually have it installed in his living room) said that it was not acceptable. Ever since, they have used FDA-approved food equipment lubricants, which are silicone based and have no smell at all.

The thing that impressed me most of all was finding out that the 31 computational digits can be split up into multiple calculations happening in parallel, with correspondingly fewer digits of precision. The Engine as it is set up at the museum is running in a 27/4 configuration. The larger set is calculating a 7th-order polynomial for demonstration purposes. The smaller set is set up with just a 1 in the first column. Which means that the output column is incremented by 1 every cycle. In modern terms: printf("%d", ++i); It's a debug statement! A convenient little printout that helps the operators keep track of what is going on! I was utterly charmed by it -- I bet Babbage never thought about doing that. This is the kind of thing that only gets discovered when you're actually using a system. It's a delightful little look into a parallel world where Babbage didn't get quite so perfectionistic and actually got the damn thing built.

Finally, here are some videos I took:

Final road test

2010-08-24T21:48:00.000-07:00

The Kalamazoo is done, more or less. And it works, more or less.

It takes more effort to drive than I expected. The steering isn't very stable, despite the caster effect, which means I really only have one hand left to pump with. It's not going to be a very practical vehicle. Oh well.

I redid the horn mounting tonight, accounting for the change in primary driver position. Tomorrow I'll disassemble the entire thing and stain the wood. That's when I'll mount the etched brass builder's plate I made. I drive it to Spokane Friday after work in pieces, re-assemble it there, load it into my dad's Burning Man trailer, and head for the playa Sunday morning. Whew.

Lanterns

2010-08-22T23:09:00.000-07:00

When originally conceived, I planned on wrapping the Kalamazoo in el-wire. It's for Burning Man, after all, and everything there is wrapped in el-wire. But as work progressed, I kept adding more and more subtle, classy design details. By the time I got to seriously thinking about lighting, the idea of adding el-wire seemed really... tacky. I played with the idea of the solar-powered fake lanterns you can get as lawn decoration, which seemed to fit the evolved look better. But they're seriously not cheap, and, well, fake. So... why not real lanterns?

I ended up going with two at the front and two hanging under the pump-arm tripods. The forward pair hang from (custom, natch) smithed hooks. I went with a pigtail curlicue instead of the same 2D spiral I've used everywhere else. As well as adding some variety, this also helps control the swing of the lanterns.

The two under the pump arm hang from some cheap brass chain.

I let one of the lanterns burn for an hour tonight to gage fuel consumption. I put in 150 ml of oil at the beginning, and at the end I poured out... 150 ml of oil. So I don't think taking enough fuel will be a problem.

Road test #3 happened yesterday, with decent results. I tightened the chains today but didn't have an opportunity to take it out again. Running out of time to make modifications, but at least my to-do list is getting quite short.

Steering lever

2010-08-18T01:15:00.000-07:00

Road test #2

2010-08-16T00:09:00.000-07:00

Spent the last week designing and building the new steering mechanism for the Kalamazoo. After last week's disastrous road test, it was critical that this problem be solved. And while what I came up with isn't pretty, it looks like it will work okay.

In short, I bolted on a lever to the side, which connects to the caster via a connecting rod and ugly universal joint system. The motion is actually pretty nice, even when stationary. (On a nice smooth concrete slab, anyway.) I need to offset the lever above the deck to make it easier for the driver to reach, but other than that it's a much nicer interface than I expected. That was always a driving motivation for the old differential braking system -- how do you pump and steer at the same time? While not perfect, the lever approach is pretty decent. And it feels very awesome. Part rudder, part steam locomotive control. Not bad.

Today's road test did conclusively show that only controlling one caster was not enough. Particularly with any back-and-forward motion it was very easy to get in a bad state where the uncontrolled casters was out of sync, sometimes to the point where the Kalamazoo was completely stalled. It will be easy enough to add a tie rod over to the other caster. I always figured this might be needed, but decided to be lazy and try it this way first. Very little work wasted.

The other big issue coming out of the first road test was the failure of my hacked-up fixie wheels under load. So this week I threw a gob of cash at the problem and bought new ones. Unfortunately, no one makes real 20" fixie wheels. But I got a system the bike store people thought might work. Today's road test showed that, no, it doesn't work. This leaves only one palatable option -- I need to switch to full size bike tires. I only went with 20" because I wanted to keep the deck lower for stability (less of a concern now that I've been on the thing for real) and because I was using 12" pneumatic casters in the front. I might just have to live with an A-Team-van-style raised back end, or shim the front casters by a few inches. We'll see.

Overall, while the broken fixies were a big disappointment, the second road test went pretty well. I was even able to drive (slowly, jerkily) up a slight slope, showing the gear ratio changes are helping. There might be one more improvement there, if an ebay shipment arrives in time. I also have a big hunk of lead strapped to the crankshaft to improve its flywheel effect, though I'm unclear how much that actually helps at these low speeds.

On the way back down to the garage I road it and mostly let gravity do the work while I steered and counter-pumped for braking. This worked great, until it got to the steep section and I had to navigate several parked cars. At one point the rear end lost its grip and I fishtailed most alarmingly towards a well. No control authority, just suddenly riding along on the world's stupidest surf board. Important lesson: don't drive the Kalamazoo obliquely across an incline.

Road test #1

2010-08-08T22:42:00.000-07:00

Tony, Rob and I pushed the Kalamazoo up to the road from the garage today to do some real testing. As the video shows, the results were pretty bad.

Time for me to eat an entire murder of crows: differential braking for steering does not work here. At all. I now have a bit over two weeks in which to design and build something completely different.

The action is really jerky, with a massive clank at the top and bottom of the stroke. Haven't quite figured this out yet, might be related to slack in the chain drive. I know how to minimize the noise, at least, but I really want it running smoother than this.

Less seriously, but still annoyingly, we broke the fixed gear connection on one of the drive wheels. Hopefully I can just pay to get some real ones made up at Recycled Cycles. A cash-soluble problem, as I like to say. If that fails, I can always flip that one wheel around. It would break the symmetry of the drive chains (grrrr), but then it would be turning in the correct direction at least when driving forward. I cut the keyway extra long on that shaft precisely to handle this contingency.

I won't deny it, I'm now worried about getting this working in time. Particularly with the Laser Medusa Helmet work I also want to get done. Overall I'm really proud of the work so far, just have to get these last problems sorted out. We'll see...