Jake of All Trades

My university’s chapter of the IEEE held a “Hacks/Mods/ Gadgets” contest a while ago, in which I entered the Telecalulograph and Vitorio-Nixie Tube. By sort of winning (there were only two other entries), I was awarded the chance to build the project of my choice–with them picking up the tab! I rather quickly realized that this would be the perfect opportunity to construct something that had been nibbling at the back of my mind for some time: a steam-powered Baird Televisor. While many people have built reproduction mechanical TVs, I am (to the best of my knowledge) the first to power it directly with a steam engine. I am beginning to see that there is good reason for this, but I am determined!

The project is in it’s infancy now (hence the “Part 1”), so there isn’t a whole lot to see yet. The physical framework and most of the basic components are built and functioning on their own, but the is an incredible amount of work to be done before the thing actually does anything interesting. I will be posting updates as I go along, so stay tuned! Photos and descriptions of what I have so far are after the jump:

An old telephone table (which, fun fact, served as my night-stand in my younger years) forms the base of the device. The drawer has a breadboard mounted inside, and brass terminals and a keyed power switch on the front. The point of this project is to keep the use of electronics to a minimum (odd for an IEEE project, eh?), so the few necessary bits are hidden away.

The record acts as a nipkow disk, though the holes have yet to be drilled. The light blinks on each time a hole is positioned wherever a “on” pixel is desired. In this way, it draws the image one dot at a time–though it does it fast enough that you perceive it as a solid image, similar to the way the Victorio-Nixe works (persistence of vision). The image a disk of this size produces is 32×80 pixels, and less than an inch wide. It is also capable of sequencing these images into video, at a frame-rate of 12.5fps. Beginning to understand why this technology didn’t go very far?

The engine is a Jensen #75 purchased for my by the IEEE from Mini-Steam. A bit of copper tubing directs the “used” steam into the flask. Things get rather messy otherwise…

To provide power to the electronic bits, I plan on using a DC motor as a dynamo. The one pictured doesn’t give me enough juice, but I’m working on it. I added a much heavier (and brass!) flywheel, which should increase torque quite a bit. The flywheel was courteously machined for me by my room-mate out of a large brass pulley I found in a scrap bin at school. I also added bearings to the two points where the drive shaft mounts to the base. They’re cute little things I scavenged from this pH meter, that happened to fit the shaft perfectly. I’ve yet to test it with these modifications (too cold out), but I think they’ll help things quite a bit. I’m trying to irreversibly-modify the engine as little as possible, as I’d like to use the engine in other things once this project’s done. I’d also like it to become a sort of heirloom someday too… Besides widening the mounts to accommodate the bearings, I also replaced some of the rivets with nuts and bolts. It killed me to take a drill to this lovely machine, but I believe that Jensen would have done these things if cost weren’t an issue. I also foolishly drilled those four holes you see to mount the dynamo before I came up with a better plan. I’m kicking myself for that, but there was, at least, a hole there to begin with…

This is what supplies the afore-mentioned pulsing light. It began life as a reproduction “Edison Bulb”, which my dad kindly burned out. I successfully hollowed it out and replaced the innards with the bits you see now. Everything was going great until I decided that some little, insignificant detail needed to be tweeked. Though I fixed that thing, I managed to crack the heck out of the glass. Typical Jake… Right now it’s glued together, but I will actually replace it when the project is closer to completion. Anyway, on with the show. Five orange LEDs (four 3mm, one 5mm) are wired in parallel to mimic the neon bulb that a “real” mechanical TV would have used. The brass tubes support the thing and carry the voltage to the LEDs, though the spiral of bell-wire is purely ornamental. The “ceramic insulator” is really the cap from a ChapStick which, along with a bit of plastic in base, holds the tubes firmly in place. The brass sheet, which I polished and covered with a layer of packing tape to improve reflectivity, focuses this light forward onto the…

Diffuser screen. To make the image look half-way decent, you need to diffuse the light evenly over some sort of surface. Otherwise, the pixels near the center will be way brighter than the others. To accomplish this, I took a clear (UV, actually) 55mm filter from an old SLR camera and sandwiched a sheet of waxed paper between the glass and the frame. Surprisingly, it does the job quite nicely! It attaches to the cabinet with the brass wire, which is jest flexible enough to allow it to be adjusted to the proper position.

This is how it all goes together. The magnifying lens was removed from my trusty “Helping Hands” long ago, because it did nothing but get in the way. On this device, however, it makes the picture bigger and easier to see (as you would imagine). It’s mounted on an Erector Set armature to allow adjustment, as well as increasing the general mad-science look of the device 🙂

The red Erector Set bits will support the mechanical linkages that connect the disk to the engine. A series of pulleys will reduce the high-RPM output of the engine to around 780RPM. The disk needs to spin at exactly 750RPM, so an as-of-yet-unbuilt device will dynamically regulate it down to this speed. There are several ways to do this that I’m considering, but actual construction is a ways off.

So, where is the picture/video information coming from? Well, the fun thing is, the image quality is low enough that the data can be contained in a signal typically usable only for audio. This means that just about anything that can transmit/receive or record/playback music can be used to control this device. In fact, most people that build similar TVs are fans of HAM/CB radio — over which the “video” signal can be transmitted. For this project, however, the signal will likely be played back via dirt-cheap and hack-a-licious JuiceBox mp3 player. PC software converts JPEG pictures and/or AVI video into stereo sound files. The left channel of this contains the actual video data, which is connected (through amplification) to the LEDs. You know how some stereos have a little light that blinks in time with the music? Well, this is the same idea. What would sound like a loud “pop” if you were to listen to this audio file, causes the LED array to blink on for a split second. As I explained earlier, these pulses of light form the dots that produce the picture. The thing is, the disk’s holes have to be in the right place at the right time for the image to be produced properly. This means that the audio signal has to start at the exact time that the first hole is over the location of the top left-most pixel. This will be accomplished though the use of a simple latch circuit on the “play” button, which will delay the command until a sensor sees that the disk is at the proper position. This is also why the aforementioned rotational speed of the disk (750RPM) is so important. The right channel contains a 12.5Hz pulse which can be used to do this. If an LED were to be connected to this, it would blink at the same rate as one that was rigged to blink with each revolution of the disk (12.5 pulses per second x 60 seconds =750 pulses per minute). If the two pulses don’t match up, the disk’s speed needs to be adjusted. This will be done automatically, but the LED thing should help to illustrate it… Read more about these specifics and even build your own here!

Still awake? Hehe, well, anyway… There’s still much work to be done, and none of it is going to be terribly easy. I’m firm in my belief that it will work, however, and am planning on having it done by September. So, thanks for reading this and stay tuned!

PS: I realize this may sound terribly snobby, but I’d like to request that you not share any ideas or suggestions with me unless I ask. Wow, that did sound awful! Anyway, the reason I say this is that I want this project to be as “mine” as possible. There are so many of you brilliant folks out there that could probably build this ten times better and a hundred times faster than me, but I’m afraid that would spoil it for me. This is all like a huge puzzle, which is no fun if someone tells you the answers! It’s about the journey, not the destination– you know? 😀 I do welcome your comments and questions, however!