Intro to Physical Computing
Fall, 2003
Jeff Feddersen

Week 2

Meat Resistor
When I was a kid, Mr. Wizard was on Nickelodeon and he cooked a hotdog with some apparatus that basically electrocuted it.
I simplified his design by taking the power cord from an old lamp, and sticking the wire ends into either end of a hot dog.
I plugged in the cord and pretty soon the hot dog started sizzling and spitting and cooked pretty quickly.
In fact, I had to run over and yank the cord from the wall when the plastic insulation started to smoke.
It tasted okay (the hot dog, not the insulation), like a microwaved one.
Hot dogs have a lot of salt and water, and salt water is one type of electrolyte, meaning a liquid conductor, I guess. But all the animal proteins in there aren't good conductors, they are resistors of sorts, meaning when electrons pass through them they create heat instead of just coasting through.

Canal Street is the place, the palace, even.
Kat said that at 269 Canal is some guy with a whole lot of stuff in the back.
Just walk up to him and say, "Hey man, how 'bout hookin me up with a little som'n som'n?"
Trust me, he'll know exactly what you're talking about.

No, seriously, I talked to him and it sounds like he deals with lots of students doing projects, so he should be able to tell you which resistor goes with which capacitor. Just remember not to mix plaid with stripes.

Still to do:

  • Pick up readings
  • Signup for cleanup time
  • Take safety class
  • Complete this week's project
  • Buy books
(hmm, that's everything that was on the list last week)

I read The Design of Everyday Things back in college for a class, but I don't have it on my shelf. I should get it.

Jeff has a thing at the Brooklyn Museum of Art with Clara Williams involving robotic marionettes.
I don't see anything on the site about it. Maybe I got the name wrong.

There's a guy from Honeybee Robotics speaking tonight at 6:30 in the Japanese Room. I almost had a gig at Honeybee last year, but the grant from Nasa didn't materialize.
I would have just been rebuilding their Web site, not making robots, but it still would have been cool.
Check out their facility if you can. It's weird to see such an industrial-looking shop floor in the Lower East Side.
Their current projects include the rock-core sampler on the current Mars mission, and some other Mars stuff, and some kind of robot that would build communications satellites in orbit instead of having to deliver them via rocket. Pretty freaking cool.

I took some photos on the cheapest piece of crap digital camera I've ever seen.
But for $40 it's amazing that it does anything.

My answering machine was in bad shape; the messages were almost unintelligible. I cleaned the tape head, but that didn't help, so I bought a new digital answering machine and sacrificed the old tape one for Science and Truth.
When I opened it I found lots of dust embedded in the switches and especially the 'lateral potentiometer' used for the volume control. Maybe that was the problem.
It has quite a lot of good bits on it that I might be able to salvage: switches, speaker, microphone, motor, lots of diodes and triodes and I think I saw few quadrodes - not really.

The bread board fits quite well in the answering machine case, so I guess I'll use that for my housing. The case already has holes for switches and cables to go through.

I had been keeping my keyboard from 1989 for mostly sentimental reasons, since it doesn't play well now and fixing it would cost more than buying a new one that's ten times more sophiticated.
So it too is being sacrificed for Science and Truth. It has some mondo speakers. I had been keeping it behind my computer monitor, which kept showing weird colors whenever I moved the keyboard. I guess the monitor uses magnets to align the display of the cathode ray tube, and the speaker magnets messed them up.

Onward and upward. And sideways.

I read a joke on Slashdot:
If it doesn't move, but it should, use WD-40.
If it moves but shouldn't, use duct tape.

I soldered the power connecter thing. But discovered my housing didn't have the right-sized hole to stick the connector through.

Fortunately, soldering irons are good at making holes in plastic. (The blurry area in the center). Watch those fumes, though, phew!

Anyone else notice that fingernail clippers make decent wire strippers? Just don't clip too hard. You have to feel the difference in resistance between the insulation and the metal.
Actually, plain old fingernails seem to work okay too.


It looks like everything is working. The LED is illuminated and none of the parts are smoldering or even warm.
"Indescribable joy"? Well, I guess that describes it pretty well.

You can even see the LED when the cover is back on. (I swear this picture isn't from before I took the answering machine apart.)

Next, the
switch. Stay tuned.

Napalm Adhesive
The problem now is how to affix the switch in the case so that it doesn't move when the switch is flipped. Even a little give could mean broken solder points.
I want to attach metal to plastic. Glue doesn't work, solder doesn't work. Melting the plastic a little doesn't work. And there isn't room to use screws.
My solution: get some nail polish remover (acetone, or kersosene or gasoline will do) put a little in a glass bowl or jar.
Get a styrofoam (polystyrene) egg carton, or coffe cup, break it into little pieces and watch it dissolve.
Mix it all up and you'll have a napalm-like substance. Add more acetone or polystyrene until you have a jelly-like texture. This is extremely flamable, but also is good for molding plastic, since once the acetone evaporates, you're left with a hard, brittle plastic. It's still styrofoam, just without the foam, since the air is gone.
I used this to mold a shell around the switch, binding it to the case. We'll see if it works. I'll leave it for a few hours to dry.

Ugh. In the future, I'll make sure to use a toggle switch, one that flips. My switch slides back and forth, and it requires a much stronger adhesion to the case.

Step 2 of this week's assignment is to wire some LEDs in series to see the effects. The picture doesn't really show it, but they all get dimmer every time another one is added (a metephor for bureaucracy?)

While placing them all in parallel leaves them all shining bright. In fact one was getting kind of hot, maybe too much of a load. I notice some LEDs shine in different colors, depending on how much juice there is. That's cool.

Step three is to add a variable resistor. Okay. Step four is to build a 'combination lock, a puzzle, a trip switch'. What?!?! I think I'll need some more hand-holding before I attempt that.

So put in a second potentiometer (the former volume control of my answering machine) - it's nice because the leads go right into the breadboard.
The first pot is an audio taper so the dimming from one LED to the other isn't linear, and the first one never goes completely off.
I guess this now qualifies as an absolute minimum combination lock - imagine a lock with two dials, each of which has only two numbers: zero and one.

pot a position | pot b position | circuit
0 0 off
0 1 off
1 0 off
1 1 on

It works, although I wouldn't use it to lock up my bike.

From talking to the other students, it sounds like most are still working on getting their first LED to light (and it's Thursday evening already). So I guess I'm ahead of the game.

Oh, and you can use a candle to melt off insulation instead of using strippers. The plastic burns off and leaves the metal behind.

Matt Slaybaugh
ms171 at