Intro to Physical Computing
Fall, 2003
Jeff Feddersen

Week 6

The site that Morgan found is really cool.
And here is another: Any site that has a link named "Fun with High Voltage" has to be good.

Take a 9-volt battery and touch it to a piece of steel wool. Instant fun!

I bought the radio kit from Radio Shack. It was kind of fun to put together. It would make a good homework assignment for people still getting used to electronics. The kit uses tightly coiled springs to hold wire leads in place, rather than solder. I think I might use those in my projects.



For the Halloween class, Ana and I will try to make a disembodied hand (when else do you ever see the word 'disembodied' except when refering to a hand?). There are many possible ways to do this, but I think we'll go with a single motor and some string inside a glove. Maybe some broken chopsticks for bones.

I was working with Christina Goodness (she's in Milmoe's section) on programming the bx24 to light up one of those 7-segment LED displays. Neither of us could get the print function to work! Everything else seemed fine, but it was hard since we could never see the pin values. We both use v2.1 of the software at home, but were using v2.0 at school. I wonder if by initially using v2.1, the chips somehow got 'calibrated' to that version of the software, and thus wouldn't work with v2.0.

I built a simple brushless motor, using magnets, copper wire, and the most essential part, one of those magnetic switches Jeff Described.

The wire is wrapped around a large iron bolt, making an electromagnet. A tube is perpendicular to the bolt, and the tube has 4 regular magnets, with the south pole facing out. At rest, one of the four permanent magnets is near the switch, keeping it closed because of the magnetic field. When current is on, the electromagnet generates a magnetic field that pushes away the closest permanent magnet, which rotates the tube about 45°. This causes the magnet on the far side of the tube to move away from the switch, which then turns off because the field is removed. The switch being off opens the circuit and the electromagnet loses its field. The tube then rotates another 45° because at this point another permanent magnet is near the iron bolt, which is no longer magnetized, and attracts the magnet.
So, the switch turns on and off 4 times for every revolution of the tube, clicking audibly
The design and parts came from a Web site which I can't remember now. The tricky part was positioning the bolt, magnets, and switch in such a way so that the tube spins with just enough momentum to bring it through an entire 1/4-turn. Sometimes I have to start the thing spinning manually, because the electromagnet isn't quite strong enough to push away the first magnet when at a dead stop.

I had to eat a few boxes of Fruity Pebbles in order to get the UPC symbols, mail them in, and get my free "Wild Planet Spy Detector".

It looks kind of like a spaceship. Maybe I can make a cool thing with the case.
I got it for one part it has, a visible light detector. Any change in visible light, such as turning on a lamp, or just waving your hand in front of it sets it off. I assume it's a kind of variable resistor. It's extremely sensitive, so quite unlike the photo-sensitive resistors in the supply cabinet. It has a lens rather than a filament.

It's fun to be able to take stuff apart and now actually be able to identify the components.

Resistors: 10
Transistors: 4
Capacitors: 4 (2 big, 2 little)
1 LED, 1 switch, 1 speaker, 1 light detector
Whoever made this used the technique of using a little hot glue to hold some of the bits together.

I owe Steve Oh a new set of jewelers screwdrivers after I stripped the heads off of his trying to take apart a cd player.

After last week's "incident" when I blew up a 7805, I rearranged my power input to run with + on the pin rather thon on te sleeve. So now I can use the DC power supplies at school. I also got a power supply that let's you switch between polarities, select from 3, 4.5, 6, 7.5, 9, or 12 volts. It also can switch between 100 and 300 milliAmps, just in case.

In 1832, New York University was founded. Of note, it was the first university in the United States to have a fine arts program. Until then, schools only taught classics, literature, and science and humanities. The first professor NYU hired for it's new department was a man named Samuel Morse. While teaching out of a room in the northwest corner of what is now the Grey Museum, Morse invented the telegraph. In some ways, the telegraph was a more significant invention than the Internet, since the telegraph was the first, the very first technology that allowed people to communicate in real-time while not actually being face-to-face. I had no idea that NYU had this kind of legacy regarding the bridge between technology and art, specifically someone coming from an art background and creating a technology (not just a device) that was truly useful, and the forebear of everything related to the Internet today.

Step 1:
Get a DC motor that runs on low voltage DC, in the 5-15V range. Connect leads to its terminals, and run if from a benchtop power supply in the lab. Try changing the voltage on it, and seeing what effect it has. Don't go over the motor's rated voltage. Connect a switch in series with the motor and use it to turn on the motor.
I've been scavenging motors from the junk shelf since we started 6 weeks ago, so now have a bunch from which to choose, although I don't know their voltage ratings. They all seem to run fine from 2 AA batteries, which is only 1.5 volts. They run on 12 volts as well, although super-fast, and I get that nice whiff of ozone when I power them up. Looks like higher voltage means higher speed.

If you want to use flux when soldering, you can just use ear wax instead. It's cheaper, and chemically identical.

Step 2:
Connect the base of a TIP120 transistor to one pin of your BX-24. etc.
The tip120 has its pins arranged as follows: BCE, so pin 1 goes to the chip.
I need a 1N4004 diode. While a transistor has three pins, NPN, and acts as a switch, a diode has 4 pins, PN, and acts as a valve. I suppose I could use a 'light-emitting' diode here, but I have the real McCoy, so... but which way is up? Fooling around with an LED tells me that the silver stripe needs to go on the ground side of the diode.
Get the motor hooked up, using a pot instead of a switch, seems okay, but my motor sounds like it has a plegm-y throat, maybe it's half-way burnt already.
Using a 9v battery to power the motor.

When I was working at PC Magazine, I wrote an article titled, "You don't Know your ASCII from your Elbow."

Can't really tell what's going on in the photos in the web instructions, but I hook up my pot to pin 8, and have a program to turn on pin 13 (the tip120)

Sub main()
	call delay(0.5)  ' start  program with a half-second delay
	call putPin(26,1)

		call putPin(13, getPin(16))
		debug.print cstr(getPin(16)) ; " - "; cstr(getPin(16))
end sub
It works: whenever the switch is on, meaning pin 16 is grounded, the motor shuts off.
But now I've been through 3 9v batteries - they run for a minute or so then shut off and I can't get them to work again. But the multimeter says they're okay. I try with a pair of AAs, and that seems okay, but now the motor runs regardless of the tip120. Hmm...

We've got quasi-CS with ICM, and quasi-EE with IPC, but I think ITP needs a quasi-ME class. Something that covers pulleys, levers, and how to mount these motors so they don't fall off as soon as they turn on.

I skip the batteries altogether and run power to the motor from the 12vDC input, in parallel with the 7805. Now the motor pulses, exactly 76 times per minute - hmm... The switch-controlled part works fine, though. I take out the diode, put in backwards, makes no difference. Why the hell is it pulsing! I take out the print line in the code, so the only thing the chip is doing is controlling the transistor switch, but it makes no difference. It must be a function of the ac power coming from the 12vDC power input - but wouldn't that cycle at 60Hz? This is cycling at about 1.27Hz

I've got it. Each time the motor turns on, the power load leaves the chip with too little current so it shuts off, which shuts off the tip120, which shuts off the motor. Once the motor shuts off, the chip has enough power to start again, etc.

Clara Williams and Jeff's marionettes at 80 Arts have a blurb in this week's New Yorker. Movies must be a hot topic for a theme issue - I've never seen so many fold-outs and insert ads.

I'm having trouble this week. I can get the bx to switch the motor, and I can get the motor to run, but I can't get the two together. I think the battery (now, a new one) is too powerful and is overriding the transistor, which is switched off, because pin 13 is 0.

I wonder if I have the transistor set up backwards, with the emitter and collector swapped, so I switch the connections, and... nothing changes. The motor still ignores the transistor. Check with the multimter, when switch is off, no voltage going across transistor - thus "off", right? Then turn switch on, and I have 5v going across the transistor - thus "on". Oh, I'm an idiot. All I had to do was look more closely at the schematic. I had the motor going to ground instead of to the collector, along with the diode. Works now.

Sub main()
	call delay(0.5)  ' start  program with a half-second delay
	call putPin(26,0)

		debug.print cstr(getPin(16)) ; " - "; cstr(getPin(16)) ; " - "; cstr(getADC(15))
		if (getADC(15) > 0) then
		call putPin(13, getPin(16))
		end if
end sub

I have my light-sensor from the little space-shippy thing on pin 15. I have a switch on pin 16, and a pot connecting the battery to the motor, so I have pretty good control. The motor has a fan plunked right on it, and the motor is held by a rubber band.

It occurred to me to make a hovercraft, but I couldn't get anywhere. Maybe some wheels.
Actually, the code above is too sensitive, and any ambient change in light in the room sets it off, so I set it to 50

Matt Slaybaugh
ms171 at