Rachel's techniques for sewable circuitry: seems somehow soldering SMB parts to make them sewable. Related idea is flexible printed circuit boards.
(i believe that) Meredith asked for a Theremin-like means for turning water on and off: the water source could be a bottle with a siphon hose and clamp or inverted bottle with hose and clamp; clamp could be controlled by some kind of motor, maybe stepper motor, which would be controlled by the Theremin-like circuit.
Theremin did create a rotating cake for Clara Rockmore's 18th birthday, when she moved near the cake, it would rotate. Here is a picture: http://www.hologlyphics.com/electriccake-1.jpg
Use MOSFETS instead of bipolar diodes: the voltage drop across a standard diode is about 0.7 Volts. It seems it's possible to use a MOSFET (metal oxide over semiconductor field effect transistor) as a diode in drain follower mode. The voltage drop is much less than across a standard diode. as more amperage goes through the diode, more power is wasted. also for circuits with very low voltage power sources, because there is very little voltage drop, more voltage is available for the circuit load. I'd like to build a circuit or two to test this. Applications include remote, low-power systems. http://electronicdesign.com/Articles/ArticleID/19871/19871.html
check out Schottky and germanium diodes if you are concerned with forward voltage drop. The parasitic diodes in MOSFETs have large parasitic capacitances and long recovery times.
"Ultimate Continuity Tester", per Electronic Design, October 2, 2008: The article claims normal continuity testers trigger through connections of as much as a few hundred Ohms, which sometimes is too crude. Sometimes what's needed is a continuity detector that will not trigger through as little as ten Ohms. The circuit described is cheap, robust, and very sensitive. I'd like to try to build this. http://electronicdesign.com/Articles/Index.cfm?AD=1&ArticleID=19768&bypass=1
Sounds like you care about the actual resistance, which you can easily measure with an ohmmeter. Something you could do is make a 555 audio oscillator with resistance across the test probes influencing the frequency. That would give you audible feedback of magnitude rather than a binary go/no-go test. 220.127.116.11 19:58, 27 October 2008 (PDT)
Amplifier for function generator: it's nice to have an amplifier and speaker combo around for testing. There are various ICs that are pretty good audio amps, given a power supply and a few components.
You can scavenge some cheap computer speakers off the street or at Goodwill, and need not worry about frying them.
Audio mixer: it'd be a fun hack to mix the outputs of function generators and other audio range signal generators.
Audio cube: the idea is for eight independent audio amplifier-speaker combos in each corner of a cube with a mixer that could send signals to any combination. the control for the mixer seems tricky: a multiple pot comes to mind as does a cubic container with semi-conductive liquid with terminals in each corner of the cube and a stick with an end that presents a signal source to the liquid. There's always the CPU method of spitting bit slices to each corner amp per user-interactive controlled algorithms (the GUI or CLI user interface seems problematic in the same way as the physical interface, though the user could set things up programmatically--though this is reminiscent of punch card programming).
Power test resistors: a few 100 Watt resistors of various sizes could be useful. If i find some, I'll bring them in.
Multiple voltage power supply: a single device (i.e. a single 120VAC power plug) that provides various positive and negative voltages, probably most limited to less than one Amp. These are available, but in limited combinations: the hack proposed would allow any kind of extensibility.
you can make a (positive DC at least) adjustable power supply like that with a wall wart and a fifty-cent LM317
and an LM 337 can provide negative DC adjustable power supply. probably both are better with a cap or two. exar makes a 4195 dual tracking +-15 VDC regulator (i like exar, they seem to have good knock-offs).
Anything with 12VDC, low voltage, low power, off-the-grid....
Is it possible to make a signal generator that drives an antenna at the frequency of visible light? Here's the tho't: we hear multiple octaves, about ten or so (piano-type keyboard devices generally support several octaves, a guitar supports four or five octaves...). But we see just a little less than one octave (we see a low frequency red through orange, yellow, green, blue, upper frequency violet; below red is infrared, above violet is ultraviolet, an octave is just below red to just above violet). I think it's possible, conceivable anyway, that we can train ourselves to expand our visual perception: to recognize a little below and a little above what we're used to and possibly stretch enough to see the octave: an infrared color that is half the frequency of an ultraviolet color. Singers exercise their voices to extend their ranges both downward and upward. It may be possible. If so, for the person who achieves this, their sense of color harmony will change, for in recognizing an octave in light, they can make better sense of harmonic relations that are apparent in music (the fifth and fourth degrees of a scale, major and minor thirds and sixths, augmented and diminished chord...). But the trick is training, and for that it seems is needed an antenna that generates light and can be modulated to produce the entire visible spectrum and beyond, both above and below.
Jonathan Foote writes: see my speculations about this at http://www.rotorbrain.com/blog/2008/10/on-harmony-of-light.html
thanks lots! i checked, i liked Q a lot, tho' i didn't like A very much.