A one chip theremin

I found a one-chip "octave down" pitch shifter to go with your one-chip optical theremin.

http://www.falstad.com/circuit/e-divideby2.html

Thierry, thanks for pointing out that the IC draws too much current in this circuit and for suggesting a voltage divider. How about a voltage divider where one of the resistors is a CdS photoresistor and the other is a potentiometer? The photoresistor would be the volume control. The potentiometer would adjust the sensitivity of the volume control to compensate for different lighting conditions. Should I put a 0.22 uF capacitor between the chip and the voltage divider as suggested in a previous post?

I had also thought of using colored LEDs with a colored filter over the chip, but I was thinking of using red LEDs, since the TSL230R is more sensitive to red than green or other colors. Ideally, the filter should be a narrow bandpass filter that only passes the wavelength generated by the LEDs, but those are very expensive. Combining several cheap filters might work.

It should be possible to use LEDs and filters of a different color for the volume control. That would allow the volume control and pitch control to be close together with little interference between them.

Gordon, the divide by 2 circuit is very intriguing. It would be interesting to use its output as the fundamental note, and mix it with the output of the TSL230R as the second order harmonic. Square waves contain only the odd harmonics of the fundamental note. That is why they sound harsh. Including the second order harmonic should make the tone more mellow and richer sounding. You could even have separate side-by-side volume controls for the divide-by-2 output and the TSL230R output. That would allow you to change the timbre of the instrument during a performance.

Please tell me more about the photocell (maximum resistance when it is dark and minimum resistance when there is light on it) so that I may calculate the voltage divider and the coupling capacitor.

Out of that, concerning harmonics: A square wave has odd harmonics and sounds harsh, that's true. A triangle wave has the same odd harmonics but less of each. So it sounds mellow, flute-like. Even harmonics will give a more buzzy, string-like sound.

Isn't it intriguing. And it generalises to a divide by n counter, so you can create any sub-harmonic - which gives lots of ways to mix your own waveform out of square waves in different ratio to one another.

You can edit those Java simulations - control-click (on a mac) on the black part of the window. (I tried feeding two square waves into an XOR gate because I thought it might be like a ring-mod in some way. It is - the difference is that instead of amplitude modulation you get frequency modulation.)

Thank you, Thierry. I have three kinds of CdS photocells. None came with specifications. I measured their resistance with an ohmmeter. My ohmmeter only measures up to 2000 k ohms. All three had greater than 2000 k ohm resistance in complete darkness. All numbers are in k ohms.

Type 1:
full sun = 0.06
bright room = 0.8
dim room = 12
complete dark > 2000

Type 2:
full sun = 0.08
bright room = 1
dim room = 15
complete dark > 2000

Type 3:
full sun = 0.2
bright room = 2
dim room = 40
complete dark > 2000

Any of these could be used in the theremin. Thank you!

Jim, please allow a few days for finding an optimal solution for you. I'm currently over-occupied but should be able to do something for you on Sunday evening.

There's a discussion about ways of turning a square wave into a sine wave on the Lunettas forum on electro-music.com.

http://electro-music.com/forum/viewtopic.php?t=29403

(Apparently "Lunettas" are synths made from cheap logic circuits - inspired by this thread I shall be building a lunetta effects box for my etherwave over the coming months.)

The easiest thing to do is just some simple analog filtering. A state-variable filter (a.k.a. a "parametric equalizer") is just the thing ... since you can vary frequency, level and Q independently.

You will lose a fair amount of amplitude, but that can easily be compensated for by following the filter with an amplifier stage.

Don

Hi Thierry, Have you had a chance to do the circuit calculations? In the meantime, I've started working on the instrument that I was planning when I was sidetracked by the TSL230R chip.

Oooops!

sorry, forgot it because there is so much other work to do. But I'll attack this still this evening - promised!