0
\$\begingroup\$

I'm trying to create 555-timer based Theremin, but I have a few questions that I don't quite understand. For now, I'm building it on a breadboard (yes, I know it's not the best option), but I don't want to commit yet to some schematic that might be wrong. My current idea is to use 555 timer in 150 kHz range.

For this, I have 50 pF C, R1 of 1 kOhm and R2 of 100 kOhm (following this convention).

Checking with an oscilloscope I get around 100 kHz. But when I connect a wire (antenna) to pin6 I don't get any change in frequency. My idea was that connecting an antenna to pin6 would create a capacitor in parallel with 50 pF, and that my hand should then vary that capacitance, thus varying frequency. But the only change in frequency occurs if I touch the bare wire, not by proximity.

Am I doing something conceptually wrong, or is my 50 pF capacitance too large (I'm guessing "antenna" capacitance is probably less than 5 pF?)

Oscillator difference

My other question is concerned with mixing of two signals for Theremin (one from fixed oscillator, one from variable oscillator - the one connected to antenna). I've seen schematics where op-amp is used to subtract signals, but I'm not sure how relative phase of the signals affects the output of the op-amp? Surely this all works only if they are in phase? How is this problem solved in other signal mixers?

\$\endgroup\$
4
  • 1
    \$\begingroup\$ Have you actually looked for a Theremin circuit on-line (there are examples)? Have you researched what "mixing" actually means in the context of a Theremin? \$\endgroup\$
    – Andy aka
    Jan 9 at 11:39
  • \$\begingroup\$ also, look at the 50 pF: That's very much in the region of stray capacitance between "clips" in a breadboard. Your breadboard choice is not "not the best option", it's having as much an unintended effect on your device's output as your intended capacitances. You simply will have a very, very frustrating time doing this with a breadboard, as the effects you want to observe are smaller than the effects you don't want but still get through your breadboard. \$\endgroup\$ Jan 9 at 12:19
  • \$\begingroup\$ Does this answer your question? What order of magnitude should I expect the parasitic capacitance to be for a solderless breadboard? \$\endgroup\$ Jan 9 at 12:19
  • \$\begingroup\$ @MarcusMüller Stray capacitance on a typical solderless breadboard is more like 2pF (so if you ground the strips on either side of a strip you'll get around 4pF from the center strip to ground). \$\endgroup\$ Jan 9 at 15:54
0
\$\begingroup\$

Theremins are based on LC oscillators, using the self capacitance of the external L (antenna) which obviously varies with hand position.

So, an RC oscillator (555) is a poor choice.

(So is using a crystal oscillator for the reference, as I found. Both oscillators should be similar, so that their thermal drifts are similar. Mine drifted like crazy because the reference was too stable! Yes, I walked right into that one.)

And as commented, mixing is at RF, producing a beat frequency which is low pass filtered to produce the audio.

You want both oscillator waveforms to be reasonably close to sinusoidal; mixing square or sawtooth waveforms from an RC oscillator - or using an XOR gate as a mixer - will produce something decidedly nasty sounding.

The classic Theremin sound relied on the much softer sound (fewer odd harmonics) of vacuum tube "detector" circuits like the "anode bend" using the non-linear portion of a tube's characteristic. You can do something similar with a diode mixer, or a transistor or FET biased just out of cutoff (not enough collector current for symmetrical operation : turn the base bias voltage down until it sounds right)

\$\endgroup\$
2
  • \$\begingroup\$ So if I understood you correctly, the pitch antenna is more like L than C, and because of that LC oscillator is more preferable? For mixing I wanted to use RC filter before to turn square into sine, but I understand what you wanted to say. Is an op-amp poor choice for mixing then, or is it an overkill, as a transistor should be enough? \$\endgroup\$ Jan 9 at 22:51
  • \$\begingroup\$ Opamps are entirely unsuited to mixing. They are supposed to be linear, for a start, which would prevent any mixing from happening. And most don't have the bandwidth required to handle the oscillator frequency. \$\endgroup\$ Jan 9 at 22:56
1
\$\begingroup\$

I don’t have Theremin recommendations at this time but two fingers about a ~ cm apart is 1pF then inverse squared , so change the wire to a small area coil or foil. You can mix 2 100kHz oscillators in a XOR gate or diode mixer. Try that.

Capacitance depends area and gap. Then 50pF fixed is about right if the same f. But stray capacitance between sensors may cause both to “lock”. Injection locked loop (ILL)

Report back with photos and sounds when done.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.