basic experiments

I am just very happy. I have got some of that air coils you need to built a - temperature -  stable theremin. Combined with clapp oscillators they are not to beat. So I can use a reference oscillator with ceramic resonator.  

Currently, the so-called motorboat effect at very low pitch is in my interest.  I suppose that comes from a 50/60 Hz hum modulation, which is captured by the high inductance coils. Another cause may be the incompletely shielded high frequency, which is scattered between the two pitch oscillators.

"Currently, the so-called motorboat effect at very low pitch is in my interest.  I suppose that comes from a 50/60 Hz hum modulation, which is captured by the high inductance coils. Another cause may be the incompletely shielded high frequency, which is scattered between the two pitch oscillators." - JPascal

I'd look first at coupling via the mixer, and second at coupling via the power supply.

Air core coils are amazing!

The 50/60 Hz problems you can have also with well grounded battery driven theremins. The reason: Over the pitch antenna and the high inductance serial coils one get up to some hundred of millivolts from electrical environment fog in houses. Fortunately this voltage is almost short circuited over the second resonance coil to ground. But not completely! The result is a modulation of the main signal and all overtones. You get additionally a more or less weakly plus/minus 50 Hz splitting of the main signal and all overtones.  That sounds a bit scratchy and not clear especially in the near of zero. 

"The 50/60 Hz problems you can have also with well grounded battery driven theremins. The reason: Over the pitch antenna and the high inductance serial coils one get up to some hundred of millivolts from electrical environment fog in houses. Fortunately this voltage is almost short circuited over the second resonance coil to ground. But not completely!"

The Phoenix uses a large (compared to the tank) coil to ground, something I thought of a while back but don't believe I mentioned here on TW.  I ruled it out for the D-Lev because the coil has to be significantly larger than the series tank coil to not impart significant phase shift, I can deal with hum in software, and the hum imparts some dither to the input, which is kind of a good thing.

In my experiments I've found that a high value resistor from antenna to ground lowers hum/noise significantly without hurting the voltage swing too much.  It forms an RC lowpass filter with the intrinsic C of the antenna & hand.  So if you use 470k and have a 10pF antenna, you get -3dB at 34kHz, which at least kills some of the super higher harmonics and other environmental noise?  Though it directly hurts Q, so I don't do this on the D-Lev (perhaps it is something I should look into though).

"The result is a modulation of the main signal and all overtones. You get additionally a more or less weakly plus/minus 50 Hz splitting of the main signal and all overtones.  That sounds a bit scratchy and not clear especially in the near of zero."

I never thought of it from that angle (harmonic modulation) before (another reason to go digital) though I have thought of the influence of the mains hum field on the tank.  Mains hum is coupled through the antenna intrinsic C, so adding a series C generally won't reduce it as the intrinsic C is so small.  And the effect is FM/PM on the tank frequency.  It acts like a hand waving near the antenna at 50/60Hz.

Thank your for the remarks, dewster. To illustrate the hum modulation effect from antenna coil here a fft spectrum with 50 Hz sidebands. Avoiding measures are here not done.

Here comes a quick and dirty video about this Motorboating Effect. Caused by 50Hz hum modulation catched over the pitch antenna a theremin may sound like a motor boat called motorboating effect. Sidebands of 50 Hz distance to the main signals and overtones overlap in the area of low audio frequencies. One can avoid this by special filtering in the pitch oscillator circuit.

I'm curious: Should a built theremin either use a rotary switch for certain registers such as cello, voice, flute or a continuous knob for the timbre, as is the case in most theremins, or both? What wishes and experiences are there?

"I'm curious: Should a built theremin either use a rotary switch for certain registers such as cello, voice, flute or a continuous knob for the timbre, as is the case in most theremins, or both? What wishes and experiences are there?"  - JPascal

Great question!  I think it's a matter of how complicated your "synthesizer" is and how many adjustments or "knobs" are brought out to the user interface.  If I understand it correctly, even the relatively (conceptually) simple EWPro sound generator has presets of a sort, with one "user" preset configurable via the knobs and the rest static "factory" presets (so it does both, as per your question).  If all there is only a "bright" and a "wave" knob like on the EWS, then you can get away with no presets.  Once you get to perhaps three or more knobs that profoundly change the timbre / pitch range / volume response / etc. then you will be doing the user a big favor by implementing a preset system.

(Also, once you have a voice preset system you can have a subset of global "system" parameters stored in one slot, which relieves you from having these settings effectively "stored" on dedicated panel potentiometers.  That is, all panel knobs and switches are potential preset state fodder - including the preset selector itself!)

Thank you for the response, Eric. So should I search for a voice setting that is the mother of all timbres? I am still looking for such a timbre yet. Some results I will post here soon. 

(My query concerning convenience and knobs I also made to the FB Theremania community. Curios about the reactions)

The minimal knobs I would see until now are 4:
1 Loudness to a low ohmic output -- potentiometer
2 pitch -- variable air condensor
3 volume -- potentiometer 
4 presets of timbres -- rotary switch

outside: delay, feedback, loop, guitar amplifier and/or amplifier with aux input. 

I'm curious: Should a built theremin either use a rotary switch for certain registers such as cello, voice, flute or a continuous knob for the timbre, as is the case in most theremins, or both? What wishes and experiences are there?

I really like the Hammond tonewheel system of producing sounds - where you use drawbars to add the desired amount of a variety of harmonics, e.g. 25% of a note a fifth lower, 10% of a note 1 octave lower. So if I was designing something, that's what I'd do nine drawbars. Once you have those, you can make any waveform you want.
 Here's a diagram (just read the "Scale Interval" row)

https://i.gyazo.com/aca1072a380cc53d2e28d99aef57ad63.png