theremin sounds like a cello

http://soundcloud.com/invisiblejelly/cello-sound-from-diodes

this one is not so furry not much buzz either ie a more clean sound cello like from just messing with diodes in the right places.This time I played or slightly murdered a tune and then went to sleep.

Invisiblejelly,

I know this is an ancient thread - I missed it when I was absent from school, LOL!

>>"But because I'm an experimenter and only have one EW theremin at the moment I took it out and found I could get a bass sound similar but not as good as the module by replacing C6 and C2 with two diodes( IN4004's) used as capacitors this proves that a diode is not the same as a capacitor of similar value.This started me on playing with diodes to see how it altered the sound by having mixed diodes in parralell as capacitors.Every diode I added altered the sound a little bit so I ended up taking out C6 (but leaving in C2) and I ended up adding 9 extra diodes till I got a sound I liked. "<<

I believe the reason for the tone change you get from using diodes for coupling instead of capacitors, is that reverse biased diodes only act as fixed value capacitors if the voltage across them is constant - Where you put these diodes, the voltage certainly is not constant!

 The waveform (pre mixer) is being distorted because the diodes capacitance is changing dynamically - the AC (oscillator) waveform is centred at about +12V (at which point there is 12V reverse biasing the diodes) and goes up to about +24V (Diode capacitance is now at minimum) and down to about 0V (at which point the diodes are hardly reverse biased at all, giving maximum capacitance -

(the above voltages are based on simulations, but will be more-or-less what the circuit gives I think) 

This dynamic, cycle-by-cycle change in the capacitive coupling to the mixer, will alter the harmonics in ways that a simple capacitive coupling cannot do - these extra harmonics will mix altering the shape of the audio output (difference) signal.

I have done a lot of playing with pre-mixer waveshaping (as in, changing the waveforms of the oscillators prior to mixing them) and, to my ears, the results are far more pleasing than if one adds distortion after the difference frequency has been extracted. Your diode modification is effectively producing pre-mixer waveshaping.

"Out of that these diodes will not act as diodes because they will be reversed-biased all the time. Thus I wonder what could make a difference there?" - Thierry

Thierry, I was a little bit puzzled by your  "Thus I wonder what could make a difference" - We discussed the changing capacitance of diodes subjected to oscillator signals several times in past threads.

Fred.

"I do have an ESPE01 module " IJ

My understanding is that the ESPE01 consists of a pair of transistor buffers (one for each oscillator) and duplication of the mixer stage - I have never seen a ESP01 or its schematic, but this is my guess.

If you could insert your diode trick before (at) the mixer on the ESP01 (as in, I suspect there are 2 capacitors coupling the buffers to the mixer - remove these capacitors and replace them with diodes) you may have the best of both worlds - oscillator isolation giving extended bass range, and waveshaping from changing diode capacitance.

Best to check with Thierry first though - as I say, I am only guessing about the ESP01 - And modifying the module will probably invalidate any warranty, and/or lead to an irate French Theremin engineer cursing me!

The other thing you might play with, would be to put a small capacitor (say 2p2 to 4p7) in parrallel with the diodes - whatever C you add, will not change dynamically - so by messing about with diode + capacitor variations, you should be able to (more) finely tune the harmonics.

Fred.

Fred  "Your diode modification is effectively producing pre-mixer waveshaping."

Invisiblejelly "Yes I could tell that from listening to the sound and by adding one diode at a time in parallel as capacitors and hearing the changes one diode at a time and also mixing different diode types.

Fred "The other thing you might play with, would be to put a small capacitor (say 2p2 to 4p7) in parrallel with the diodes - whatever C you add, will not change dynamically - so by messing about with diode + capacitor variations, you should be able to (more) finely tune the harmonics."

Invisiblejelly"Thanks Fred for your suggestion I'll definitely try that when I have more time.

I have run some simulations - Not great- problems with modeling the diodes - but some tentitive suggestions.. Hopefully I will sort out a fully working simulation soon, and provide a link.

1.) Simulation showed the diodes being forward biased briefly, as my oscillator waveform dropped to -3V.. Not sure how "real" this is, but the resulting clipping might be the 'buzz' you are speaking about.

2.) You may find that putting a series capacitor between the oscillator and diodes - perhaps something like 47pF, will improve matters - If the diodes are forward biased at ant time, there will be a DC loading on the oscillator through these diodes and the mixer resistors - this could cause serious distortion (?).. well - its enough to make my simulation crash!

3.) because I cannot get the simulation to show me what happens when the diodes are forward biased, I cannot know how musical this distortion will be.

4.) My guess is that adding the capacitor (2) will reduce the buzz harmonics greatly, but that you will be able to increase this in a more pleasant form if you duplicate the modification on the reference oscillator (as I understand it, you have coupled the reference osc via a 15p capacitor?)

Fred.

Hello Fred ,I can control the buzz in two ways (1) via the graphic equalizer on the amp..(2) also if I turn the brightness control to the right I get more buzz than if I turn to to the left..my prefered settings are waveform all the way turned to the right and brightness turned all the way left.

I replaced both 15p caps with diodes only one 1004 on the reference oscillator and I have used about 37 or so(current amount of 4148's as replacement for the 15p cap on that side.I tried many different arrangements) in parrallel on the other oscillator which seems like a bit of a silly thing but I kept adding one diode at a time to get the sound I liked..plus I put a couple of extra diodes connected to D4.I haven't encounted any serious or dangerous distortion from this setup I've been playing my theremin with this setup for months now and am enjoying the nice bass and cello like tones everyday.

Romain, first of all: you have a lot of courage, this cello suite from Bach is technically almost unplayable for the best thereminists in the world.

Two things for your practice. The left hand dynamic gives shapes more like a bassoon than a cello. Your track is the upper one. The other is a professional cello player.

I do not own an etherwave. But the setting used by you give too few overtones (brightness). I suppose the etherwave is there strong limited. Your track is also the upper one, now in a time-frequency diagram. You are a bit slower, what is a good idea.