Modifying the tone of the Etherwave Standard

I think I get what you mean. Let me explain it back to you and see if you agree.

Starting at the zero beat zone and moving towards the antenna we first encounter the "prelock" or "Geiger counter" zone, where we hear rapid clicks. As these become faster they take on a raspy or gravely character which diminishes as the pitch rises. One can achieve a similar effect with the voice (I used to do this as a kid to creep out my parents - they were adamant that it would hurt my throat) or by belching.

I don't normally notice this as (1) I usually play with some reverb added, which diminishes the effect a little and (2) I don't consider the very low notes to be part of the playable range of the etherwave. (Or when I do use them it is in the expectation that this is what it will sound like.)

For comparison, here's Clara playing in the lower range of her instrument. It's still a bit raspy, but it's a nicer rasp.

http://www.youtube.com/watch?v=jEl4zCQBv2c

Carrice,

I think that what you are hearing at the low end, is the increased harmonics due to the oscillators starting to synchronise.. This gives a 'pulse' effect which is extremely (in my opinion) unpleasant.. Trouble is that, as it is on the LF side, it is also extremely difficult to fix... You may find that the low pass filter will help a little, also you may find that a tiny bit of reverb will help to blur the 'dur dur' and make it smoother.. But yes - it is horrible - and all the EW's do it.

[i]"This is why I originally asked the community at large about different "higher grade" theremins, I feel like the EW is a beginners instrument. But despite my best efforts it seems "get a degree in electrical engineering" and "play a moog" are my only options, as no, no one produces better theremins that a musician can buy."[/i]

I agree with all the above - I get private emails almost daily from people requesting features from the Theremin I am designing, and the line [i]"no one produces better theremins that a musician can buy."[/i] is echoed in the majority of these.. It is one of the reasons I am taking so long to bring anything to market.

I have now "played" a few theremins [i](far too few - one of my biggest mistakes was not going out and really getting familiar with the different models before I started.. Back at that time I would have been far wiser to spend some money buying Theremins rather than wasting it on projects I abandoned because they turned out to be crap)[/i] and the one with the nicest tone (to my ears) was GordonC's Enkelaar - The over-all winner (and I am an absolute beginer when it comes to playing Theremins, so please take my opinion with an EXTREMELY large "pinch of salt") in terms of sound and playability, was Lydia Kavina's Tvox.

I hated the EW standard's I have played, and was not impressed by the EW-Pro, even though I had expected that I would be.

I believe that a Theremin which overcomes all the "problems" (range, linearity,sensing distance,interference,reliability,sound quality,waveshaping options,correct CV tracking to name a few) is not only possible, but realisable - and on its way.. Even the most advanced Theremins ever made are quite cheap in terms of electronics (valve Theremins, because of the cost of the parts, are justifiably more expensive).. Whilst I believe I have solved the "problems" above (at the moment I cannot be sure about reliability, as it is protoptype phase.. I am aiming for highest possible reliability, but cannot make any claims for that now) I have created one big problem - the component count and cost of my Theremin is at least 4x that of any other Theremin - including the Ethervox.. My first priority is to get a fully running version of this Theremin, Then I must concentrate on optimising the design to reduce the cost, without compromising the quality, so that i can put it to market at an affordable price.

Back to possible modifications..
And here I must remind you that I am NOT an EW expert..

As I see it, there are a few routes for oscillator synchronisation on the EW.. one of these (which I think will be the lesser) is inductive coupling via the coils, the cther is capacitive coupling (between tracks) - But the biggest is the HORRIBLE way that the oscillator signals are fed to the HORRIBLE mixer.

C2 and C6 (outputs from reference and variable oscillators) are connected together - this gives a direct path for oscillator synchronisation.

I think the best improvement would be achieved by replacing the whole mixer circuit with a LM633 IC .. but this is a lot of work.. I wonder whether..

Replace C2 with two 33pF capacitors in series -
to the centre point connect a 10k resistor, the other end of which is taken to ground.

do exactly the same to C6.

You could also try connecting a small capacitor (10pF to 22pF) in parrallel with the 10k resistors.

This should reduce oscillator coupling, and if you add the parrallel capacitors should reduce coupling even more and make the waveforms more sine like..

worth a try?

I have created a page here with schematics, to reduce the possibility of confusion / misunderstanding.

Fig 4 shows the latest suggestion - it wont cause any damage, but I cannot predict its effectiveness.. In particular, I have not done any calculations or simulations on the effect reduced amplitude will have on the diode mixer.. it might make it perform better, it might make it worse, it might make no difference - or it might not work at all.

link to page is here (http://www.therasynth.com/html/ewmod.html)

Can anyone tell me what the ACTUAL amplitude of the waveforms from the reference and variable oscillators of an ACTUAL EW Standard are?

My experiments with EW circuits have never been done using genuine EW parts .. so I would like to confirm what the real thing does..

AAARRRGGGHHHH!

Just seen that my latest mod could seriously detune the oscillators.. Adding capacitance / resistance from the mid-point of C2 and C6 modification is similar to adding these to the antenna circuit..

I dont think the latest modification (Fig 4) will work .. it should work if one added a series resistance in line with the modification, so that the impedence 'seen' by the oscillator is unchanged.. But this impedence will alter for every value one experiments with, and, to keep the oscillators at their correct tuning, you would need to calculate a new series resistance for each change... And there would, with a network like this, be the destinct chance of forming a bothersome tuned circuit from this network..

I will leave the page for now - but Fig 4 is most likely not worth trying.

Everything else I can think of requires active components (FET voltage followers to buffer the oscillator signals and isolate them from each other, and a simple variable coupling so one could give the oscillators some deliberate, controlled coupling if you found you needed it, would be the easiest next step).. But if you are going to add active components, I would add a LM633 and possibly solve all the problems -

I cannot help any more this month - Sorry!

If you can wait until the weekend I'm ready to open mine and connect the oscilloscope (with a 1:10 probe ;-) ).

Why not change the values of C24 and/or R31. It seems to me that these two would have a more pronounced effect on the tone than C23. I would think replacing R31 with a rheostat type configuration would give a control that would allow the sound to change from a very harmonic-rich sound to a very hollow timbre.

[i]"If you can wait until the weekend I'm ready to open mine and connect the oscilloscope (with a 1:10 probe ;-) )." - Thierry [/i]

Thanks Thierry - Yep, 1:10 would be a good idea .. Even then, the capacitance still causes problems if one was wanting to keep the frequency stable.

[i]"Why not change the values of C24 and/or R31. It seems to me that these two would have a more pronounced effect on the tone than C23. I would think replacing R31 with a rheostat type configuration would give a control that would allow the sound to change from a very harmonic-rich sound to a very hollow timbre." - Etherspiel[/i]

Good idea! I was focussing on Pre-OTA (VCA) because Clarice made it clear that only sine waves were wanted.. I thought that lowering the levels prior to the OTA would reduce subsequent distortion produced by the OTA.. But there is no reason why C24 should not be increased - either as a possible alternative modification, or in addition to the modification of C23.
[b] Edit -> [/b] Changing C24 would, I think, also give a volume dependent roll-off (as volume increaced - due to increase in Iabc - the R*C of the filter would change) There are other filters which could be changed without this 'complication' C26 for example could be increased [b] <- End of Edit [/b]

However, I now think it more likely that the problem relates to oscillator coupling - This is certainly the case in the example "dur dur" at the low end.

Do you have any ideas about the wierd volume control effect from the brightness control?

Fred -

Yes, I think you are correct. The Etherwave already produces a volume-dependend roll-off just like the RCA did. I particularly like this feature of the etherwave circuit, the "blooming" or "throaty" effect when the volume is increased.

Another possibility would be to change the values of C26 and R34 - these might be the best choice for post VCA processing. Again a rheostat type control to replace R34 would have a very nice effect of controlling the harmonics, no?