Kit choice

[i]"Would I be way out of line in thinking that it would be relatively straight forward to alter an etherwave to have a similar sine/square control, and probably even retain the option of the moog's "signature" sound?"[/i]

Depends what you mean by "relatively straight forward".. If you are able to understand schematics and construct quite simple circuits, then it is "relatively straight forward".. otherwise its not! All the stuff I discuss below is available on synth sites, and one can buy cheap PCB's to implement these ideas..

If you have an EW+ (or EW with CV outputs)the above might be quite simple.. You could take the audio pre-VCA (the "preview" output) and square this with a comparator or over-driven Op-Amp.. take the original audio and the new square output (via suitable resistive attenuator) to a simple X-fade mixer, so one can adjust the waveform from 'original' to 'square' with a single potentiometer..

Now, feed this X-fade mixer output to a VCF (LM13700 standard synth LP VCF - Loads of schematics on synth construction sites.. Example (http://nutsvolts.texterity.com/nutsvolts/200603/?pg=47)) and take the pitch CV to this.. Tuning this VCF will controll the roll-off frequency, and one can get sine from the original simply by rolling off all the higher harmonics - The VCF will track the pitch, so the waveform will be constant over the pitch range.

Then take the VCF output to a VCA (another LM13700 - or the other half of the LM13700 used for the VCF, if a single stage VCF is employed) and use the volume CV to contol this.

The above does work - but has problems at the low frequency side, because the Pitch to Voltage converter jitters below about 200Hz. You can overcome this by clamping the CV into the VCF so that below 200Hz the CV is clamped at 200Hz equivalent.. This means that the filter will not track below 200Hz and the harmonic content will become greater (sound becomes 'brighter').. This scheme is no use for VCO's but actually sounds quite pleasant (to my ears) when applied to a VCF.

Filter tracking is not nearly as critical as oscillator (VCO) tracking - even being a few semitones 'out' is inaudible unless one has the Q set high (one does not want this - one wants absolutely no 'peaking' at the roll-off frequency - just a nice, smooth, 12db/octave maximum rolloff).

GOlly, Gosh, Fred!
You are the new 'M'!
teslatheremin

[i]Depends what you mean by "relatively straight forward".. If you are able to understand schematics and construct quite simple circuits, then it is "relatively straight forward".. otherwise its not! All the stuff I discuss below is available on synth sites, and one can buy cheap PCB's to implement these ideas..

If you have an EW+ (or EW with CV outputs)the above might be quite simple.. You could take the audio pre-VCA (the "preview" output) and square this with a comparator or over-driven Op-Amp.. take the original audio and the new square output (via suitable resistive attenuator) to a simple X-fade mixer, so one can adjust the waveform from 'original' to 'square' with a single potentiometer..[/i]

That's about what i'd had in mind with some comparator squaring. I hadn't noticed that the pitch preview was pre vca, so that ought to make things at least a little simpler. I've only quickly loked at schems... exam time means that any time on here or looking at schematics is procrastinating.

[i]Now, feed this X-fade mixer output to a VCF (LM13700 standard synth LP VCF - Loads of schematics on synth construction sites.. Example) and take the pitch CV to this.. Tuning this VCF will controll the roll-off frequency, and one can get sine from the original simply by rolling off all the higher harmonics - The VCF will track the pitch, so the waveform will be constant over the pitch range.

Then take the VCF output to a VCA (another LM13700 - or the other half of the LM13700 used for the VCF, if a single stage VCF is employed) and use the volume CV to contol this.

The above does work - but has problems at the low frequency side, because the Pitch to Voltage converter jitters below about 200Hz. You can overcome this by clamping the CV into the VCF so that below 200Hz the CV is clamped at 200Hz equivalent.. This means that the filter will not track below 200Hz and the harmonic content will become greater (sound becomes 'brighter').. This scheme is no use for VCO's but actually sounds quite pleasant (to my ears) when applied to a VCF. [/i]

I'd been thinking take the pre VCA signal, blend that between square/sine, and have a switch between the mod and the original ouput, but yours seems simpler and probably more functional... my (minimal) electronics experience isn't really with synths, so 'm still getting used to the idea of having relatively stable pitch tacking and possible CVs everywhere... seems like there's way more possibliities than I can imagine....

Thanks for taking the time to reply (and everyone else that replied as well!), its great to get some input on this from those with more experience. I'm really thinking an etherwave is the way to go for me, so when I get around to pulling the trigger, that's probly the way I'll go.

[i]"GOlly, Gosh, Fred! You are the new 'M'!" - teslatheremin[/i]

Alas, there will never be another 'M' .. Thanks for the great complement, but I am nowhere near the great 'M' - I just feel honored to have shared existance in the same 'time slot' as him, and to have benefited from his innovations (and willingness to share these) and the instruments he created.

[i]"seems like there's way more possibliities than I can imagine"[/i]

LOL! - This is the biggest problem when one has voltage controlled modules and some CV's.. the possibliities are huge - and deciding what to include and what to leave out is what drives one nuts!

For example.. Do you ONLY want the VCF to track the pitch CV 'correctly'? - Having the tracking variable adds a whole dimension of 'dynamic' tone variation .. Some of the smaller Moog synths give a VCF tracking switch with "Off" "Half" and "Full" tracking - but I prefer a potentiometer going from 0 to 2 (0=Off, 1=Full, 2= 'overshoot'x2)..

Then, do you want to allow the volume CV to affect the tone? Add another potentiometer and you can do this..

With a Theremin, that's probably all one would want - with a synth (or a Theremin driving a more complex set of VC modules) one ends up with a panel full of knobs to cater for even a few of the possibliities.

[i]"I'd been thinking take the pre VCA signal, blend that between square/sine, and have a switch between the mod and the original ouput, but yours seems simpler and probably more functional"[/i]

The original audio output (unmodified) will still be available.. If you took the output from the new VCF/VCA to a seperate output socket, then you could choose which to connect - or take both to a mixer - or even have a stereo output!

Effectively, one can add this modification without altering the basic EW at all.. You could even have it as a seperate box - just have a connector carrying +12V,-12V,0V,Preview Audio,Pitch CV,Volume CV.

[i]I'm not sure that one will be able to get a more sine-y tone from the etherwave. While the oscillators are a very good and stable design, the mixer section and the post-mixer sound processing are rather poor.[/i]

I'm not sure I would call the design of the mixer/processing "poor". It seems a very efficient way to get a very traditional Theremin sound. The distortion created from the volume circuit is especially effective in this regard.

I agree that it would be difficult to get a straight sine wave with the Ethewave. (Let us keep in mind, however, that the original RCA/Theremin sound is hardly a straight sine wave, and is actually more similar to the distorted wave produced by the Etherwave).

However, it certainly would be easy to change the tone of the Etherwave by adding a variable control to the rolloff frequencies - I have been considering doing this to my new E+. One could also change values of the input section of the VCA to get a much truer sinus wave. Both of these would be easy to accomplish and get a tone very close to, say, the Theremax sine wave. At the same time, I think doing so would rob the Etherwave of it's rich sound.

Etherspiel, you are right, "poor" is the false word, but I could not find better with my restricted English vocabulary.

What I wanted to express is my personal opinion: If distortion is wanted, it should be naturally given. But if it's unwanted (i.e. by people who do not like the RCA sound but prefer the 1994 A. Henk prototype sound) it should not be given.

So a good circuit design (in my eyes) should produce no distortion/harmonics at all by default with the opportunity to add some, only if wanted (kind of additive synthesis). Later filtering is only curing symptoms.

Although I like the pitch and volume response of the Etherwave very much, its wave shaping circuit gives me only the choice between pest and cholera...

[i]Thierry wrote: choice between pest and cholera[/i]

one of the many things i like about thierry's posts, is that even if they are done with "limited english vocabulary," the words that he does choose are pretty colorful!

Ha! I agree, Thierry - I adore your use of english, and for the record, your command of english is certainly better than my french!

How is your work with the Theremax coming along? That should be giving you the pure sine wave you are seeking?

[i]"So a good circuit design (in my eyes) should produce no distortion/harmonics at all by default with the opportunity to add some, only if wanted (kind of additive synthesis). Later filtering is only curing symptoms."[/i]

There are some real, practical problems with this approach - and, in my opinion, the effort and expense required to maintain 'perfect' sine waves from the oscillators is not justified.. Yes, it is reasonably simple to apply controlled distortion to the oscillator waveform to add harmonics pre-heterodyning - If one is designing the Theremin.. But when one is talking about modification to an existing design (such as an EW) an entire re-build of the front-end (oscillators,controlled distortion,mixer) would be required.. one may as well design / build the Theremin from scratch.

With a tracking VCF, "filtering is only curing symptoms" is not relevant - if one wants a sine wave, you can get a sine wave (and a far better sine wave than you could get from redesigning the oscillators) and one also has tight control of the amplitude of harmonics if you want to let them through.

The above is subtractive synthesis, and is the way that all classic analogue synthesisers work.. Start with a harmonically rich waveform, and get rid of the harmonics you dont want.. Because the filter is tracking the fundamental (via the pitch CV), the harmonic spectrum remains constant over the pitch range.. It is simple and effective, and one can get sine waves of extremely high purity. When one talks of a VCF, people often jump to the conclusion that they give the classic 'Moog sound' - but this is only because the VCF Q ("resonance") is set high, giving a peak at the cut-off.. When set up purely as a filter one does not even hear that there is a VCF in circuit.

I believe that pre-mixer waveshaping is the better route as discussed here "how do I distort thee" (http://www.thereminworld.com/forum.asp?cmd=p&T=3906&F=1) and that having adjustable filter/s pre-mixer would give the best of all worlds..

But for those who want a better tone pallette from the instruments they own, or those who do not want to (or cannot)wait for someone to bring a better (affordable) Theremin to market, I believe that additional post-mixer waveshaping and a tracking VCF is the answer. I also believe that both the E-Pro and the T-Vox employ a post-mixer VCF for tone smoothing.

Electronic design is all about balancing requirements - call it compromise if you wish - and I do not think that starting with 'pure' sine oscillators would be a good choice.. In fact, mixable square and ramp waveforms fed into pre-mixer high frequency filters allows the mixer output waveform to be adjusted from 'pure' sine to a large range of harmonic palletts.