Superior to the Ethervox !? TWO

Fred: If you still have copy of these photos, I would love to have them!

**********************

I would be happy to send you all the links to the photos. The jpeg files are HUGE and can be magnified to the point where a small transistor will fill your entire computer screen without losing definition. I am sending you a private email with the links to the site where the photos are archived.

Thierry mentioned that he had hoped his familiarity with the E'Pro circuitry would help him understand the E'Vox. Apparently it didn't.

[i]”my personal opinion is that the ethical thing is to release it. What would be unethical would be to have it under lock and key. It would be a pity that people were prevented from producing better musical instruments for being similar to existing ones” -AlKhwarizmi [/i]

I agree with the above - I think I failed to accurately describe my "ethical dillemma".

My method of Voltage-controlling 'theremins'(VCT) can be adapted to any Theremin.. I have my own Theremin designs which I have applied this to - and I personally prefer these.

My reason for including an EW VCT would be to capitalize on the fact that this Theremin is so well established and known (and liked).. But it is Bob's design, not mine! The things which give it its appeal and its tonal qualities (the way the signal is distorted, the change in waveform / filtering as a function of volume, etc) are things I would not have done on my designs..

The improvement that putting a Theremin into a 'closed loop' so that it is locked to an incoming audio frequency gives is that one does not have any problem with the HF oscillators going into lock, or (worse) stuttering between being in lock and out of lock - There is a tricky balance with 'normal' Theremins.. One wants the oscillators to interact with each other so that the waveshape changes as a function of difference frequency, and one wants them to interact so that they 'pull' each other and therebye linearize the field at the bass end.. But one also wants them to stay unlocked in the audio range down to as low as possible.

By making the front-end entirely seperate from the tone source, and by driving the VCT in this manner, all the above (except for field linearization - which is not required as this is taken care of by the front-end) are achieved.

No modification is actually made to the EW-VCT to make it 'sound better' - This happens simply as a result of the VCT being in a closed-loop (and this improvement applies to all VCT circuits).

So the ethics really comes down to the fact that I dont need to use the EW circuit - And that I would need to advertise these as EW 'clones' for people to know what they were buying.

Perhaps I should contact Moog and at least let them know what I am planning.

Fred.

[i]I have my own Theremin designs which I have applied this to - and I personally prefer these.[/i]

And so does Nick Franglen, as I recall. I haven't heard enough to form an opinion, but generally my methodology is of the "this is what I have now let's see what I can do with it" camp rather than "this is what I have in mind now how can I achieve it" so my priority is to have a good range of options rather than a specific set of timbres. I do like the etherwave sound, but I'm not married to it.

Some points.

1. Fred, which do you trust better, your ears or your ability to evaluate market preferences? In every other aspect of your designs you are innovating rather than following trends...

2. There are also plenty of people who are not keen on the etherwave sound.

3. A frequency tracking beat frequency oscillator is not a theremin, it is a modular synthesizer/theremin module. This is something thats natural environment is one where the waveform it outputs is a starting point for further audio processing.

4. I suspect what will sell the instrument is hearing a great thereminist play it.

Thank you, Gordon - What you say makes a lot of sense.. Particularly "which do you trust better, your ears or your ability to evaluate market preferences?" - LOL! Nail on the head!

There were other reasons that I was looking closely at copying the EW though - The EW is an extremely simple circuit by comparison to mine.. Even my simplest (The South-Bank H1) has multiple sound / harmonic sources into a mixer (sub and 'natural' shapable square waves, natural analogue heterodyne waveform with simple shaping control - 5 controls in all) and the mixer has an adjustable filter and adjustable soft clipping (compression).. So it ends up being quite a monster! (it only looked simple because all the controls were on presets, and not accessible to the public).

My other reason for thinking "EW" was that I know there were people who loved the EW-Pro's range switching and linearity, but did not like its sound - I have met a couple of people who would love to be able to have the E-Pro features on their EW.. and I could provide this..

So maybe - just maybe.. I will be selling modules with which a CV outputting Theremin can be built (and all the other data / modules required for taking this to whatever level the constructor chooses) before I put complete instruments on the market (I would, in fact, like to sell these modules to OEM's, and never need to put complete instruments together - let someone else who has space / a factory do that! ;-) .. I could add a module to allow people to convert their EW to a voltage controlled unit - this way all possible ethical implications are overcome (I will even need to buy a damn EW in order to get the module / kit right! )-8 ..

For now though, I will stick with standard VCO/VCF modules as the basic sound generator, and a variant on the H1 as the VCT option.

BTW - My reference to the VCT tracking audio was misleading.. It is not possible to feed a tone into the VCT and get it to track the pitch - The CV into the VCT drives an extremely strange circuit with PLLs, a dedicated (custom) mixed-signal processor, and lots of other gobbldygook ;-) .. it is the most wierd, complex, time consuming circuit I have ever built.. Yes, what it tracks is "audio" - but not audio as we know it, Jim! .. If it was tracking audio, it would fall under the limitations of the "laws of physics" [i](as in, it takes 50ms to get a single sample of a 20Hz waveform, so one is limited to frequencies above about 100Hz - the same problem all pitch to voltage converters struggle with)[/i] and as Scotty says...

Fred.

[i]Thierry mentioned that he had hoped his familiarity with the E'Pro circuitry would help him understand the E'Vox. Apparently it didn't. [/i]

Sorry for the late reply, but I'm actually overloaded in my day job and with some rather personal tasks, so that I cannot spend as much time with theremin related subjects as I would like.

I hoped being able to discover the principle of operation of the non-midi part of the Ethervox by finding similarities with the EPro (which I know well in the meantime) and with the series 91 (which I don't yet know as well). What I understood up to now is that Dr. R. Moog had basically 2 different structures for each function block of a theremin which he combined in several permutations and at different quality levels into the E-Standard, the Series 91, the Ethervox and the E-Pro.

But this knowledge is not (yet) enough in order to understand the Ethervox (especially because I never could put my hands on it), because the oscillators are (I could see that thanks to Peter's pictures) closer to those of the Series 91 while those of the EPro are rather improved EStandard oscillators.

I will have the occasion to work a lot on a series 91 in December or January and as soon as I will have acquired more insight in this technology (which is very different from the EPro and the EStandard) I will share it here.

Thierry,

First, I hope things went well at Lippstadt, hope you get a chance to tell us about it!

My primary reason for getting "into" the E-Vox is that I will be attempting to fix one on 2nd Dec, and having some advance knowledge is desirable when one is faced with repairing an unknown instrument, hopefully quickly. (not sure about the grammar of the last, LOL! ;-)

Frem Peter's Photo's, I have now drafted unverified but sensible looking schematics of the pitch and reference oscillators, range-switching dividers, and mixer - next I will attempt to verify that what I believe is an F to V is, in fact, an F to V, and to attempt to draft a schematic for this, and for ascociated circuitry. My aim is to get as much of this done prior to 2nd Dec, as I believe this is the section where the fault is most likely to be.. When I have the instrument, I should be able to quickly verify the sections of the circuit I have drafted.

I plan to publish any schematics I derive, on Element-14, together with any notes / waveforms etc.. I intend to open an E-Vox 'project' page there, which would be an on-going 'centre' where anyone with insights into this instrument, or any documentation of any kind, can publish.

I would love you to open a '91 series 'project' on E-14 when you get into this instrument - I think the two projects could complement each other well!

Fred.

I have now spent a day examining the E-Vox photos - I am not *sure* about anything really (even with these excellent high resolution photos, one cannot see tracks running under the IC's or be sure of connections running between the legs of the IC's).. but..

What I thought was range-switching circuitry is, I now believe, the wierdest 'mixer' I have ever seen! I do not believe this is a heterodyning Theremin at all - (I may be wrong!) - I think that the reference and variable oscillators independently act on Cmos switches in a way slightly similar to that used in the EW Pitch to Voltage converter (but at much higher frequency, therebye overcoming the F to V problems at low frequency) and that the front-end, which at first glance looks almost conventional, is a capacitance to voltage converter.

Range switching is achieved by simple summation of voltages which are added to the voltage from the front-end.

I have located what looks like two 'tone generator' blocks - These, I believe, consist of a VCO and a VCF each.. They are each identical and are built with a pair of LM13600 Dual OTA's - One looks like a VCO, the other looks like it is configured as a LPVCF.

Voltages (from the Front-end and from the MIDI-CV converter) appear to be routed to these tone generator blocks.

There is a circuit block that has a quad transistor array, and I suspect this is an exponential converter - but again, I am not sure - I have not been able to trace actual connections of voltage paths, just been able to see 'groupings' of traces from blocks which guide me in my hypotheses.

I have more questions almost, now that I have come this far, than I had when I started. Much of the circuitry is quite crude, the oscillators (which I have the most complete schematics of - almost certainly 95+% accurate) are simple single transistor designs, much less sophisticated than the standard EW oscillators - I do not believe that they, on their own, could give the stability that thev E-Vox is, I believe, renowned for... It MAY be that the nature of the 'mixer' behaves in a way where drift (which is likely to affect both variable and reference oscillators similarly) is cancelled out.

I believe that the good linearity is achieved by the unusual method of tuning the Theremin - Onlike most other designs I have seen, tuning is done on the variable oscillator. A variable capacitor sits across one winding on the Equalization coil (not directly on the tank) - On first seeing this, I thought there would be an adverse effect on linearization by doing this - but I now realize that, if the matching of the oscillator frequencies are carefully done (there are no variable inductors on this design, but there are 'pads' on the oscillator circuits on which 3 capacitors are placed in parallel - I suspect these are selected on test) and if the inductors are extremely high quality, then tuning in this way will simultaneously adjust frequency and linearity to compensate for background changes in capacitance.

At this time, I am both highly impressed and also quite unimpressed - This instrument had / has so much more potential than was realized.. A few jack sockets giving CV outputs would have extended it usage.. CV tracking the full audio range without the problems one gets from the EW+ at low frequency. I also do not understand why, after having made a good CV outputting Theremin, Moog reverted back to inferior technology for the EW Pitch-CV. I believed that my method of getting full-audio-range CV from a Theremin was unique and original (and it may yet be).. But Bob managed to do this by (I believe) a slightly different method, Years ahead of me! Why has this been hidden from the world? I feel this is so sad.

Fred.

As far as I understood, the signal from the smaller tapped/tuned tank circuit gives a voltage which is proportional to the resonance current of the "big" tank circuit. This is (at least in the Series 91) rectified and used to drive a VCO.

[i]"As far as I understood, the signal from the smaller tapped/tuned tank circuit gives a voltage which is proportional to the resonance current of the "big" tank circuit. This is (at least in the Series 91) rectified and used to drive a VCO." - Thierry[/i]

Most interesting, Thierry - Your description tally's with what I was told by another Theremin designer - that the 91's pitch topology was much the same as the standard volume voltage generating topology.. As in, there is no reference oscillator.

The E-Vox does not use this scheme - if you look at the photo of the board, you will see that the VFO is on one corner, closest to the EQ coil, and the reference oscillator is on the other corner - an identical circuit with heatsinked 2n3019 transistor.. The outputs from these oscillators are coupled by a capacitor on the edge of the board (just below the VFO transistor) and the visible track running on the edge of the board carries VFO output to this capacitor on one side, and Ref Osc output to the other side of this capacitor.. There are a pair of resistors running down from each side of this capacitor forming a potential divider / filter with capacitors + resistors going to 0V. The oscillators are not combined here - and I will need to probe this area to confirm what is going on - But I suspect that both VFO and REF signals are being converted to 'logic' level (or at least being current limited to ensure that the IC inputs can clamp the levels safely) and being used to switch the dividers etc to produce a (DC) voltage proportional to frequency difference between the oscillators.

But the above, at present, is at least 40% speculation.

I have tried the resonant "volume" circuit for pitch CV generation, and never managed to get good results from this topology.. It now looks like there are quite a few different schemes used by Bob Moog:

Standard good quality oscillators and analogue heterodyning, as with the EW;

Extremely high quality oscillators followed by digital range switching and mixing, on the E-Pro;
Resonant "volume topology" voltage outputting, VCO sound generation on the 91's;

VFO and reference oscillators being 'mixed' in some as yet unconfirmed way, and generating a CV which drives VCO/VCF tone generators and also a MIDI controller.. and has a MIDI to CV board which can feed these VCO/VCF tone generators.. Fixed 'span' with register selection implemented by simply adding a voltage to shift the frequency - the E-Vox.

I think Bob just liked playing with Theremin designs! ;-)

Fred.

PS - I think that the VFO and Reference oscillators may be maintained at similar thermal operating points due to how they are driven.. The transistors are driven hard, and dissipate quite a lot of heat (far more than is required for them to perform their function).. I suspect that, by operating above ambient, and by their dissipation being constant, and the layout of each circuit being (thermally) similar, they will be less subject to differences in temperature between them than if they were driven at low current where local thermal effects would cause a greater temperature difference between them. I have often wondered why transistors in some designs are driven harder than seems reasonable - perhaps this is the reason.

I have created a seperate thread for the "reverse engineering" discussion here (http://www.thereminworld.com/forum.asp?F=3&T=4706&cmd=p&p=1)

And have posted the Photos Here (http://www.element-14.com/community/groups/theremin-general-resources?view=documents)

Fred.