New Solid State Keppinger Design Theremin

Mr. Mark Keppinger has informed me that his all solid state design/development work is nearly complete, and his given me permission to foward this information. More details to follow soon, but the design is not a FET adaptation of his renowned tube model, but a completely different approach, similar to a Big Briar Model 91 in some ways.

Philip Neidlinger

That's great news! Mr. Keppinger sent me a copy of his initial schematic back in January of 2007. It would be interestng to see how it has evolved since then. I decided not to say anything about it so he would not feel any unnecessary pressure from any of us.

Wow! amazing news, thanks for keep the hopes to all of we doesnt have a Epro, the 2009 will be very interesting, ill separate some bills to be preparated.

Jeff

If it was Jan 2007, Mark likely abandoned that approach and started over, per his notes. I'll post a summary when I can soon.

PAN

Here's some information. PLEASE COMMENT. Mark has not finalized his prototype and is very open to feedback at this point.

"I've retained only the volume and pitch oscillators; everything else is completely different. In a traditional Theremin, the designer is restricted by electrical theory as to certain aspects that control linearity, etc. So I decided that instead of beating two oscillators together, I would set up one
oscillator for each the pitch and volume with a corresponding antenna and a loading detector circuit that would give me a DC voltage proportional to the hand position; nominally +2 to +8 VDC, with the higher voltage being present when the hand is furthest away, and lowest when the hand is closest. Each of these oscillators is on their own circuit board, has varactor ("vari-cap") tuning with an onboard potentiometer, and also an indicator LED on the circuit board that is brightest when the circuit is tuned to resonance (which is what you want). The adjustment will normally be "Set-and-forget", since there are other controls for the pitch / volume responses; more on that later.

The 'main board' contains everything else up to and including line out / pitch preview out. The pitch and volume circuits each contain two adjustments for hand positioning; in the good ol' op-amp days, we would have called these "GAIN" and "OFFSET".I'm not really sure what to call them in this case; that's one of the decisions you get to make. Anyway, for a given hand position range, let's say 24" to 12" from the antenna, "GAIN" would increase the range from, for example, one octave to one and a half octaves. The "OFFSET" would shift at one to one and a half octaves either up or down,
depending on which way you turned the control. Does this make any sense? This might be easier to show with a graph.

Okay, this is where the 'traditional Theremin circuitry' ends. The rest is more like an early Moog (or others) monophonic synthesizer. The above "GAIN" and "OFFSET" controls for the "PITCH" go through an inverter circuit (so that the higher voltage translates to a lower pitch, etc), but the "VOLUME" circuit stays non-inverted. The "GAIN" control voltage goes into a traditional "One volt per octave" (remember, you can scale it up / down with the "GAIN" and "OFFSET" controls)VCO (Voltage Controlled Oscillator) which creates a sawtooth (ramp) waveform that goes into a triangle wave converter, which in turn goes into a sine wave converter. Additionally, the ramp generator goes through a precision ramp frequency doubler, and again to a triangle wave converter and then a sine wave converter. The results are sawtooth (ramp), triangle, and sine waves at both the fundamental frequency and the second harmonic frequency. These six controls allow the user to come up with just about any sound they want. I thought about adding a square wave, but this really doesn't seem to belong in a Theremin, so I didn't. If you think it does, I could probably be talked into adding it in, although
The volume control I presently have is a traditional VCA (Voltage Controlled Amplifier) used in early solid state synthesizers, but I'm planning on switching over to using a TDA1524A volume / tone control chip that does a great job with few external components. The one part I haven't put in yet is an old school spring type reverb. This was the # 1 request I had on my tube design when I turned Portland's musician community loose on my prototype.

While I'm a big proponent of 'universal power supplies', I have a new appreciation for Bob Moog's decision to use an AC power pack, although I would never use a DIN connector. My power supply requires +15 and -15 VDC (regulated), which I can get from a 16 vac doorbell type transformer. My inclination is to keep this simple for the folks who want to build their own, and have the components be something they can come up with on their own.

Okay, so what should I call it?

Some of us are either designing theremins or at least thinking of how to design the ideal theremin. So I expect a controversial discussion here. At first we should thank Mark that he is about to throw a new real ball into our playground instead of sitting, reflecting and playing around with Maxima (yes, the linearity correction problem can be approximated) and LTSpice only as I for example do.

Moog has shown with the Big Briar 91 model that a non-heterodyning theremin may work and be playable. I rejected this idea for my own virtual design because I prefer having this audible slight hesitating of an RF oscillator when making quick and big frequency jumps.

Marks wave shaping ideas do sound a little bit as if he was about to re-invent the 8038 or 2206 function generator IC... Technically it seems ok to have such a good and fully parametric control over harmonics, but what will be the audible musical result. What if players would decide for using just 3 or 4 "really interesting" sounds and rejecting the rest of the synthetical spectrum? (This is a rhetoric question, resulting from my experience with the Etherwave standard where I make use of only 3 "presets" of the waveform and brightness pots.) I haven't yet taken a final decision on how "my" wave shaping will be designed but I imagined that an organ like control over harmonics (additive sound synthesis) could be musically more interesting. But I'm not sure if eight or ten Hammond-organ-like slide controllers will be accepted.

Using a spring reverb in a progressive design? Hmmm, my first reaction is thinking of "noise, distortion, microphony" keywords which fit better with a tube based circuit. If the rest of the circuit shall already be innovative, why not use a single-chip digital reverb. You will still have noise and distortion, but at least no microphony.

OK, this was the first of my five cents. I hope I'll have the occasion to come out with the rest during a long, interesting and copious discussion here.

Interesting.

How big? Does it use those massive air coils? Ease of transportation is something I would be looking for.

I note that it uses internal control voltages, but no mention of CV outs. It should have CV outs. There is no pro level theremin on the market that has them - the only ways to generate CV from space control currently available (that I know of) are the Paia Theremax and the Doepfer A-178 Theremin Control Voltage Source Module, both of which are a bit naff.

Synthesis rather than heterodyning. It's a funny thing, isn't it. Give people a theremin with one heterodyne voice and they'll be mostly happy - they might feel that they want a bit of variety - a few more voices - but they can nod their heads gravely and say "yes, it's a true heterodyning theremin" and be content. But settle on any other synthesis method out of the many available and you can be sure that most people would have preferred some other method.

My guiding principle in these situations is "If you can't please all of the people equally, displease them all equally instead." Question: is there any reason why the synthesis circuitry can't be a voltage controlled heterodyning oscillator? (If that's the right terminology.) I think that would make everyone mostly happy, and CV outs would give the opportunity for people to generate other timbres by whatever method they want.

Likewise for reverb. There are so many reverb options available - I know I'd keep the spring reverb switched off and feed the signal through a digital reverb pedal instead - that's my preference for live performances - for studio recordings I'd add stereo reverb in post-production.

Name? I think the name should indicate the selling points of the instrument, which are - 1. it's a theremin, 2. it's a Mark Keppinger design, 3. it has CV outs. So my vote would be for [i]Keppinger Theremin CV2[/i] or something like that.

Gain and Offset. Name things for what they do, not how they do it - it is going to be played by musicians, not electrical engineers. If adjusting the gain changes the range of the instrument, call it [i]Range[/i]. If the offset decides whether it the instrument is bass or soprano or something in between, call it [i]Register[/i].

Mark was considering "Synnemen".

"Question: is there any reason why the synthesis circuitry can't be a voltage controlled heterodyning oscillator?"

None at all.. but it is difficult, particularly if one wants a 1V/Octave CV so that the CV driving it can be taken out to other equipment. All one needs is a HF reference oscillator, and a HF VCO, and mixer.. Problem is scaling the CV so that the mixer output tracks at 1V/8Ve.

A way round the above problem is to derive the output CV from the audio produced from the mixer .. but this means that, at low frequencies, the CV out has long latency (the fastest it is possible to get a CV from a 100Hz signal is 10ms - but 20ms is really needed)

I do not think a simple heterodyning VCO is worth the effort - BUT - a really fancy one with Moog 901 style harmonic selection is most certainly worth doing.

Apart from the oscillator, the VCA (and, I believe, VCF) VC Synth 'components' are still required.. I believe it is the VCF (and how pitch and volume CVs are routed to this) which is key to getting the sound right - I am sure that both the E-Pro and Tvox have their appeal primarily due to the subtle use of VCF.

>> Following comments have been edited out by FredM on 20 Dec 2008

Fred Mundell
Fundamental Designs Ltd.
Electronics Consultant.
<- See Profile Image for Email.
Designer of Theremins and other alternative electronic music controllers and instruments.

I've had a bit of dialog with Mark myself. He has some interesting ideas. His design does not use a function generator IC, contrary to what it sounds like. He reports those are just not stable enough.

I've got to go find that company that makes an inexpensive DSP effects development system. It would be ideal for the reverb. As Thierry pointed out, they can be more more trouble than they are worth. They do sound good, but they are a "one-trick pony". DSP gives you a wide range of things to play with.

Fred -- my condolences :( It's tough to go it on your own. I have no idea what the work climate is in the UK, but here it is pretty bleak looking for engineering jobs.

Don