Goals for a TW Theremin

I think we should also have as much of the relevant response stuff accurately simulated.  Things like linearity and sensitivity.  That way people can play around with the simulation and do a bunch of "what if" with the basic design, and there won't be as much hand waving when it comes to describing it to others.

I'm kind of against modularity. 

I'm hoping the component cost for a basic model (pitch & volume) can be kept below $50.  Above this you might as well go digital, sidestepping most of the drawbacks of analog and reaping the almost infinite benefits of logic.

Might we consider a mixed design: analog resonant elements combined with digital oscillators, but still use heterodyning for tone generation?  This might help a lot with initial calibration, linearity, tuning, and stability.

Sorry, I probably tend to get out my digital hammer too often, which makes every Theremin issue seem like a digital nail.

Dewster, a single, not modular board would be preferable. Construction would be simplified and the chance of wiring errors greatly reduced.

I'd like to present my case for modular.

It could incorporate as much or as little complexity as the builder desires. It could be adapted to any oscillators. CMOS, EW, etc.

Those oscillators could be built with or without linearizing coils with just value changes.

The builder could load it up with digital, analog, or any mix they chose.

I guess what I'm suggesting here is a TW theremin that's the builders idea of a perfect instrument, not ours. Lets each of us contribute an element of it, and stay in touch through this forum, E14, and e-mail to be sure our modules will play nice with each other. Also, I'm all for using the same, or compatible editors and simulators.

I don't see why a basic pitch and volume version could not be made for very little cost, and then added to so much as to rival an EW pro.

Fred's compromise double sided board sounds good to me. There are tricks that can produce double sided boards with the toner transfer method. I'll post a link as soon as I remember where I saw it.

Nieradka's specs sound reasonable and achievable to me.

Documentation of the theory behind each section and how they interact will be one of the most important things we do IMO. One of Jason's goals is that this should be a learning tool. Understanding how a theremin works makes building it and playing it much more rewarding.

Through this documentation, the builder can make informed decisions about what they want the finished product to do and sound like.

More to come....Rob.

"I guess what I'm suggesting here is a TW theremin that's the builders idea of a perfect instrument, not ours." - Rob.

Yes, I agree..

But herein lies a big "problem" - There will be (are) a huge number of personal preferences amongst thereminists, constructors, experimenters etc.. No way to keep everyone happy.

So, a modular gives people the ability to choose their route - but doesnt keep those who want a single board theremin happy.. LOL.

I know I get boring .. But it comes back down to specification and development plan..

Lets say that the ideal would be a single board theremin which catered for the majority of builders and users, AND the option of a modular for those who want this..

The logical progression, IMO, for the above, is to start with modular prototype boards, develop these, combine them to form complete theremins, THEN select one 'combination' and put this onto a single board.

At this stage one has designed a range of modules, so all the work is done - and one has a "standard" build for the majority.

Which brings up the specification aspect. The first specification required would be an interface specification, so that modules are compatible. This specification needs to define at least:

Supply voltages

Signal levels for both HF and Audio signals

Control Voltage level for VCA, and (if one goes the CV route, or uses Pitch Related CV's for internal filters etc - a more advanced matter) Voltage levels for this.

Impedences for all output signals (including VFO and Reference HF signals - I personally believe all signal outputs should be low impedence - ideally not greater than 1k, certainly not greater than 10k - as in, a resistive load of 10k does not affect the source.. Likewise, capacitive loading needs to be specified - as in, for example, the HF oscillators (their buffers) should be able to drive a capacitive load of 100pF through a 1k resistor without problems.

With a specification established, people can start designing modules - I am talking here about electronics / interface specification which does not relate to performance specification - that is another seperate matter.

"I think we should also have as much of the relevant response stuff accurately simulated.  Things like linearity and sensitivity." - Dewster

Sadly, simulation of the front end is no easy job, and rarely IMO gives anything close to what one gets from a prototype - I love simulation, and have used it for the last 20 years in design.. Simulation saves many hours and highlights problems long before one touches a soldering iron - in most cases.. But for LC circuits and antenna linearization, I have found it almost completely useless... For mixers, Wave-shaping,VCA's, audio circuits, power supplies etc, Simulation is great (some simulators actually allow one to output the audio generated - As Proteus does this, I have not explored if LT-Spice can). 

Fred. 

while I love the enthusiasm,  Ive been reluctant to jump in the conversation.... seems that I have a vastly different idea of what a "good" new theremin design should be.  

just for consideration,   Here is an excerpt from a Levnet email where, theremin guru,  Howard Mossman describes his ideal new theremin design:

"Might I suggest that any new design NOT be a tube type. These designs tend to radiate badly, use high voltages and cause "birdie'' like spurious interference without extensive shielding.
 The RCA design is excellent but instead of tubes, they might investigate solid state FET's (Field Effect transistors) which have similar characteristics to tubes, i.e. high input and low output impedances but require only low voltage power requirements. The pitch coils MUST be as exactly as designed in the RCA using two Armstrong oscillators for the pitch circuits. The large pitch coil in the RCA can be substituted with a modern smaller coil of the same impedance.The mixer is the key circuit as it will determine the ultimate basic tone of the instrument, and must be designed (biased) very  carefully to achieve the final desired sound.
 The volume circuits should be copied from the Moog Melodia as they work perfectly and are solid state; also operate without the annoying "chirp" that is emitted when the Etherwave volume loop is touched.
 This new theremin might be just what is needed if it can be made in kit form to keep the cost down."
 Uncle Howie

while I love the enthusiasm,  Ive been reluctant to jump in the conversation.... seems that I have a vastly different idea of what a "good" new theremin design should be.  

just for consideration,   Here is an excerpt from a Levnet email where, theremin guru,  Howard Mossman  gives some suggestions on a new theremin design:

"Might I suggest that any new design NOT be a tube type. These designs tend to radiate badly, use high voltages and cause "birdie'' like spurious interference without extensive shielding.
 The RCA design is excellent but instead of tubes, they might investigate solid state FET's (Field Effect transistors) which have similar characteristics to tubes, i.e. high input and low output impedances but require only low voltage power requirements. The pitch coils MUST be as exactly as designed in the RCA using two Armstrong oscillators for the pitch circuits. The large pitch coil in the RCA can be substituted with a modern smaller coil of the same impedance.The mixer is the key circuit as it will determine the ultimate basic tone of the instrument, and must be designed (biased) very  carefully to achieve the final desired sound.
 The volume circuits should be copied from the Moog Melodia as they work perfectly and are solid state; also operate without the annoying "chirp" that is emitted when the Etherwave volume loop is touched.
 This new theremin might be just what is needed if it can be made in kit form to keep the cost down."
 Uncle Howie

"Timbre is consistant across those 3 octaves. " - nieradka

This is the only point I dont agree with - I think that timbre changing as pitch changes can make the sound far more musical and interesting.

The ability to adjust this variation from zero to 'quite a lot' LOL, is one of the things I have put on my theremins - and if they dont have an adjustment for this, I will give a little variation as standard.

Fred.

"The pitch coils MUST be as exactly as designed in the RCA using two Armstrong oscillators for the pitch circuits. The large pitch coil in the RCA can be substituted with a modern smaller coil of the same impedance.The mixer is the key circuit as it will determine the ultimate basic tone of the instrument, and must be designed (biased) very  carefully to achieve the final desired sound." - From Uncle Howie.

Oh dear! - There is only one thing I agree with in the above, and thats : "The mixer is the key circuit as it will determine the ultimate basic tone of the instrument"

IMO, large pitch coils achieve nothing. Armstrong oscillators might be worth looking at - but I dont see any intrinsic mechanism that makes them superior to other topologies.

But - Its Uncle Howie! And I greatly respect the man, and past consultations with him have been informative and useful, so despite my reservations regarding what he says above, his words should never be discarded lightly.

Fred.

Fred, On paper, you are probably( most certianly) right... but there is definitely some "magic"  to Lev's design. (and it is certainly simple enough!)   As far as I know, all of his instruments used basically the same pitch circuit as used in the RCA.  Other tube- design theremins(keppinger, for example)  just dont have "that" sound.  Could just be the Armstrong oscillators, not the size of the coil .... kep has large osc coils, but uses  a hartley  configuration.

philip should chime in here...

(Question... anyone familiar with the later Lev-built "soviet" theremins posted on RCAtheremin.com? Interested in how or if his design changed  )

Mark Keppinger chose the hartley for low output impedence. If we buffer the output, we have that no mater what type we use.

Melodia VCA. Well, even though it's uncle Howie, I think we can do better than that, although mine works pretty good.

What about the greatest Lev built ever? Clara's.

What about the greatest Lev built ever? Clara's.

-"Clara's" is  the RCA pitch circuit but with a few variables (pots and a cap) for tweaking