Goals for a TW Theremin

How about this: come up with a simple, basic, bullet proof schematic that we can point first-time Theremin builders to rather than leaving them to all the crying game ones out there.  Make the design as simple and as trouble free as possible.  Get people hooked on the Theremin by giving them a great first experience rather than discouragement.

But I personally don't think this is very doable without the benefit of modern electronics.  If it means abandoning all the legacy stuff (tubes, control voltages, heterodyning, etc.) will the magic of capacitive control at a distance really disappear?  I understand how the long legacy is attractive - it is an indication of a field tractable by enthusiasts - but I think we could all have incredible instruments for a lot less if we looked to present day solutions a bit more.

"But I personally don't think this is very doable without the benefit of modern electronics.  If it means abandoning all the legacy stuff (tubes, control voltages, heterodyning, etc.) will the magic of capacitive control at a distance really disappear? " - Dewster

Personally, I disagree ;-) I see no reason why a simple theremin cannot be made cheaply and quite easily. The Moog EM was simple enough really, and matching this design or bettering it now should be attainable.. The main factor why good cheap theremins have not appeared is, IMO, down to manufacturing economies and profits.

I am not suggesting that there is any major rip-off.. Running a business is expensive, people working for you need reasonable wages - and all this must come from the sales.. But - Do a quick costing of what an EW will come to if one only included parts - then scale this down if one was looking at buying parts / PCB's for 100 units.. I think we would be clocking in at about $100 or less.

But - add assembly costs and overheads, and you need to build / buy more than 100 units to  break even at the EW RRP.

For people willing and able to build their theremin, I believe a good theremin could be done for under $100. For a project like this, where there is no development cost to recover, and no "physical" overheads, I do not see the need to go too radically off the standard path.

Fred.

I just ran something like that on Simetrix, and got similar results. 42IF106 is a good choice. Cheaper than dirt.

Also playing with a dual gate mosfet mixer. By messing with it's bias, the product can be mangled quite nicely, or made very clean. I wanted to try that after looking at the tetrode mixer in Clara's.

I didn't post a diagram because it's absolutely conventional.

"Also playing with a dual gate mosfet mixer".. w0ttm

Yeah - dual gate fets look like they could directly emulate tube circuits quite nicely - but do you know of any readily available parts ? .. And even better if you know of any which have spice models available..

I have simulated a mixer with 2 series jfets -  And will move my schematics to a "Scratchpad" thread so as to de-clutter this one.

Fred.

Guys,

I have the materials to wind another set of replica RCA osc coils and will be  ordering  a bunch of components in a few weeks, so....

If you engineers throw together a list of possible fet candidates, I can add a few of each to my order and attempt a real build.   I also have the RCA antenna coil, a bunch of miller 6304's, and a spare EM circuit board.
-  Ill Build the Pitch osc and mixer circuits and patch/ run the signal thru the volume/ VCA half of the EM. (also use EM powersupply) - we can  instantly hear how well it works!  

Does that sound like a TW theremin or what?

Again Im no engineer, so humor me ...

   ("ive heard...") running theremin simulations can be problematic and dont shed much light on how the instrument will sound (what's really important, to me anyway)

-lets start with Howie's advice, and replicate the rca pitch and mixer circuit exactly, but w/o tubes     - then start playing/experimenting with tiny modern coils and making changes.

"lets start with Howie's advice, and replicate the rca pitch and mixer circuit exactly, but w/o tubes    " - Chobbs

The above is not possible. The word "exactly" makes it absolutely impossible.

Its also starting to look like perhaps no one has ever really understood the operation of the RCA oscillators.

It is my feeling that if we fully understand the operation we will be better placed to replicate (or emulate) its functions/s using semiconductors.. I do not think there is any point in copying antique hardware - I do not think the "secrets" are  in these.

Regarding the Fet oscillator being explored at Theremin Circuit Scratchpad - This has evolved into exploration of the operation of the original RCA oscillators (as well as simulations of a Fet implementation) ..

It would be interesting to compare modern inductors with copies of the original RCA coils - So if you had a set of the coils to compare against, this would be interesting..

The circuit itself is incredibly simple and could easily be built on a tiny prototype board . I have ordered some off-the-shelf transformers and inductors which have a spec close enough to the original (160uH rather than 165uH) to allow testing, and I have commissioned a technician to build me a couple of oscillators and mixer onto small prototyping boards - I have a load of high quality inductors to construct the antenna circuit.

Use of the EM board is probably not a good idea IMO - The VCA on this is not right for RCA replication, and the resulting output will prove nothing - IMO, if we want to replicate the RCA we need to stay with FETs for all functions like mixing / VCA etc (as in, the multipliers must be FET based).

There is also the potential problem regarding interference from  volume oscillator frequencies - I think that starting without a volume circuit, one can still assess the tone quality.

Fred.

It's threads like this:

http://www.thereminworld.com/Forums/T/28791/help-with-em-theremin-construction?last=True

that make me think the first TW design should be plug and play as much as possible.

IMO it's kind of weird looking inside a modern Theremin and seeing what looks like an AM radio from 25 years ago staring back at me.  Or in the case of a tube Theremin, 50 to 100 years ago.  Shouldn't we consider updating things a bit?  We're living in the golden years of electronics.

Dewster, I'd love to escape from LC. I just don't have a frapin' clue how to.

The same thing also fits with a lot of other instruments. My Stratocaster and Les Paul are virtually unchanged from the first prototypes built in the 1950's. Same with amplifiers. The best sounding ones have tubes.

There have been advances in capturing the tube sound in silicon, most notably with DSP. They are still not quite the same, but they come darn close.

If those techniques could be applied to a theremin, well I'm all for it. The model 91 is a good start.

A sampling instrument is a thought. Record the greatest theremins, and let the software correct linearity, volume response, etc.

I'd rather not write the code for that one...

I can see the argument for "going modern" -

But I cannot see an argument for abandoning a principle which is beautiful, simple and elegant, just because its not "modern".

Look at the simplicity of the Fet oscillator version of the RCA (Termen) oscillator - It is only a simulation at present, but it will be possible to make it work even if it didnt work on first build.

Ready made, low cost wound components will be found (or could be manufactured)  which would remove the difficult aspects of construction (I have found a transformer at <£1 which looks like it could do the main tank inductors job) .. Build a simple VFO, simple Refo, simple FET mixer, wire a bunch of low capacitance inductors together (the most expensive part), have a screw adjustable length antenna for tuning, a simple volume antenna and opto-FET VCA, and for about £50 you have a RCA clone.

I do not think one could beat that with anything more "modern" than Fets, BJTs, Opto-VCAs and Opamps!

Looking at the RCA theremin, and understanding it for the first time, I am forced to recognise utter genius - This is more than a modified ancient radio - Everything is crafted to perfection.. The oscillator is incredible, its interaction with the antenna resonator is incredible - In fact, not sure yet (its only simulation) - But the resonances of the combined tank and antenna tuning circuits look to me extremely familiar.. there are twi destinct resonant peaks and the proportional positioning of these looks similar to a vocal formant -

I should just say that these are peaks generated by the strange series wiring of the oscillator tank - and these peaks are replicated in the reference oscillator despite it having a slightly different topology - This does not look like an accident to me! But it is only simulation - the simulator could be producing mathematical artifacts which dont exist in reality.

It may be just my mind ascociating familiar signature and imposing my interpretation, but it looks deliberate, and if these harmonics do actually appear after mixing - well - genius is too small a word.

I think the mistake has been to automatically attribute the qualities of old instruments, to the technology used to implement them..

I think the reason they sound so good is down to the care and knowledge of those who created them - An art which we have forgotten.

Fred.

What you wont be able to do is to extend the range - you will be limited to a top playable frequency of perhaps 1500Hz*, so if you need a higher register, you need an EW type design to give you one more octave - If you want more, then perhaps more "modern" implementation is required. *Well, there is a way, but I cannot share it.

Im pretty excited about fred's experiments with the armstrong/lev oscillator, and if it works out I think we should go with that as the basis of the design, albeit, with comparatively more modern parts and no hand wound coils, for ease of construction, in the end, I want to be able to unceremoniously toss it in the back of a van when going to a gig, not sure that large air coils, as nice as they are, will like that very much.

Id prefer a professional level instrument over an introductory one, there really isnt a better instrument than an ew standard you can easily buy (new) or build. Sound quality is the most important factor to me.

But, on another more conceptual level, a theremin to me is an idea, which i dont know is conceptually, compatible with digital (please convince me otherwise). That and an experience with cheap digital effects is probably influencing me against it. But id think it would be nice, conceptually, to make a good instrument, that has a historical connection to levs design. The theremin is, well, a historical instrument, not a contemporary one, there would be something nice about knowing there is some history in the instrument, when you step up to play.

But it seems like we are all keep bringing up all our same pet concepts. Id say, lets build a theremin that sounds like an rca/claras, to the performance specs i mentioned if there is no objections (minus the timbre, mostly there i was thinking of, i believe, the paia theremin, that suddenly goes from violin-like to suddenly flute like which is awful) and to the interface specifications, which fred should probably propose, in whatever the best way possible, with open minds. Its an interesting challenge.