RCA THEREMIN

"Perhaps its time I left the theremin community and went to work with some real engineers again.."  - FredM

I can see how real engineers like get totally sucked in by the siren song of Theremin design, it's much deeper than it appears on the surface.  Femto-Farads running the whole shebang and not getting nailed by thermal or environmental noise/disturbances is amazing and  counter intuitive IMO.

It seems clear to me that the output waveform can not be a sine wave but has to be distorted and that the FETs have to go into overdrive and cutoff. Bob Moog noted in his "repair notebook" concerning Clara's Theremin that he found amplitudes of about 70Vpp (this version of the document mentions even 100Vpp)at the grids of the mixer tube. Since the 224 has a cutoff voltage of about -10V, I did the simulation with the same proportions, using FETs with -1.5V cutoff and amplitudes of 10Vpp. Finally I choose the DC offsets and the "too high Z" termination on the secondary side to come close to the desired duty cycle waveform. If you look at the waveforms on the RCA theremin page you will find that the duty cycle varies somewhat with the frequency. This is in my eyes another indicator for the use of a kind of "phase cutting" technology.

Nothing about moon phases...

Here are the RF bursts I was talking of:

@Fred

You had mentioned finding a 160uh transformer.

Please let me know where to get such a critter.

I FINALLY got my hands on a couple of fets and I'm suffering from soldering iron withdrawal.

If I don't build something soon....

Hi Rob,

The transformer I got doesnt really work - Yes, each winding is about 160uH, but the ferrite has extremely high permeability, so when connected in series the inductance was 1.2mH !

I have ordered some other transformers with lower AL, and will let you know how these measure up.

I am also looking at a minor re-design - having a smaller ferrite transformer with say 50uH per winding, and combined inductance of something like  300uH (High Al core) and a seperate 100uH inductor an both top and bottom 'arms'.. This should give a total of 600uH, with 150uH on each 'leg' and still provide more than enough 'tickler' feedback .. These inductors in the 'arms' could be adjustable (Cap less IFT) if required.

However.. I am starting to think it might just be simpler to wind these ourselves - I do not see that we need a large former - I think that plastic former as used in small pot-cores should do the trick - All we need is to maintain the independent winding inductances as they are on the RCA (168uH or thereabouts) and the combined inductance to be about 600uH..

If these other transformers I have ordered turn out to be unsuitable, I might get some transformers made to my specification by a local company - I think that these transformers will be quite simple (certainly simple compared to others I have had them build, LOL ;-) - I can probably get a few samples free (or for no more than about £5 each) and could probably get them made for about £1 each if I ordered 100.. If I then sold them for £5 each, I would break even if 10 people bought the required 2 per theremin.. But, for now, I am hoping we can buy what we need from Farnell.

I was only able to "use" the transformer I obtained by putting (tunable) parallel inductances across each winding - in fact, I ended up with only one (across the tickler winding) to reduce the total inductance without seriously reducing the "lower" winding inductance - but this unbalances the configuration and is probably an extremely bad idea!

For now, I would advise winding a simple transformer - I know you are able to calculate the turns required etc! (I know from the speed you did it for the RCA windings, that you are faster than me! ;-) .. I would start by computing the number of turns  required for 600uH, then see what inductance 1/2 this number of turns gives.. If this is anywhere near 160uH, you have all you need.

Fred.

"Bob Moog noted in his "repair notebook" concerning Clara's Theremin that he found amplitudes of about 70Vpp (this version of the document mentions even 100Vpp)at the grids of the mixer tube." - Thierry.

Oh, I do hate it when people take a comment entirely out of context. - What Bob Moog also wrote in his notes was THIS:

" Early theremins use the venerable 24A as a mixer tube. Here is where the audio wave is synthesized. "

Bob says quite a bit about the operation of the mixer, and its effect on the waveform:

"The mixer tube is biased beyond cut off even though its cathode is grounded. Bias is obtained not only from the large negative peaks of RF voltage applied to the 24A grids, but from a steady component developed in the cathode circuit of each HF oscillator. The RF sine wave fed to each grid is very large -on the order of 100V p-p or more. Only the peaks of positive voltage are large enough to produce pulses of plate current in in the 24A. This plate current forms pulses corresponding to the frequency difference between the two HF sine waves. The magnitude of the high frequency pulses, but not the steady component of the bias, is determined by the pot adjustments. Therefore, it is a complex system, because turning either pot not only alters the magnitude of the synthesized audio frequency, but its waveform. The RCA theremin is very similar in this regard, but lacks rheostat adjusters."

And not a single word about how the transformers "create" the waveform through "impulse response" - In fact, not a word about the transformers influence on the sound at all.

Until proven otherwise, I will take the word of someone who has actually examined these theremins, someone who grew up with tubes, who had unquestionable competence, whos analysis makes technical sense..  over yours.. Particularly as this person wsa Bob Moog.

"It seems clear to me that the output waveform can not be a sine wave "

I dont think anyone has said it was - I certainly havent.. What I have said is that I believe that understanding the mixer, and the WAVEFORM PRODUCED BY THE MIXER, is of paramount importance - You, on the other hand, have effectively declared that the mixer is of minor importance, and that the waveforms are "created" by the transformer charactaristics.

But this is really getting pathetic - I have said all along that you MIGHT be right, I have not said you are wrong, I have simply presented my analysis and declared my uncertainties and admitted that I am out of my depth. However you have not, IMO, given me any reason to accept your statements as absolute truth, you have given me no reason to think that you actually know what you are talking about..

What you want from me? Faith and worship ???

Fred.

Lets wrap this up.. My last words to you on this matter (I hope):

I disagree with you on the following:

"don't waste too much time with the 227 and 224 tubes as I did.
 The secret of the waveforms is elsewhere in my opinion."

"And I continue insisting on the fact that the oscillators, the mixer and their waveforms are only of secondary relevance, the "magic" is in the impulse response of the audio transformers."

"The transformer's impulse response makes the waveform."

But I fully accept that the charactaristics of the components following the mixer (including, but not limited to, the transformers) could - no - will - have a profound effect on the final audio output.

My plan and stratergy has been clearly stated - I believe that, to "clone" an RCA or Claramin, the waveform from the mixer should ideally be crafted to match that of the output waveform from the 224 mixer as closely as possible, and from this starting point, following circuits devised which behave in a similar way to the circuits following the 224 mixer.

Our major area of disagreement appears to be that you believe a transformer is required to produce the waveform from the mixer - I fundamentally reject that idea.

"What you want from me? Faith and worship ???" - Fred

No, nothing, and please accept my apologies for having contributed my view of these things.

I must admit that my actual point of view is rather intuitive. When I was young I learned to look at the step response of a system and to derive its electrical properties. Based on that, my above "theoretical explanation models" seem clear and obvious to me.

They do not only seem clear and obvious to you, but you seem to feel offended by my conclusions, thus I'll stop sharing my thoughts in this thread, especially since I don't seem to be a real engineer in your eyes.

Guys, we are all in this together.

We all have our own, different areas of experience, different world views, cultures, and languages. This can, and apparently does cause friction.

In the high end audio world, there are heated debates about the virtues of the single ended triode power amplifier. The engineers say it's crap. Loaded with distortion, and terribly inefficient. The audiophiles claim it as a gift from God, with magical properties, and they are willing to spend insane amounts of cash to buy them.

Well, I don't care. Life is too damn short.

What can we bring to the table here?

Thierry has probably seen the inside of more theremins than uncle Lev.

Fred has classical engineering training and experience.

Dewster is Mr. digital.

I eat, sleep, and breathe RF.

And so on.....

We all make a theremin stew. Sometimes there will be bits of bone and gristle. Just spit those out, and enjoy the rest.

"I am not "offended" by your ideas - I believe they may be as valid as mine. The only "offensive" issue was your "animosities" comment - this really pissed me off." - Fred

Ooops! Now I see my big mistake, I used the term "animosities" as I would have used "animosités" in French where it is a synonym for "excentric traits" in the current language. Now I looked it up and found that in English it means rather "hostilities". This is NOT what I wanted to say and I understand now why you were upset, Fred. Thus please accept the sincere apologies from someone who is not an English native speaker. 

eccentricity =  French animosity;  French "animosity" != hostility.

"pollute the thread with my animosity" = "pollute the thread with my eccentricities.."

Oh F**k.

Apology accepted, I likewise apologise for my reaction.

Fred.

Im sure these have been asked before and I can remove this post if I am being redundant or I am in the wrong forum. I will research the posts tonight but only have time to shoot this off.

I recently acquired an RCA theremin and have paired it with a 106 speaker. she sounds lovely and the linearity is pretty awesome. 

I am having some drift though. If this issue is discussed somewhere else I apologize and please send me in that direction.

I have noticed that exchanging the 27 tubes and the 24 tubes greatly reduce and/or amplify this drift. 

Am I to keep switching the tubes around to find a set that work together? Is there another tube in there I may want to try to switch? 

Any help would be greatly appreciated. Thanks!

Nico