EM Theremin - Problem with VCA and VCA processor

Hi i'm new on this website! Some months ago i built Thierry's theremin as a first experience and now i'm trying to build the EM Theremin from Mr.Moog's paper.

I started by simulating the circuit on LTSpice and everything is OK until the Detector. For the VCA and VCA processor simulation i downloaded the Spice model of the LM13700 from the TI's website. I checked the connections of every pin but for some reason at the 'Audio Output' i get a voltage between 0 and -45 nV which is too low!

Anyone have tried this before?? Is it normal??!!

PD: I simulated the antennas as capacitors of less than 1pF

https://www.dropbox.com/s/qe1sb3v03ll57x0/Screen%20Shot%202016-05-21%20at%2014.59.07.png?dl=0

 (i don't know how to include a picture, so here's the link to a picture of the simulation...!)

1pF is definitely not enough. The volume loop has a static capacitance of 7pF + (0-1.5pF) for the hand, the pitch rod 9pF + (0-1.3pF) for the Hand.

When the volume oscillator and the series resonant circuit formed by the 5mH and 2.5mH inductors and the volume loop are in resonance, you'll find -4V at the interconnection of D1 and the 1Meg resistor which will make +12V at the output of the volume CV processor and finally the full amplitude (0.7Vpp) at the audio output.

YUHU!!! That was the problem thanks =).

Now i have another question, related with the construction of the instrument. I'm from Chile and obviously i won't find the 10mH inductors (neither the 5mH or the 2.5mH). Also is not an option to order them from internet. Which is the best way to replace those inductors? I'm planning to make them rolling up some cable, but i don't know what would be the best way...

"Which is the best way to replace those inductors? I'm planning to make them rolling up some cable, but i don't know what would be the best way..."  - geduran

Good luck with that, they are carefully engineered (scramble-wound donuts spaced away from each other and spaced away from the ferrite via paper) to have maximum inductance, minimum self-capacitance, and reasonable stability for their size.  Not saying they're entirely ideal (the Q is kinda low and the ferrite causes drift) but they do what they are supposed to do fairly well in that circuit.

[EDIT] You could maybe put a single tight layer of 1085 turns of 38AWG (single coat insulation) on 2.5" (73mm OD) schedule 40 PVC pipe, which should give you a ~40mH air core solenoid with a height of ~125mm.  I don't think working with 38AWG would be all that fun however.

Forget about replacing these coils with whatsoever. The EM circuits are designed to take into account the (relatively) low parasitic capacitance (and thus the elevated SRF) of the 3-Pi wound Mouser 63xx series coils. The only confirmed working replacement are the Hammond 1535 series coils. With all other approaches you will have bad linearity or a restricted pitch range, or both. 

Don't tell that the big international distributors like Mouser or Digikey won't ship to Chile. Chile is AFAIK not a third world country...

"Forget about replacing these coils with whatsoever ... With all other approaches you will have bad linearity or a restricted pitch range, or both."  -- Thierry

Theremin himself employed single layer air core solenoids here, no?

Theremin designed his circuits to deal with the lower SRF of single layer air core solenoids by using a lower oscillator frequency and a different L/C ratio.

A theremin oscillator in conjunction with a linearization circuit is like a clockwork or a gear box. You can't modify a single gear wheel, everything has to fit together.

dewster, I don't know if you have worked as much with the Etherwave circuit as I have, including repairing/fixing, upgrading, and fine tuning >100 instruments. Thus, you can be sure that I know what I'm talking of when it comes to replacing critical components. You should not mislead beginners with your theoretical ideas which you have never proven in practice - at least as far as I know, none of the renowned professional thereminists has ever played an instrument which you have built, repaired or upgraded in a public concert or a CD recording.

I recently downloaded the same model from TI and got it to work correctly with no issues.  How did you use the model to create the LTSpice part? I have a VCA sim based on the LM13700 I can send you if you want. You can send me your sim if you want me to take a look at it. I also have a sim for the LM13700 as a mixer and a limiter that works. I have built the mixer and VCA on a breadboard and they also work quite well.

Mark Allie

"Forget about replacing these coils with whatsoever. The EM circuits are designed to take into account the (relatively) low parasitic capacitance (and thus the elevated SRF) of the 3-Pi wound Mouser 63xx series coils. The only confirmed working replacement are the Hammond 1535 series coils. With all other approaches you will have bad linearity or a restricted pitch range, or both." - Thierry

As an experiment I picked a coil I wound some time back that measures 9.105 mH with my LC tester and 94.5 Ohms DCR.

I directly stimulated one end of a coil with the output of a function generator, and hooked one scope probe to this as a reference.  The other end of the coil was left unconnected and a second scope probe placed some inches away.  The coil was placed on top of a small plastic box to distance it from the workbench.  With this arrangement the self resonance is 850kHz and the -3dB bandwidth is 7.2kHz, which gives a Q of 118.  I'm pretty sure the Q is a bit higher than this because the function generator output is 50 Ohms (which damps things).  And I'm sure the self resonance point can be made a bit higher because the center tap wires are just hanging out, as is the passive end, which increases intrinsic C.

I don't have a 10mH Bourns 6306 to compare it to, but the datasheet lists an SRF of 0.71 MHz minimum and a Q of 95 (measured at 79.6 kHz, whatever that means [EDIT: I suppose it means if the L were padded with a perfect parallel C to resonate at 79.6 kHz, the Q of the combo would be 95]).

I do have a 50mH Bourns 6310 on hand, the datasheet lists an SRF of 330 kHz minimum and a Q of 100 (again, measured at 79.6 kHz).  With my setup I measure a self resonance of 470 kHz, a bandwidth of 14.9 kHz, and a Q of 31.5 (at self resonance).

From the above, I think it's fair to say that one probably could replace the Bourns inductors with single layer air core solenoids, perhaps a series of them to keep SRF high, though series resistance might have to be added to kill the Q if that is an issue.  But what do I know.

Almost like the old days, TW needs conversations.

I enjoy engineer saber rattling and believe in keeping it simple like my circuit designs which I often use the smoke test. I am on the side of exploring why the builder cannot buy the coils? Valuable opinions can come from those that have actually demonstrated what works in practice, the real world.

Christopher