Let's Design and Build a (mostly) Digital Theremin!

Posted: 3/1/2014 6:27:24 PM

From: Northern NJ, USA

Joined: 2/17/2012

livio, our discussions have been quite fruitful, I hope you don't think I don't appreciate them because I really do!  You made me think about going to higher frequencies which can be quite freeing from an enclosure standpoint.  Enclosures had me dead in the water!  And you got me thinking about plates rather than rods (I'm preparing a spreadsheet for the comparison).  Please don't leave TW over some misunderstandings.  It's just that your project could probably benefit from a separate thread where the merits could be discussed there.  This thread is kind of losing focus.

Posted: 3/1/2014 6:37:47 PM

From: Northern NJ, USA

Joined: 2/17/2012

"That shielded antenna looks quite familiar! ;-)"  - FredM

All credit to you Fred for thinking of shielding in the first place!

"You do need temperature compensation when using this kind of shielding - I used a stereo jack socket so I could have a thermistor inside the antenna, and sum this into my TC circuit.. "

I'm kind of married to the UHF connector, though I'm starting to forget why.

The reduction in sensitivity that shielding brings could be made up by heterodyning, but maybe at the cost of stability?  I'm kind of averse to throwing away sensitivity, but maybe that's just another thing I'm married to.  I'm definitely averse to temperature sensitivity which I guess is the clincher (for now).  I'll give it another spin when I get heterodyning up and running.

Fred, what do you think of my interference conjecture up-thread?

Posted: 3/1/2014 6:50:12 PM

Joined: 2/2/2014

OK.. re-subscribed... 

I also promise that I will try to write only few and short messages!!! 

I do not know if I will make a new thread...
If someone more experienced could start it, it would be great, I'm old and I'm not very good at these "social" things  

In any case, thanks for have beared (often really too much) passion.

Posted: 3/1/2014 7:03:16 PM

From: Northern NJ, USA

Joined: 2/17/2012

"I do not know if I will make a new thread..."  - livio

Click on Forums | All Forums

From here either Theremin Construction or Technical Theory (up to you), then click on the New Thread button at the top and start a new thread (Title and first post).

Good to have you back!

"In any case, thanks for have beared (often really too much) passion."

You're not alone, passions tend to run rather high around here, mine included.

And I do agree with you that the EW coil emits RF, I'm just not sure how much:


I've always been fascinated by AM antennas, it seems they work by a combination of resonance and strong broadcast signals.

Posted: 3/1/2014 7:14:39 PM

Joined: 2/2/2014

"EW coil emits RF, I'm just not sure how much:"

A loaded antenna can be a very good antenna also if very short. Am radio receivers demonstrate this. The losses are all concentrated in the loading inductor resistance. If you measure the total DC resistance of your EW bobbins we will calculate the real emitted power.   

"...AM antennas, it seems they work by a combination of resonance and strong broadcast signals..."

AM antennas are not efficient like a true quarter wave antenna, because they use only the magnetical part of the electromagnetic waves. But, if correctly resonating with high Q, they can be very efficient. In the night, when the propagation is high, they can receive signals from the opposite side of the earth. 

Posted: 3/1/2014 8:30:20 PM

From: Northern NJ, USA

Joined: 2/17/2012

"If you measure the total DC resistance of your EW bobbins we will calculate the real emitted power."  - livio

For the 10 mH Bourns 6306 this is 31 Ohms (max) and there are 3 or 4 of them in series on the EW.

Posted: 3/1/2014 8:45:18 PM

Joined: 2/2/2014

With 120 ohm about 0.5 mW are loosed to heat the inductors. But there is also another consequence, the "Q" of the circuit is reduced and the voltage and current are reduced too. So the power on the resistor that simulates the electromagnetic field is decreased from about 26 mW to about 16 mW. I do not copy the image of the simulation to be concise. If you need i will post it. As you supposed the output power is rather small then my initial 50 mW, but equally to be considered.

Now we can also do other interesting considerations (very roughly). A perfect antenna (quarter wave) probably would emit around 60 mW and would therefore be about four times more efficient. This same relationship will be also while receiving. So there is not a very good coupling between the EW, but (my guess) will be difficult to get them to work at less than 10 meters from each other. If we eliminate the large coil changing it from 33mH to 330uH (going up to 3 MHz) and enclosing it in ferrite, this can significantly reduce the coupling between the Theremins. I am very glad to had the opportunity of explain this. Thanks for your help!

Posted: 3/1/2014 9:24:20 PM

From: Northern NJ, USA

Joined: 2/17/2012

The orientation of the inductors is probably important too.  On the pitch side of the EWS the first choke in the string is oriented East, the next North, the next East.

Posted: 3/1/2014 9:41:43 PM

Joined: 2/2/2014

L1 and L3 are in the same direction, not opposite. If you de-solder one of them and re-solder inverted it does not changes the winding versus. One solution it would be buying different inductors with opposite windings.... Another would be to invert PCB connections, but it is evident that this has not been done. 

However, your idea to reduce the field is right and should work. You might not get a perfect cancellation, but an attenuation of 20 dB is feasible and would certainly be a good step in the right direction. But the number of inductors should be equal (two or four)

I still prefer 330 UH ferrite inductors. They are very stable with temperature. I think TDK inductors are more stable then those of the EW. Maybe one day we can try it with a hair dryer, heat by 10 degrees and test the frequency change?

Posted: 3/1/2014 10:10:27 PM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as WaveCrafter.com . ...................................

Joined: 12/7/2007


Your comments on loss of sensitivity from using passive shielding are extremely valid - My experiments with shielding led me to the following conclusions:

Driven shields:

1.) Active shielding works extremely well, does not have temperature problems, and does not reduce sensitivity.

2.) If one has high antenna voltages, implementing active shielding is difficult, expensive and troublesome - One needs to provide a high voltage supply to drive the buffer.

3.) Due to the above I played with much lower antenna voltages, and got good results using 20V P-P on the antenna which allowed me to use a +/- 12V supply for the buffer, and this worked well.

4.) I was (am) worried about driving the shield with from a low Z driver, in terms of possible increase in radiated signals - whilst the shield stops other theremins or people from interfering on the non player side, it did not protect unshielded theremins from picking up signals radiated by the shield.

5.) The fix for the above was to have another ground shield covering the driven shield - this blocks all electrical field radiation from the shield, but adds extra capacitance the buffer must drive - it was about this point that I got distracted and moved on to something else.. But If I was to produce a high-end theremin, I think I would revive this design.

Passive shields:

6.) I have concluded that the problems with passive shielding and the cost of correcting these makes them quite limited in application

7.) If one has temperature compensation circuitry allready, and can simply add a thermistor to the error summer, then it may be worth using passive shielding.

8.) the best passive shielding involves having a fat antenna assembly so that one can get maximum distance between the sensor and the shield while still maintaining a focus - this reduces shield capacitance, and also greatly reduces thermal problems (but does not eliminate them) and increases sensitivity (or at least the reduction in sensitivity is reduced)

9.) For extremely low cost 'toy' theremins where thermal drift may not be regarded as a problem, passive schielding is easy - a simple Jaycar/SC can be fitted with a schielded antenna (and neon ;-) and the tank capacitor reduced "to taste" - because the shield replaces a lot of the tank capacitance, sensitivity is unchanged.

10.) for any "pro" theremin, I decided that low voltage double shielded (active and ground) antennas would be my choice.

Fixed frequency:

11. ) Active shielding is a lot easier with fixed frequency antennas (like my upside-down topology) - For one thing, it is easy to generate the higher voltages synchronously with the fixed frequency (a big problem I had was using assynchronous SMPS to produce this HV) but even for low voltage shields, having a known fixed frequency (and known capacitive load when doing double shielding) makes the buffer design a lot easier.


In all the above experiments, when there was adequate shielding (either passive or active) I never found any problems with theremin interactions. The worst interactions I ever had was when I had a single active HV shield near another unshielded theremin, and the LV version didnt improve matters much. Double shielding entirely eliminated interactions from the LV version, but didnt manage to entirely eliminate it from the HV version.

I am inclined to think that double shielding will be needed to meet EMC standards.


I should just say that the above experimens were spread over a long time - I focussed on one idea, did it to death, left the matter for a long time, then came back and played with some other ides - the organised account above is NOT a reflection of what goes on in my lab - and my lab was trashed 6 months ago so chaos rules and documents / prototypes are in complete dissarray - Everything written above is from memory and a few salvaged scribbles and bits.

You must be logged in to post a reply. Please log in or register for a new account.