Oscillators locking together?

Hi,

I am having some trouble with the pitch of my theremin. I have used a pair of Colpitts type oscillators operating at around 350kHz and have successfully filtered an audible signal from the mixed frequencies. However the audible signal is rather high pitch even at its lowest point. 

I suspect that the oscillators are locking together when the frequency difference between them is slight, and therefore no audible frequency wave is generated upon mixing for low frequencies. Resulting in the signal dying down to nothing at all.

Each of the Colpitts oscillators output is amplified by a Darlington pair of npn transistors prior to mixing. The audio frequency is then filtered out and amplified using an Op-Amp.

I have tried reducing the frequency of the oscillators, which did give a slightly lower pitch at the lowest, but also reduced the adjustment range from hand to aerial.

What can I do to reduce the point at which oscillator locking occurs? 
I feel I may have built the two oscillators too close to one another, can RF interference between the two circuits result in them locking up together?

Or am I completely barking up the wrong tree?

Thanks in advance.

"I suspect that the oscillators are locking together when the frequency difference between them is slight, and therefore no audible frequency wave is generated upon mixing for low frequencies. Resulting in the signal dying down to nothing at all."  - Hinnif

Are you regulating the voltage going to the oscillators?  A three-terminal regulator is the first line of defense when it comes to Theremin stability: it obviously tightly regulates the voltage, but it also provides a low impedance source which reduces oscillator interaction.

The next thing I'd try is filtering the individual oscillator supplies with RC lowpass filters (maybe 10 to 100 ohms in series from VCC, and 0.1uF to ground on the oscillator side) to further reduce supply interaction.

"I feel I may have built the two oscillators too close to one another, can RF interference between the two circuits result in them locking up together?"

Putting some distance between the oscillators can definitely help reduce coupling, as can orienting the coils at right angles to each other (particularly if they are air core coils).

"Each of the Colpitts oscillators output is amplified by a Darlington pair of npn transistors prior to mixing."

A Darlington to buffer is probably overkill, a single bipolar or FET (voltage follower) can get you down to subsonics (if everything else is OK).

[EDIT] I haven't investigated this particular coupling angle, but I have pondered it: If the fixed oscillator has a small tank capacitance (say ~10pF) it could be susceptible to capacitive coupling.  If so, then increasing the fixed oscillator C and reducing the L (to maintain the resonance frequency) would reduce coupling.

Thanks for the prompt reply!

Yes, the voltage supply is regulated, however I think you've hit the nail on the head thinking of supply side interaction.

I have placed the filters you've suggested but still seem to have high frequencies present on the VCC rail. 

I have then checked the earth itself and it has the disturbances present on it as well :/. It looks almost as if it is bouncing. A large sin wave appears, decays to almost nil, and then reappears. 

I have found that the disturbance is fed back into the supply only from the variable oscillator, and only when the aerial is connected, could it be picking it up from the aerial itself?

Also, I suppose my earth is a bit pants if these waves can stand on it?

I don't think the capacitance values should be a problem as they are up nearer 100pF (plus a variable cap on the fixed and the aerial on the variable oscillator). 

The proximity of the circuits is also a problem from what you've said. I have placed them on the same board so I guess I'll have to whip the saw out to seperate them! I'll also reorient the inductors at right angles, good tip.

Thanks again for the speed of your reply, it is much appreciated 👍.

"I have then checked the earth itself and it has the disturbances present on it as well :/. It looks almost as if it is bouncing. A large sin wave appears, decays to almost nil, and then reappears."  - Hinnif

You really need a good ground, if you don't have that then all bets are off and just about anything can happen (I use an old power cord with just the ground pin connected at the plug, and with the other wires stripped out - the USB ground from my PC seems to be sufficient as well).  Fix that and see where you are before you get out the saw.  Oscillators can be relatively close to each other on the same PWB, but distance can only help.

Also: are you absolutely sure the oscillators aren't interacting somehow via their buffers?  If things remain confusing, maybe try disconnecting the buffers and see if the individual oscillators are still locking.  If the PWB trace at the LC high impedance point of the fixed oscillator has significant area, it could be acting like a small antenna / C plate.

Hinnif, I think you are getting excellent advice but why keep us guessing. Post a picture or schematic of what you are doing. There are a dozen ways to use a Colpitts oscillator and 10,000 ways to mess up a theremin design.

Oscillator locking is often misunderstood and has little value as it eliminates the access to the lower octaves.

Are you after good proximity control or a Classic Theremin sound?

Christopher    

Edit: dewster stop being weird, this site has six months left.

"I don't think the capacitance values should be a problem as they are up nearer 100pF (plus a variable cap on the fixed and the aerial on the variable oscillator)."  - Hinnif

That capacitance seems somewhat high?  The intrinsic of the antenna is around 10pF, and the hand in playing position is less than 1pF.  It could be that you're using a smaller than normal difference in frequency to produce the audio beat pitch, which could make the oscillators more prone to locking.