Help, my Theremin is polyphonic.... (Jaycar)

I just completed the build of my Jaycar kit (MK II, 2009) which went entirely without any problem. Plan was to add fat antennas, linearization coil, large enclosure etc. But the thing has a peculiar sound problem which rends it useless as a musical instrument. Instead of a single pitch I get an overlay of two voices. Both change frequency with the pitch antenna, but are not at a fixed interval - throughout the pitch range there are spots of unison and of interference. The problem seems not to be with the volume oscillator as the phenomenon does not change with the volume antenna, which works to change the volume. I also tuned T1 and T3 at different intervals, it does not make a difference.
I'm dumbfounded, flabbergasted ie. totally at a loss. Does anyone have an idea what could cause this, technically? Or is it a flaw in the design? I only found theoretical musings on the Mk II design here, no success stories or sound samples.

Thanks,

Dirk

Hi Dirk,

it's difficult to diagnose this problem at distance, while I'm sure that I could find it within minutes when connecting my oscilloscope. So I have several suggestions for you, since I'm living in Alsace, not too far away from you:

a) You may do a weekend-trip to Colmar and we look together what your instrument has. Or...

b) you send it to me if you don't want to drive so far and I send it back after repair.

In ever case you may contact me by email theremingenieur(at)gmail(dot)com

Wild guess here.
I have seen a similar problem in valve guitar amps when they have far to little power supply filtering.
The 120 Hz ripple (or 100 Hz in Europe, etc) mixes with the signal to produce sum and difference frequency's. Just like your Theremin's reference and pitch oscillators mix to make a tone, but they are far above the audible range.
Just a guess. Good luck........Rob.

This is likely to be some extraneous signal beating with the VFO.. If the VFO / REF oscillators have been retuned and volume oscillator does not seem to be the cause, then it is worth looking for other possible sources of HF..

One of these can be a dodgy voltage regulator.. these can oscillate and produce a huge spectrum of harmonics - some of which could beat with the reference and/or VFO frequencies.

My first suggestion would be - put another 100n good quality ceramic capacitor directly on the voltage regulator pins (solder these capacitors on the underside of the PCB with short leads between 0V and output).

If this doesnt fix it, you need a good HF 'scope - So a visit to Thierry could be your best bet! ;-)

Fred.

In the spirit of checking the simple stuff first...

Be aware that a surge protecting power strip can cause unwanted "ghost tones". There are various electronic components inside these devices that can wear out and go bad. I have experienced this phenomenon myself.

Thanks to all of you! Sometimes it helps being told the obvious :)
I'm happy to report that Rob was right. When hooked up to a trustworthy power source (ie. the 20 pound poured concrete contraption that my father had custom made in his company when I was a child geek and that I never exactly ingratiated him with any thankful attitude for) I get clean sound. Simple as that. I swear there must be a ring modulator in there...
I had tried different wall warts but all were simple transformer type ones of course, as called for.
Now I have one thing or other to talk over with Mr. Sieve. He is a 470uF capacitor which seems to be operative and connected. Shouldn't that be enough??

470 uF sounds a bit light.
IMHO, too much filtering is just enough ;)
Mine only has 6 uF, but it runs on battery's.....

I have not really sorted it out yet. I have tried adding a 2200uF capacitor which does not really change anything, but different PSUs make a difference. It doesn't go away entirely though. Maybe I have to add a small=fast capacitor instead?
Presently I'm waiting for the compulsory towel holder and curtain rod components to replace the soup ladle ;)

[i]"but different PSUs make a difference. It doesn't go away entirely though.[/i]

This sounds like a HF problem rather than a LF one - As in, it is less likely to be due to low frequency (50Hz / 100Hz) ripple on the supply, and more likely to be due to the impedance of the supply rails to HF signals.

If the lowest voltage on the incoming supply is gteater than the regulator voltage + 1.5V, then supply ripple should have an insignificant effect.

I would be looking for an interference signal on the regulated (after the regulator) supply.. Also, check the grounding.. If there is a high Z (inductive or resistive) on the ground line, induced HF can cause problems.

The fact that increasing the smoothing capacitor makes little effect, and the fact that different supplies give different results, makes me think that the regulator IC / circuitry is the most likely cause of the problem.

Strong external HF signals which couple to the supply (or other section) could also be responsible of course - Different PSU impedances could act to change the coupled signal levels.. One common source of interference (I have found the Jaycar circuit particularly prone to this) is these awful economy light bulbs.. They can emit strong HF at the 'right' frequency to produce ghost tones.

[i]" Maybe I have to add a small=fast capacitor instead?"[/i]

I would say this is the first place to start on a problem like this.. Back in the '70s when I was repairing Radios and TV's, the senior tech had a fix for 90% of all the weird problems we saw - Any time I was struggling he would say "stick a point one in!" and he was right - a 100n capacitor or two between supply and ground fixed most 'wierd' problems!

Try a 100n on the input to the regulator, and (probably more important) a 100n on the output of the regulator/s.

Fred.

My neighbour had that kit and I put it together for him. It surprised me by not having any problems and working first time. I do remember something in the instructions regarding modifications to the design relating to switch mode power supplies.