Oscillators Tunning

Hi, I just finished building a Theremin using Moog's etherwave schematics. Every module seem to work correctly. I then started to adjust my oscillators to match my antenna's resonant frequency. I did the volume oscillator first. The problem is that I am not able to find a good tuning that would give a decent range.

First, I didn't used exactly the same component. I wasn't able to find the variable coils described in Moog's article. Instead of a 47uH variable coil with a 33uH fix inductance I used a 11.8uH variable coil with a fix inductance of 68uH. I know that my variable coil range is now narrower but I thought I would be able to fix it by adjusting the capacitors used in the tank. I also changed the LM13600 for a LM13700 but it is not important I'm pretty sure ;).

The thing is, I changed the 6800pF for other values and looked at the frequency of the oscillators as well as moving the variable coil from one end to another. I found that if I placed a 1000pF instead, and adjusted the coil or the potentiometer associated with the volume (P1), I was able to invert the direction of the antenna. In other words, at, example, 80% of the coil, the antenna reacts normally (less volume when the hand is closer) (but sadly with a short range of about half a foot) and then at 81%, I abruptly hear the sound going down and the direction of the antenna is now reversed (more volume when the hand is closer).

Additional information : All the above tests are made with an oscilloscope reading the net L7-C12-R17... The frequency at the tipping point is between 694kHz and 730kHz depending on the potentiometer. When I unplug my oscilloscope from the net, the sound disappear and I can't here anything.

My questions are : Is this a normal behavior or can there be an error elsewhere?

Can I find the perfect tuning by simply changing the capacitors for different values?

Is there a way to measure and/or calculate the perfect frequency for my setup?

P.S. I have access to a wide variety of debugging tools including oscilloscope, capacimeter, ohmmeter, etc. (and also my earphones :D) Also, I would prefer not having to change the inductance/coils.

You can't connect a scope to much in a Theremin without throwing everything out of whack, just too sensitive to environmental capacitance (it's what they do).  For the pitch side try clipping a short wire to the scope probe and setting it near the pitch antenna (maybe 6 inches away).  There's usually enough voltage swing to see what's going on by turning the scope sensitivity up, but you need to ground the scope probe to Theremin ground or 60Hz will make it jump around too much.  For the pitch side, tune for maximum voltage and tweak from there.  For the volume side someone else will have to help you.

Theremins, particularly parallel tank with EQ coil types, are weird enough.  Changing component values in a known working design before you get it up an running in the first place is asking for trouble unless you really know what you are doing.

Having a buffer on the oscillators (as on this thread) not only makes the theremin better, but also enables setup / debugging - one can connect a 'scope or frequency counter etc to the buffers output without these loading the oscillators.

Fred.

Thanks both of you for the quick reply!

I was able to observe the oscillations without touching any wire as you said, just by approaching the probe. I didn't thought about that to adjust my oscillators, thanks I will try it today. This means that I might not need to add a buffer, or at least not for debugging purposes. The thing is, I would like to keep the same PCB. So my best approach is to change some of the components.

In that optic, what would be the best option for me? I am not a pro in Theremin at all but I figured most of the necessity of the components present in the circuit. Also, I would like to know if what I described in my first post was a normal behavior.

Thank you

"Also, I would like to know if what I described in my first post was a normal behavior."  - billJohn

For the pitch side it's hard to know if it is capable of normal behavior until you get it at least close to properly tuned.  So play around with it some more and see what you can get.

Not trying to beat you up, but using the specified components gives you some expectation that normal behavior is even possible.  If you deviate from the components then you have to make sure the EQ inductors have low enough self-capacitance (high enough self resonance), and that the tank LC can be tuned to the same frequency as the combined EQ L and antenna C (the antenna C is often around 10pF) .  F=1/(2*pi*sqrt(L*C)).

I got my EW as a "kit" - the PWB was already populated so all I had to do was solder the pots on - and it still gave me a dickens of a time getting the pitch side tuned.

Thanks for the support,

I'll try to work on the pitch oscillators a little more and let the volume on the side for the moment.

The trick of approaching the probe instead of connecting it works great. I even found that at a frequency of 790kHz, if the probe is attached or not to the net described in my first post, the oscillation will or will not have an amplitude high enough to generate sound. This means that the added probe modifies the circuit in a way that enables it to "work". Also, if the probe is disconnected and the capacitor is a 1000pF, there is no tuning in which the volume goes down when my hands approach it, the amplitude of the sine always goes up. The odd behavior described in my first post never shows up, but the amplitude isn't high enough when my hand is far and it never goes down when my hand approaches.

I'll post feedback on the pitch circuit when I'll have more details. In the mean time, if someone as a clue about these behaviors, let me know! Is it normal that it never goes down if it is not tuned properly? I checked my circuit many times and it seems to be fine.

Additional informations: I also tested it on a breadboard before making it on a PCB and the sound was going down when my hand approched it. On the other hand my antennas where little length of wires turn on themselves to look like a coil.

Thanks for the help,
billJohn

When you are tuning the pitch side tank inductor (L5 in the schematic), you should see a broad but definite peak in voltage on the scope.  This tells you that the tank is stimulating the EQ L (L1, L2, L3, L4) and antenna C correctly. 

After adjusting for maximum voltage, you should adjust the pitch side fixed oscillator (L6) so that the pitch goes down to zero beat.  If you can't do this then you don't have enough adjustment range with the fixed oscillator.

Hi, I just finished working on my theremin and I might have found the good frequency for the variable pitch.

I don't know if I found the point you were talking about tough. At points where the voltage was higher, I didn't had any sensitivity. But in my tests I was adjusting L5 when suddenly I realized it changed abruptly of amplitude. I then realized that it was my head bobbing a little that influenced the antenna. I then adjusted L5 until I got the maximum range (distance from antenna) I could find. It is at about 210kHz. At approximately one feet from the antenna I can move a little my hand and it makes the amplitude vary a little, it's magic. Now, when I approach my hand from the pitch antenna, the amplitude goes down, and I think it then inverts (or change phase by 180°) and rises a little. I will work on the fixed pitch oscillator tomorrow.

Thanks for your input, if I can adjust the fixed oscillator at a similar frequency (meaning a frequency of 0Hz after the detector) then I'll give you feedback on this "sensitivity method". Otherwise, I'll search for maximum amplitude and adjust the fixed oscillator again.

It worked! Believe it or not, the sensitivity method worked perfectly.

By changing the fixed pitch oscillators' capacitance, I was able to get a frequency of 0Hz when far away and 3kHz when near. The physical range fits my criteria which is about a foot or a foot and a half.

I am now working on the volume oscillator and the strange behaviors have now disappeared...I'll work more tomorrow.

Also, I might have discovered a way to find easily the resonant frequency of an antenna. The method consist not alimenting the circuit and connecting an oscilloscope probe to the junction L7-C12 and the reference part to the actual reference of the circuit. Then you zoom to the maximum you can (2mV in my case) and you go to a sec/div that makes sense (around 2.5uS/div for a signal of around 400kHz). And if you look closely you will see a small oscillation. This oscillation is, what I think might be, the result of the an excitation of the antenna and the inductance by the 60Hz that "floats" in the air. But the antenna can't oscillate at 60Hz so it reemits it at its resonant frequency. If you want to have a clear look at the frequency without a lot of noise, you can set your oscilloscope to trig on a higher voltage then the noise and set it to a single sequence. This will result in the oscilloscope waiting. You then just have to create a short pulse of electromagnetism near your antenna (but not on it) in which there will be a near infinite amount of frequency. Your antenna will pick up the one it can amplify and it will result in the oscilloscope triggering on a clearer waveform. Try it for yourself and tell me if it works on your theremin or if it's just a random luck that it worked on mine.

Have a good day,
billJohn

"By changing the fixed pitch oscillators' capacitance, I was able to get a frequency of 0Hz when far away and 3kHz when near."  - billJohn

Yes, it seems you didn't have enough adjustment range in the local (fixed) oscillator.

We really could use a simple TW Theremin, something made with readily available components, something you would have to really screw up in order for it not to work.  These parallel tank / series EQ things are just too picky and weird for the average novice builder.