Just built Etherwave doesn't work

1.) They sum up both pitch oscillators' signals for additive mixing over the demodulator diode D4. Since the latter has an exponential U/I curve, the result is a multiplication of both signals. e^(a+b) = e^a * e^b

2.) When nothing is in proximity of the volume loop, the series resonant circuit formed by the static loop capacitance and L7 to L10 is in resonance with the volume oscillator. Thus a resonant current flows through L7 to L10. The current which flows through L7 makes that a voltage proportional to this current will appear between both ends. This voltage is rectified by D1, C12 acts as a load capacitor. A simple, classic one-way rectifier...

3.) D4 acts as mixer/demodulator diode (see my answer to question n° 1). R23 and R24 form together the load for the demodulated signal and a voltage divider to adapt the signal level to the input of U-3b. At the same time, R23, R24 and C23 form a low pass filter to eliminate remaining RF from the demodulated signal.

4.) As I stated in one of my above postings, Hammond 153x series or Miller 630x series, values as in the above schematics. Some more recent industrially produced Etherwave Theremins use only three coils on each side. This requires a much more sophisticated circuit board layout with less parasitic capacitance. When building the circuit "by hand", the use of the original 4 coils on each side is strongly recommended for stable operation and easy tuning. Once everything works perfectly, you may still experiment and short L10 and L4.

5.) Big electrolytic capacitors are good for buffering DC but they have very bad RF properties (high parasitic series inductance). Thus C29 and C30 (ceramic or tantalum capacitors, not electrolytic!) are added to form a virtual ground path for the RF and preventing the integrated voltage regulators from oscillating themselves, causing interference or heating up.

6.) The third schematic with C10 and C11 = 100uF and C11 to ground is much more stable. The older variant with 0.47u or 1u capacitors and C11 to +12V was less stable and had the tendency to make the reactance tuning stage around Q5 oscillate itself at around 330Hz under some conditions.

In general, you should understand that all RF components in a theremin are much more critical than in a radio or tv set. Thus you should always use the indicated and approved components and not try to do wild replacements by parts with "almost similar" values and properties, because the latter will almost always give unstable operation and/or unpredictable results.

Hi all.

Finally I built a new circuit board with the original components. But it's still unworkable. Exactly pair volume oscillator + volume antenna circuit is unworkable.

I built the pair according to schema 2 from my first post of the thread. I used Toko 154ANS-T1017Z as L11 and J.W.Miller 70F225AI (22uH) as L14.

Voltages at emitters of Q2, Q4, and Q7 are exactly the same as on schematic diagram.

Using an oscilloscope I checked output signals of oscillators. It's clean sinusoidal signal on every oscillator's output. Changing values of L5, L6, L11, P2 and P1 I registered the following frequencies:

1) Variable pitch oscillator: 277.8 - 307.8 KHz.

2) Fixed pitch oscillator: 271.7 - 307.8 KHz.

3) Volume oscillator: 455.9 - 571.4 KHz.

Looks like the frequency of volume oscillator is too high because in HotRodding document stated that the designed(expected) frequency should be about 450KHz. So it doesn't correspond to L7-L10 SRF and a current flowing throw L7 is too small.

Therefore I need to shift the frequency to make it lower.

1) What (and how much) would you suggest me to correct in the oscillator to shift the frequency to a normal rate.

2) For possible future cases the same question about shifting frequency of variable pitch oscillator and fixed pitch oscillator.

3) If ground pin 12 of U3 then we can hear the loudest sound of theremin. I checked sensitivity of pitch antenna and found out that bringing hand near the antenna changes frequency not too much as in industrial Etherwave from a video from Internet. I suppose it's ok for self-built theremin. But how can I increase the sensitivity of the antenna? I mean can be there any solutions excluding removing any metallic stuffs (not connected to the antenna) near the antenna and so on?

Thanks.

About correcting volume oscillator's frequency... I wouldn't like to change L14. Is there any another way to resolve the problem?

On photos from Internet I saw that some axial inductor is used as L14 in an official kit. Would it be too critical to use any another (not J.W.Miller) inductor there? I think that 22uH or 33uH is not too big value for that. What do you think?

I mistook about registered variable pitch oscillator above. It changes within 277.8 - 322.8 KHz on my device.

Anyway I think that frequencies for both pitch oscillators are correct.

Make sure that the space around the volume loop is clear (to not to mute it accidentally) and measure the voltage delivered by D1 towards ground while tuning L11 from one end to the other. While the voltage will be most times 0V, it should go down to -4V when there is resonance between the volume oscillator and the volume antenna circuit. If you can't get this resonance peak, try replacing L10 by a short wire to raise the resonant frequency of the antenna circuit. At least now, it should work.

That has nothing to do with the SRF of L7 to L10. The SRF is only important for the "out of resonance" behavior (field geometry). The elementary resonance of the volume antenna circuit is given by L7+L8+L9[+L10] and the static capacitance of the volume loop towards environment/ground (ca. 7 - 8pF)

BTW: Maximum volume is not obtained when U3/P12 is grounded, but -6db. At the moment when you have the -4V at D1, you should find almost +12V at U3/P12 which will give max volume.

Ok. Thanks. I'll try it and post results later.

But why exactly L10? Would it be so critical to replace L8 by short wire? That will adjust the SRF to 490-500 KHz. If we replace L10 the SRF would be about 550 KHz (according to my calculations).

It's not a problem for me but just interesting. Do I correctly understand that these big inductors make a EM impact from one to another? And replacing one from a middle (not from the end) of the series would affect it by someway.

The effective SRF is always lower than the calculated one

a) because of the individual SRF of the single coils which increases the inductive component slightly. Knowing that the reactance of a parallel resonant circuit (coil's inductance with parasitic capacitance) is infinite at its own SRF and decreases with the frequency towards 2 * pi * f * L when f << SRF, you'll find that the effective inductance can be between L0 and infinite, and even become capacitive when f > SRF.

b) magnetic coupling between the coils will further lower the resonant frequency of the antenna circuit and should be avoided. That's one of the reasons why moog slightly raised the oscillators' frequencies and uses only 3 coils on each side in a | — | geometry instead of 4 in a |||| geometry (which was an idiocy from the beginning and explains why early Etherwaves had much worse linearity than recent ones).

Best example is the RCA theremin. There are two linearization coils in series, a 11mH one and a 39mH one. Thanks to their parasitic capacitances and magnetic coupling between both (one inside the other) the effective inductance seen by the oscillator is > 80mH...

When I told you to bridge or replace L10, then it was thought as a temporary idea, just to create a significative jump in the SRF, but not too much, so that we see if you can now obtain resonance, and if yes, at which frequency, to be sure that there is no other problem with your circuit. It was not intended as a definitive circuit modification.

Don't force me to explain always each of my suggestions, most are intuitive and a result from many years of practical experience with Etherwave circuits...

Well, I replaced L10 with a short wire. But now I can see a new unexpected behaviour. Unfortunately...

When I just replaced L10 with a short wire and powered the circuit on, it was workable! On frequency 545 KHz it was fine reacting for my bringing hand to the volume antenna. It was muting when I was bringing my hand close to antenna and was becoming louder when I was moving my hand out. The loudness was changing regularly and right on videos of industrial Etherwaves from Internet. But not for long...

It was ok for about 10-20 seconds. For this time the sound was regularly becoming commonly silenter and silenter and after these 10-20 seconds the sound was muted completely. And no more reaction for moving my hand near the antenna. Nothing could help: not adjusting of L11 or any of potentiometers, not power off + power on. The Eherewave was keeping silent.

If I ground pin 12 of U3 then the sound become loud, and reacting for moving my hand near the pitch antenna.

I checked all connections and the set of potentiometers - not excess connections. But the behaviour is still the same.

On next day I tried to power on the circuit board again. And just after I powered on it became fine workable again. But not for a long as well. After about 30 seconds it became silent. But there was some reaction for bringing my hand near the volume antenna. If I was keep my hand about 10-15 centimeeters over the antenna the strange hoarce sound appears. It was like usual sound but horce and sawing. And no adjustments could change the bahaviour. But if I bring my finger near any ground of it's power supply then the horce sawing sound disappears. Also If I try to measure any voltage to ground then the sound disappears too. Some phantom...

I checked the voltage on anode of D1 - it's continuously about -0.6 volts. On anode if D1 os about 0 volts.

All signals from oscillators have a correct form and frequencies as described in another my post above.

Actually I meet the strange behaviour not the first time. When I built my Etherwave last time with "wrong" components, there was the same effect. And it's completely brings my into deadlock. I don't understand what makes the sound to be silenter and silenter when I powered on the circuit board. And why does it keep silent if volume oscillator generates the same frequency signal as L7-L9 SRF? What the strange effect there?

What do you think?

So voltage of pin 13 of U3 doesn't change when I bring my hand near the volume antenna. Very stange for me...

It looks like the volume oscillator is running away... Connect a frequency counter to the base of Q7 before powering on and observe the frequency during these 30 seconds if it really keeps constant or if it drifts away.

I suspect that you have a problem in the reactance stage around Q8. This would explain the too high initial frequency of the volume oscillator, the instability and the "ghost tone" problem. Connect an oscilloscope across C17 once and then across C18. In both cases you should only find DC, no RF and no AF. You can try to use bigger capacitors (22 or 47uF instead of 1uF) to make this stage (which tends sometimes to oscillate by itself) more stable.

But first of all, I'd "wash" the bottom side of the circuit board with 90% alcohol or isopropanol to eliminate these tons of brown resin residue which often causes trouble.