Looking at the EM theremin schematics (https://www.cs.nmsu.edu/~rth/EMTheremin.pdf) I found that it has very high capacitance in its pitch oscillator (3900 pF). I know that antenna capacitance changes only by few pF when one plays theremin. So the reasonable question is how such small relative changes of capacitance (about 1/1000) can change oscillator frequency by 5000 Hz? I suppose that the trick is in using L1,L2,L3,L4 10mH coils. If yes, what physical effect is it and how can I choose similar coils for my own oscillator design? Can anybody write any math considerations?
High capacitance oscillators in EM theremin
Posted: 1/28/2021 1:57:18 AM
The high capacitance is just in the stimulus oscillator, and the series coils transform the impedance, giving much higher C sensitivity.
I recommend you play with ILYA's very excellent Theremin Explorer software:
Posted: 1/28/2021 9:30:57 PM
You can also use this formulas, change the values.
http://thereminworld.com/forums/T/31
Posted: 1/28/2021 9:32:16 PM
"This is a really interesting software!" - andreyrogatkin
It is! If you're interested further, I collected data of my own arm, as well as finite element analysis of a virtual arm. If you go at things from a C-sensor standpoint with digital processing you can get remarkably linear response, way better than analog approaches, and you can pretty much kill mains hum too. But you have to start somewhere, and real data and/or accurate sim data is the best place, overly simplistic closed forms are the worst place (led me astray for years).
Posted: 1/30/2021 1:22:51 PM
"what physical effect is it" - andreyrogatkin
The physical effect is "coupling resonance circuits" ("связанные колебательные цепи" for russian)
A sensitivity of ones in EM/EW/RCA etc schematics (in terms of df/dC) is almost the same as of single LC circuit with small C (well, a bit worse)
Interesting phenomenon of coupling circuits is bifurcation (doubling) of resonant peak even the individual resonant frequencies are exactly equal to each other.
Theremins (in linearization purpose) utilise the upper peak, which has a limit in movement. The upper peak have to be stronger than lower one - to prevent oscillation at improper frequency. It's possible only when the individual resonance frequency of antenna circuit is slightly lower than individual resonance frequency of tank. Otherwise, the frequency jumps from peak to peak when the hand is moving.
"Can anybody write any math considerations?" - andreyrogatkin
Refer http://antiradio.narod.ru/theremin/etherway/index.htm
The coupling coefficient k=sqrt(L5/(L5+L1...L4) in EW is 0.05...0.06.
Just keep it.
You must be logged in to post a reply. Please log in or register for a new account.