Theremax Optimal Pitch and Volume Oscillator Frequencies

All,

Thanks to a number of you on this forum I have managed to get the noise level of my Theremax to an acceptably unnoticable level [linear regulator, oscillator isolation, extra comparator decoupling, adding comparitor output series resistors, and moving the pitch audio chain away from the comparitor on a separate board] and I am now moving on to coupling and pitch linearization.

My first question is should I work on coupling and tone first and linearity second? - I think so since I have read that coupling changes linearity in the first place but I think that the oscillator frequency will also impact the coupling.  Right now I have copper walls between the oscillators.  I plan to play with fixed caps, varactor diodes, and resistance and maybe even inductance.

I have two pitch antennas, an 18" piece of 3/8" water supply tubing and a LEV antenna.  I have found that the linearity is very dependent on the oscillator frequency and as an example at 820kHz the linearity response of both antennas is the same (C2-C7).  I know that I need to change the 100pf caps on the pitch oscillators to be able to move in the 900 -1000kHz region for the LEV antenna to be in its sweet spot.  I can do that by changing to 68pF but if I do that then I will probably have to change the volume caps as well to keep them far enough apart.  I assume, but don't know that the inductor in series with the volume antenna will only do what it is supposed to do over a limited oscillator frequency range and I think I want to keep the volume oscillators at a higher frequency than the pitch oscillators.  I have 10, 22, 47, 68, and 100pF NPO caps to play with.  I also have a "grab bag" of inductors.

My gut feeling is to change the pitch caps to 68pf and the volume caps to 47 and potentially change the volume inductor.  The IF transforms tune over the better part of an octive at least with 100pF.  (480 - 850 KHz or so).  Maybe I can't move the frequencies too far from the stated resonate frequency though. 

Anyway, if anyone has advice on a good place to start for the two frequencies that would be helpful.

Thanks,

Greg

I don't design analog Theremins so grain of salt and all, but I think I'd put the volume section frequency below that of the pitch section.  You're already running the pitch section rather high (compared to most other Theremins) and are considering pushing it higher.  This means your linearizing inductor can be a smaller value but also must have a higher self-resonant frequency.  Linearity can be a bear, you might want to start wrestling it as soon as possible.  Though you're right, coupling (to change the waveform) can profoundly influence linearity at the low end.

I'd be very interested in your future experiences with the LEV antenna.

Greg said: "I know that I need to change the 100pf caps on the pitch oscillators to be able to move in the 900 -1000kHz region for the LEV antenna to be in its sweet spot."

Antennas are a taboo subject and I could be ban from TW for bringing up my simple opinion and approach to linearity.

I have never worked with a theremin design outside of my own but the Theremax does use a NPN driving a Hartley oscillator similar to my own but our design similarity ends there.

Viewing my 1996 PAiA schematic change C6 & C10 to 68pf (47pf) and C14 & C18 to 100pf. Basically your swapping the oscillator frequencies. I do prefer the 47pf.

If the oscillators still oscillate you're home free!

The Lev Antenna needs about a 1/4" to 1/2" stretch. It has an insulated coating between turns. Using the internal coil support for a more natural look you will lose the linearity within the 1" next to the antenna which is negligible.

The best way to roughly estimate the pitch oscillator frequency is using an AM Radio so you don't load down the circuit for a false reading. 900 kHz to 925 kHz is ideal. The antenna has a broad response band width. High Q oscillators, low Q antenna.

This must be a direct connection to collector of Q1, no inline chokes, no capacitors and no connecting wires near ground. A larger inline capacitor like .01uf is fine. On my tube board I add this to prevent DC shock hazards.

The antenna begins at the circuit board and must leave the board free and clear of any grounded object or linearity tuning will be affected!

To much shielding worries me, maybe just have the IF cans grounded and place a piece of grounded copper clad board under the oscillator circuitry of the board. Copper side away from the board. (-'

I can't help with the noise issue not hearing it but in the Null area I think theremins sense either molecule motion in the antenna or cosmic radiation left over from the Big Bang, more like a course white noise. There is a interesting phenomenon going on here which needs further study. LOL

Possibly your detected audio signal is too weak, for some reason, to rise above all the background noise?

Christopher

I think I'd put the volume section frequency below that of the pitch section.  You're already running the pitch section rather high (compared to most other Theremins) and are considering pushing it higher.

I figured if the volume oscillator frequency was higher than the pitch frequency it would be easier to tune the two to avoid interference with the fundamental or any of the harmonics or other intermodulation products.  Since a few people have suggested to basically swap the caps I'll try that first.

I'd be very interested in your future experiences with the LEV antenna.

I'll post my results - I was thinking of measuring the linearity of both antennas for different oscillator frequencies. 

Christopher Said:  (I don't know how to actually use the Reply with quote button or the forum in general...)  Lots of good comments - I put Christopher's in italic - [my responses below in brackets]

Viewing my 1996 PAiA schematic change C6 & C10 to 68pf (47pf) and C14 & C18 to 100pf. Basically your swapping the oscillator frequencies. I do prefer the 47pf.

If the oscillators still oscillate you're home free!

[I'll try it with 47pF and see what happens]

The Lev Antenna needs about a 1/4" to 1/2" stretch. It has an insulated coating between turns. Using the internal coil support for a more natural look you will lose the linearity within the 1" next to the antenna which is negligible.

[I haven't seen anything on the internal coil support - must not be looking in the right place.  Right now my implementation has a /2" stretch and is mounted inside a PVC pipe that screws on same 3/8" compression fitting as the "normal" antenna - basically shoved the compression nut in the bottom of the tube as suggested on the LEV page I found but had to carve out the tube a little to get the nut in.]

The best way to roughly estimate the pitch oscillator frequency is using an AM Radio so you don't load down the circuit for a false reading. 900khz to 950khz is ideal. The antenna has a broad response band width. High Q oscillators, low Q antenna.

[I do two things right now.  First I put the scope probe directly on the oscillator collector (which loads it and reduces the frequency) just to make sure it is swinging 12V p-p.  Then I probe the collector of Q8 - the beat frequency amplifier and find the two oscillator frequencies with the scope FFT display.  I'm not sure I have an AM radio except in the car which I used to see if there were any strong stations in the area...]

This must be a direct connection to collector of Q1, no inline chokes or capacitors. A larger inline capacitor like .01uf is fine. On my tube board I add this to prevent DC shock hazards.

The antenna begins at the circuit board and must leave the board free and clear of any grounded object or linearity will be affected! The antenna and the oscillator circuit behave as two separate components trying to find harmony.

To much shielding worries me, maybe just have the IF cans grounded and place a piece of grounded copper clad board under the oscillator circuitry of the board. Copper side away from the board. (-'

[The antenna connection is through about 6 inches of coax to the compression fitting - I'll take some pictures and post them - http://www.mediafire.com/?3s8w8dd53bokp20 contains a PDF which shows the antenna connections but not the isolation walls]

I can't help with the noise issue not hearing it but in the Null area I think theremins sense either molecule motion in the antenna or cosmic radiation left over from the Big Bang, more like a course white noise. There is a interesting phenomenon going on here which needs further study. LOL

Possibly your detected audio signal is too weak, for some reason, to rise above all the background noise?

[For the Theremax, the biggest noise issue is introduced by the comparitors used in the pitch and volume CV generation.  The open collector output drives too fast and is directly connected to a cap so there are large current spikes and high edge rates.  After I improved the power supply I found that if I manually drove the VCA so there was always an output signal most of the noise issues including "radio noise" went away when that volume comparitor was not switching on and off.  It didn't matter what the volume oscillators (assuming they were tuned to a frequency that did not interact with the pitch oscillators) were doing or the beat frequency as long as it was post-audio - it was the comparitor output noise which many have mentioned.  I'll post a rundown what I did in that area]

If you found the Lev Antenna why did you not look further into what else that website was giving hints about. RS Theremin has a new direction for his theremin research and may be pulling down the info to focus on something very special.

[I'll see if I missed something...]

"You can bring a horse to water but you can't make him drink."

Christopher

gnsmith116 said: [The antenna connection is through about 6 inches of coax to the compression fitting - I'll take some pictures and post them - http://www.mediafire.com/?3s8w8dd53bokp20 contains a PDF which shows the antenna connections but not the isolation walls]

Your mention of coax set off a red flag, I need to see what you are doing, hopefully not the antenna wire running through a grounded shield, not good.

I found no cooperation on that website, they have to many angles to put software on my computer when a simple PDF link is all that's needed.

Email the file to me and I will post a link. I have Gigabytes of online storage with no commercials.

theremin@oldtemecula.com

Christopher

"Antennas are a taboo subject and I could be ban from TW for bringing up my simple opinion and approach to linearity." - Christopher

Really? Banned? - Are there dark forces at work.. ? I dont really see the hosts of TW banning anyone for talking about antennas Chris - or at least, not unless there was some powerful incentive 'forcing' them to do so.

AFAIK, there have been people banned from TW - but only for breach of the conditions - which, AFAIK, do not exclude tabbos like antenna constructions ;-)

LOL - I just realized - Its your sense of humour again !!! - Gets me every time ;-)

Anyone wanting a repository for anything technical related to theremins can post files here:

Element-14 Theremin Document Page

Or start a "discussion" here:

Element-14 Theremin Discussions Page

where one can upload up to 5 items per posting.

The big advantage of Element 14 is that it is massive, hosted by one of the largest electronics component suppliers, and does no funny business (no pop-ups, adverts, no "downloaders" required, no software installed on your PC) and it is reliable and completely free.

Every picture I link to from TW is hosted there, as are any links to .pdf or whatever.

Fred.

Christopher Said [Your mention of coax set off a red flag, I need to see what you are doing, hopefully not the antenna wire running through a grounded shield, not good.]

It's a Theremax - of course it is running through a grounded shield!  Maybe that should be changed.  I'll get you pictures later tonight.

gnsmith said: "it's a Theremax - of course it is running through a grounded shield!  Maybe that should be changed.  I'll get you pictures later tonight."

I never knew of this coax as I never owned a Theremax. Could this be why many may have failed to experience the phenomenon of linearity and were too kind to report back to me. 

Oh my. . .

A direct connection to the pitch oscillator at 900 kHz is the point of magic. I admired John S, there must be a reason I am not aware. Digital noise maybe, what a nightmare!

A 3/8" wood dowel is inserted in the coiled spring antenna. A washer is slipped in between coils at the top and a drilled hole at the bottom for a nail to slip through. The wood extends 6" beyond the bottom for mounting. I used a grinder to remove the loops on each end of the spring.

Perfect pitch field linearity is either on or off, there is no it seems like it is getting better!

This worked on my vacuum tube designs and every transistor version as long as I got just over 900 kHz.

Christopher

"This thread is about to roll over, so don't forget me!"

All,

I have uploaded 3 really rough docs with lots of pictures and some text to Element 14 as Fred suggested.  I certainly hope to refine them.  I'm not sure where on Element 14 I would put just a single pic.  Anyway these are the paths:

http://www.element14.com/community/docs/DOC-48675/l/theremax-base-build-characteristicspdf

http://www.element14.com/community/docs/DOC-48676/l/theremaxbuildpremodpdf

http://www.element14.com/community/docs/DOC-48677/l/theremax-digital-noise-mitigation-and-lev-antenna-build-detailspdf

My apologies about the other service - I didn't realize they made you download anything.

Greg