A new vacuum tube theremin

I recently finished building a new vacuum tube theremin that I've been developing over the past year or so.  I'm quite happy about how well it works.  I wrote up an article about it, with photos, schematics, and a brief sound clip.  Here's a link to the article:

A Vacuum Tube Theremin

I hope you find it interesting.

John

John,

Very nicely done. I enjoyed your webpage. I wish we were neighbors instead of just being in the same time zone. Things in the world of theremin have really slowed in the last few year. I myself accomplished what I had set out to find 12 years ago in the theremin, now I'm slowly moving back to some old hobbies. Like writing VB code in Excel to exploit the stock market. Twenty years ago I found that to be the most profitable adventure.

Visit this link which sums up my theremin journey.

I wish you the best of success in all your future projects.

Christopher

Christopher,

Thanks for your kind comments.  I read the thread you posted and listened to your sound clips.  All I can say is, WOW!  That is beautiful tone.  I hope to achieve that kind of sound someday.  I fully agree that lots of 2nd harmonic is crucial, and the asymmetry of your wave shape shows that.  My theremin, as it currently stands, produces more or less a reverse sawtooth wave (sharp leading edge, gradual ramp back down) at low frequencies, changing to something resembling a sine wave at higher frequencies.  I think I can get more 2nd harmonic by changing the bias of the mixer stage and/or the volume modulator; or, if necessary, by adding another stage with the unused half of the volume oscillator's 12AU7A.  It'll take me awhile to get there, though.  Thank you for giving me a holy grail to aim for!

John

Very nice write-up John!  It's quite clear that you know what you are doing when it comes to the theory and implementation, more so than most it seems (that's not a knock at anyone here).  I'm a bit surprised at your reported thermal stability, though I suppose some of this is the large open back letting the heat out?

My own Theremin investigations, though digital, have paralleled yours to a surprising extent, particularly the going back to first principles, hand capacitance measurement and characterization, coil winding, oscillator topologies, etc.  The digital approach allows you to use the period rather than the frequency, which can dramatically linearize the pitch response without the need for bothersome resonant series coils.  It also enables a 2 axis left hand without too much extra trouble, as well as interactive frequency displays.  But digital does remove some of the charm of it all.

I recently finished building a new vacuum tube theremin that I've been developing over the past year or so.  I'm quite happy about how well it works.  I wrote up an article about it, with photos, schematics, and a brief sound clip.  Here's a link to the article:

A Vacuum Tube Theremin

I hope you find it interesting.

John

Very nice INDEED !

Doug

John - that's amazing!  I'm going to post about your article on the news feed so more folks see this, and I've shared it on our Facebook page.  It sounds great.  Have you considered building more?

I was on Lopez Island recently. What a friendly and beautiful place. The rumors were true - people really do wave as you drive by.

Thanks, everybody, for the interest and the kind words!

Dewster, I've been surprised at the stability, too (with respect to both temperature and supply voltage).  One big factor seems to be the cathode-coupled oscillator topology.  I noticed when I breadboarded it that the frequency shifts hardly at all over a very wide range of supply voltages.  I don't know why, exactly.  The topology has enough gain that I was able to use a pretty big capacitor (4700pF) in the LC circuit, so the tube's inter-electrode capacitances are small in comparison.  Also, the power dissipated by each pitch oscillator is low: just a few milliamps each at about 100V.  (In contrast, the Colpitts volume oscillator draws >10mA. at a supply voltage of ~150V.)  So, the pitch oscillators run cool.  Incidentally, the oscillators in Robert Moog's "EM Theremin" in the Feb., 1996, issue of Electronic Musician used a solid state version of the same topology.

Jason, thanks for posting about the article on the news feed.  I am honored!  I hope you'll give me a holler next time you come to Lopez Island.  By the way, I'm afraid you're mistaken about the drivers here waving to everybody.  What we're really doing is practicing our theremin moves ...

John

John, I've noticed quite good stability over voltage for common base Clapp, and believe this spec is important for overall stability.  My favorite is common source Clapp, which oscillates almost no matter what.  Have seen quite low current draw with common emitter Colpitts, 0.3mA including buffer!

That said, I don't believe there is a "magic" oscillator out there that impacts linearity in any significant way, and that the search for it is a canard.

Have you considered air-core for the series coil?  The ferrite cores you are using are likely the main source of whatever small drift you are experiencing.

Dewster,

John, I've noticed quite good stability over voltage for common base Clapp, and believe this spec is important for overall stability.  My favorite is common source Clapp, which oscillates almost no matter what.  Have seen quite low current draw with common emitter Colpitts, 0.3mA including buffer!

I appreciate this advice, and whatever else you feel like telling me.  I'm solid on the theory, but am lacking in practical experience with theremins.  If there's one thing repairing guitar amps has taught me, it's that experience counts for a lot.

I tried out tube Clapp oscillators in the common-anode configuration, which is like common-collector or common-drain in the solid state world.  They worked well, and I almost used them in my first build.  I was using capacitors of 56pF (the "Clapp" capacitor), 680pF, and 2000pF, with a 2.5mH RF choke as the inductor.  With the pitch antenna directly connected to the 56pF capacitor (no antenna coil), I got good range and linearity -- both in simulation and in a real circuit.  But, for various reasons, I ended up going with the cathode-coupled oscillator.

That said, I don't believe there is a "magic" oscillator out there that impacts linearity in any significant way, and that the search for it is a canard.

I agree with that 100%.  The frequency is determined by the total effective L and the total effective C.  The rest of the circuit really doesn't matter.  My simulations taught me that pitch linearity depends on (1) how the capacitance varies with the distance of the hand -- which depends only on the antenna geometry, and (2) the relative sizes of the C or L in the oscillator on the one hand, and the effective C or L of the antenna + hand on the other hand (no pun intended).

Have you considered air-core for the series coil?  The ferrite cores you are using are likely the main source of whatever small drift you are experiencing.

Thanks!  I did not know that, and now you've got me interested.  I chose the RF chokes because of their flexibility.  How do you make it all work with an air-core coil?  Do you just make a coil of about the right inductance and then tweak the oscillator frequencies to make it all work?

One thing I'm concerned about is the self-capacitance of an air-core coil, which is quite a bit larger than that of the RF chokes I'm using.  It seems like the additional capacitance would diminish the sensitivity of the thing to hand movements.  Any thoughts about that?

John

Hi John,

First off, I have to tell you I haven't built much in the way of Theremins, certainly nothing commercial.  Though I've been thinking about them and performing direct experimentation on and off for several years now (yikes!).  Much of my work is here at TW in my digital Theremin thread, though some is in my analog Theremin and Excel simulation threads (all of which have languished while I work on another vanity project).  The threads are here:

http://www.thereminworld.com/Forums/T/28554/lets-design-and-build-a-mostly-digital-theremin

http://www.thereminworld.com/Forums/T/28492/lc-tank-linearizing-coil-excel-simulation

http://www.thereminworld.com/Forums/T/30562/lets-design-and-build-a-simple-analog-theremin

I don't know for sure as I don't know exactly which RF chokes you are using for the series coil [EDIT: on closer look I see they are Hammond 1535D and 1535G] but I'd bet their small physical size and low relative permeability almost guarantees they have more self-capacitance than a well designed & constructed air core (beginning and ending windings near each other are a good indication of self-capacitance, and low permeability means more windings).  I don't have gobs of experience with ferrite, but from the literature it seems the ones that concentrate the field the most have the most temperature dependence.  This can be ameliorated to some degree with an air gap, but the gap dramatically reduces overall permeability, so you might be better off (when forced to use ferrite) going with a lower permeability in the first place.  Axial chokes are inherently air-gapped.

I placed a Bourns 6310 choke in the freezer and measured it as it came up to room temperature, saw +757ppm/C. Things they don't list in the datasheet!  This single experiment pretty much turned me off to ferrite for precision LC work, though I know it shows up regularly in IF circuits, and indeed in Bob M's Theremin designs.

Here is my latest air-core solenoid spreadsheet if you want to play with it:

http://www.mediafire.com/download/s1zk9ya761xlvmc/solenoid_design_2015-04-29.xls

I'm sure you're aware Theremin himself use an air-core solenoid for the series coil - it can be physically quite large, though I suppose that's part of the charm.  Air-core solenoid Q is generally much higher than ferrite core inductor Q.  The aspect ratio influences self-capacitance to some degree, so the Theremin's long solenoids make sense.

I decided long ago not to go down the "linearizing" or "EQ" inductor path because doing so puts you in a tricky place when it comes to aligning the thing due to the double resonance (parallel LC tank in the oscillator, series LC with the series coil and your hand/the environment).  I ownded an EtherWave Standard, so I know first hand how tricky aligning the pitch side can be.  TW member Fred Mundel had schemes to vary the series inductance via core saturation.  Digital period measurement (whether offset heterodyned or direct) lets you completely sidestep all off this and gives much better basic pitch linearity, particularly near the antenna.