GT Theremin

> You may find that lowering this frequency (increase L) to about 200kHz will improve linearity and extend the sensing field

Just a guess, but lowering the nominal pitch oscillator frequencies should reduce the pitch sensitivity somewhat. This could lead to the greater linearity you describe since you're moving over a smaller frequency range where the inherently nonlinear frequency-reactance relationship for capacitors is less noticeable. The early versions of my pitch oscillators ran around 700 kHz, but I found them a bit too sensitive and backed off the frequency to 500 kHz.

As to extending the sense distance, I have no idea there. I would assume that the geometry of the antenna and its placement would have the greatest effect on the sense distance since the interaction method is capacitive. Reducing the coupling between the pitch oscillators might also help as doing so lowers their ability to lock on to one another at the very close frequencies implied by greater hand distance. Then again, I haven't done much investigation into optimizing the sense distance, so my assumptions could be off.

If memory serves, I observed a bit longer-range sense distance in the prototype, which had the antenna set on a wood stand relatively far away from any ground planes. In my mind this gives some credence to the dependence of sense distance on antenna geometry/placement. Originally I had wanted to use a slightly larger enclosure than this one; you can see how tightly everything is packed and imagine how carefully it had to be laid out and built. However, we needed to build something fairly quickly and found a box previously used for another purpose. If I did it again I think I'd try to move the antennas farther away from the active circuitry and any strong grounds.

That's not to say it's unplayable in its current form. I can stumble through some simple motifs, limited more by my own lack of playing ability than the instrument (this is the first Theremin I've actually used). I think there's certainly room for improvement, but using aerial fingering I can pretty easily play some scales without feeling too crampt. Of course, you should take my word with a grain of salt since I have played no other Theremins to compare against...

> I especially love the RCA tube right up front. It's like a nod to the original RCA theremins of yesteryear. Fantastic!!

The 6AL5 twin diode (the tiny tube in the back on the left) is also an RCA tube, and I believe the 12AX7 may be a late-40s or early-50s RCA as well from its construction. The 12AU7s (ECC82s) are all new tubes made by JJ/Tesla. We have several drawers full of tubes here which probably haven't been touched since the 1960s (the cabinets are at least that old). I only wish I could've found a use for the very cool-looking 6AW8A triode/pentode tube I found which had half-exposed grids in the pentode section.

Thanks for the kind remarks, everyone.

I went ahead and recorded a quick sound test (http://mattb.alwayssleeping.com/files/theremin_test_1.mp3) with the GT Theremin. Sorry for the "sound effect" content, but I was just messing around and I'm too unskilled as of yet to play much beyond a botched "Mary Had a Little Lamb". :P

What a great design with that shine of electron tube coolness all about it! And, it has a lovely voice that seems quite controllable from the posted soundfile. I hope that you will aspire to play and submit some other soundfiles in the future.

Good Luck!

teslatheremin

Here's a demo video (http://www.youtube.com/watch?v=hZUk8hKIUjU) we recorded for introducing the GT Theremin to folks who will be seeing it here. You already know the information in the video, but you might like the GT Theremin in action.

Excellent video demo! I love how the LEDs come up with the supply voltage, then the big RCA tube fires.

You might want to add a little "don't try this at home" disclaimer about opening up the chassis with power applied! The 250V or so rail voltage is enough to knock you on your rear, scraping forearms on sharp sheet metal edges in the process. And they tell you not to touch a hot circuit with both hands at the same time because the current can flow from one arm, through your chest and out the other ... a potentially lethal thing even with low currents. (Been there, done that on an old radio years ago. It scared the hell out of me at the time!)

We have all gotten very lax with all these low-voltage solid-state things. The HV on old TV sets used to be the worst ... the shock was bad enough, but in the jerk away from the thing, you invariable ripped your arm open on something on the chassis.

Don

> You might want to add a little "don't try this at home" disclaimer about opening up the chassis with power applied!

Good idea. Done. I'm actually only running the HV supply at 75 V (the voltage drop over that glow discharge tube), so the incoming line voltage is higher. In reality everything inside the case is so thoroughly heat-shrinked that you would have to actively try to shock yourself. There are only a couple of exposed points where high voltage is present, and my hands weren't near them. However, you're right that we get a little bit lazier than we should when we've worked around this stuff for awhile.

That's not to say I've never been zapped... I got hit by a charged 160 V power supply filter cap a couple months back, which was rather unpleasant.

Thanks for the tip!

Good tubes! I'd love to see them in a box chosen for looks rather than expediency.

I like the timbre. Classic. (Seems to be a bit of an extra tone in the 12+ kHz range as you get closer to the pitch antenna on the video - between 2:41 and 2:43. Not sure if that's a computer related artefact.)

Is that a Fender Frontman 15R? That's what I use. :-)

A thought about the pitch field - 12 inches. Is that by any chance the same as the distance between the pitch rod and volume loop? You need to have your hand over the loop to play, and when it is the nearer hand to the pitch antenna it determines the pitch, not your pitch hand. When you had the antennas on a stand and noticed a larger field, how far apart were they?

Also a question. Is there a particular reason for having the volume rise as you lower your hand, rather than the other way around?

And, seriously, thank you for doing Hoffmanesque sound effects rather than playing a tune badly. My ears appreciate it.

> Good tubes! I'd love to see them in a box chosen for looks rather than expediency.

Thanks! Oh I had some pretty grandiose plans at the outset, but time and money are factors... Actually I'm fairly pleased with the way it came out despite the recycled box. It would've been nice to find some hinges in the right color and perhaps to have used brushed aluminum (or at least less scratched up), but I sort of dig the look we came up with.

> (Seems to be a bit of an extra tone in the 12+ kHz range as you get closer to the pitch antenna on the video - between 2:41 and 2:43. Not sure if that's a computer related artefact.)

Hm, probably just the poor mic on the camera at work... The pitch circuitry shouldn't be capable of producing a tone quite that high. You can hear a recording straight to my sound card here (http://mattb.alwayssleeping.com/files/theremin_test_1.mp3) which is better representative of the real output sound. If you can still hear it then we have a mystery on our hands and I'll have to go get my good headphones rather than relying on crummy laptop speakers which probably cannot produce a 12 kHz tone.

> Is that a Fender Frontman 15R? That's what I use. :-)

It's a 15G; no reverb circuit. It actually belongs to a friend. Interestingly I get a little bit of additional distortion using this amp as opposed to more "transparent" amps intended for music reproduction, even with the gain knobs as low as possible.

> A thought about the pitch field - 12 inches. Is that by any chance the same as the distance between the pitch rod and volume loop?

The 12 inches was just a guess. I think it's actually a little longer when tuned properly. I ought to measure it. Actually I think the distance between the rod and loop is less than 12 inches. I doubt there's a lot of electrical field between them due to their orientation and the strong ground near by, but it's possible that their proximity is partially to blame for the sense range. I'm more inclined to simply blame the big grounded box, however, since the output of the volume and variable pitch oscillators are very stable, and I can typically see phase noise on the scope if there is a coupling problem between them.

> When you had the antennas on a stand and noticed a larger field, how far apart were they?

I never could run the entire Theremin in its bread board form! Too much parasitic coupling! So I never had the pitch and volume sections even hooked up to each other until I built the current box. With only the pitch section running, however, the antenna stood a bit over a foot away from the circuitry and anything else.

> Also a question. Is there a particular reason for having the volume rise as you lower your hand, rather than the other way around?

Yes. I ran out of time. I was working with VCA designs for a couple of days, but by that time my team mate had begun construction of the box and I had already committed to a certain number of vacuum tubes. So I had limited leeway in what I could do circuit-wise to get the more traditional volume behavior. I finally decided to just go with this topology because it worked quite well. Actually, from a playability standpoint, I think this has some advantages because the instrument is silent until approached, allowing me to step back from it without having to engage the mute switch. The backwards direction would probably confuse long-time Thereminists, but I think it's just as usable as the opposite direction.

If I wanted to modify the design for the more traditional volume control direction, I might need to opt for a more clever method since I'm unsure whether my current VCA topology could be easily tweaked to work. Some possibilities:

* Increase the control voltage (a lot) by properly impedance-matching the ceramic detector and replace the VCA with a pentode. Would require a custom-wound transformer or two.

* Do something cr

[i]the antenna stood a bit over a foot away from the circuitry and anything else[/i]

I'm curious about that. I don't know much at all about electronics, but I understand that the wire leading to the antenna is part of the antenna, so having a long wire would affect the circuit. I'm thinking about my TV - I have moved it several times and lengthened the lead from the aerial to the TV without having to retune it. Is that because the lead is coaxial - shielding the wire from external interference? Would that work with a theremin? It'd be neat to just have the antennas on a stand for playing, with the circuitry in a separate box.

> the wire leading to the antenna is part of the antenna

Well yes and no. First of all, understand that Theremin antennas aren't antennas in the RF sense; radio antennas are usually built to an exact geometry according to the frequency they operate at. Theremin antennas are just half of a capacitor (and you are the other half). Radio antennas are usually fed by a waveguide of some sort, like coaxial cable. The waveguide is designed to contain the RF energy rather than radiate it, so it really isn't part of the antenna it feeds. On the other hand, with Theremin antennas acting like capacitors, the wire connecting them could be considered as an extension. Really, any wire connected to the oscillator circuits serve a similar purpose to the actual antenna; moving them around will change their parasitic capacitance somewhat and alter the oscillator frequency (like your hand does).

> I have moved it several times and lengthened the lead from the aerial to the TV without having to retune it. Is that because the lead is coaxial - shielding the wire from external interference?

Partially. The coaxial cable does a pretty good job of containing the RF signal from the antenna without radiating it, and the cylindrical outer conductor screens external electric fields. The bigger reason is that a property of the coaxial cable called its characteristic impedance is matched to the TV and antenna. When transmission lines are properly matched to their source and load, the length doesn't matter.

> Would that work with a theremin?

Again, Theremin antennas aren't RF antennas, so impedance matching them isn't of much benefit. Using a shielded cable like coaxial should prevent the line from being sensitive to your hand capacitance, but changing the length of the line will change the frequency of the oscillator it's connected to. So sure, you could connect the antennas with shielded cables, but the Theremin would have to be tuned for the particular length of cable you're using.