Theremin World Recent Forum Posts

Big issues with Etherwave after a few months of non-use and a move (by JPascal)

That's really good news! Unfortunately, grounding issues and hum problems seem to be constant companions of the theremin: It's great that you coud fix this.

Big issues with Etherwave after a few months of non-use and a move (by nclm)

Three years later and another move: it wasn’t the apartment, it was a dry solder joint in the theremin preventing it to be grounded properly! Now fixed, ready to start theremining again

Moog Etherwave repair and acquiring variable inductor or make modern replacement (by FatherCondor)

Hahaha, any chance you recently got rid of that at a music store in Georgia? I just picked one up with the exact same issue, and I'm having the exact same experience you're describing. I'm trying to find replacement SLOT TEN-5-10's and really struggling. I suppose if anyone here snagged some up in bulk and has any up for grabs let us know!

I found someone on ebay selling some SLOT TEN-5-11's and looking at this Coilcraft data sheet I'm hoping they have enough overlap in performance to get the volume circuit into the right range. I might also experiment with some fixed value inductors to see if I can at least get the thing working, I'll post an update here if I have any decent results with either approach.

As an after note, I'm kicking myself for hamfistedly removing the component from the PCB after realizing it was broken, I think in our case there is a world where the ferrite core can be backed out and replaced with a core from a similar component in the coilcraft lineup (not necessarily a hard-to-find TEN-5-10, could maybe work with TEN-3-14 core). I wound up pulling the casing off and breaking the magnetic jacket surrounding the coil in the process, so hopefully others can learn from my particular mistake. I should have been more gentle and thought a little bit more before jumping in with a soldering iron and some pliers. Lesson learned!

Let's Design and Build a (mostly) Digital Theremin! (by dewster)

"In the D-Lev, there is no heterodyning, just the RF antennas and capacitors controlling an oscillator (as the analog stage) and from the subtraction of the oscillator frequencies, everything is digital, yes ?" - skrishna

It's true that there is no heterodyning in the D-Lev. The electric field generation oscillators are almost entirely digital, and spit out frequency numbers at a 48kHz rate (the audio sampling frequency). A constant value is subtracted from these numbers, which I suppose can be seen as a sort of static heterodyning, but unlike classic non-linear mixing of two waveforms, it doesn't generate additive frequency components that need to be removed later via low pass filtering.

"So then since the D-Lev does try to roughly sound like an analog theremin (but with a difference), the rest of the circuit somehow shapes the difference frequency to "catch up" with the harmonic content ?"

The numerical frequency difference is further processed to linearize, size, and position the fields. The final pitch and volume numbers are then sent to the DSP synth, where the oscillator and filter settings control the timbre. The synth can also use those numbers as modulation inputs to the filters and such.

Let's Design and Build a (mostly) Digital Theremin! (by skrishna)

"I don't have much analog Theremin design experience, but the heterodyning stage is where you get the harmonic content, so you might try modeling that somehow.

In the D-Lev, there is no heterodyning, just the RF antennas and capacitors controlling an oscillator (as the analog stage) and from the subtraction of the oscillator frequencies, everything is digital, yes ? So then since the D-Lev does try to roughly sound like an analog theremin (but with a difference), the rest of the circuit somehow shapes the difference frequency to "catch up" with the harmonic content ?

Or am I getting it (somewhat) wrong ?

Thank you !

Let's Design and Build a (mostly) Digital Theremin! (by skrishna)

"I don't have much analog Theremin design experience, but the heterodyning stage is where you get the harmonic content, so you might try modeling that somehow.

Or am I getting it (somewhat) wrong ?

Thank you !

Let's Design and Build a (mostly) Digital Theremin! (by dewster)

"But it also occured to me that rather than training those models, perhaps one could physically model the electronics from first principles - since it is a static system with known components (after the RF antenna stage), this modeling would have to be "simpler" than modeling a violin, for example (famous last words)." - skrishna

Definitely simpler than a violin!

I don't have much analog Theremin design experience, but the heterodyning stage is where you get the harmonic content, so you might try modeling that somehow. You may already be aware of this, but when naively generating digital samples you will also be generating aliasing, and it can be a difficult thing to eliminate, or at least sufficiently suppress.

Let's Design and Build a (mostly) Digital Theremin! (by skrishna)

Hello,

I have a low-latency camera-based hand-tracking system. I would like to use the keypoints to drive a theremin. I am working with some fairly accomplished thereminists. While I can program and know the basics of DSP processing, I do not have much actual experience.

To recreate the theremin sound, I am currently considering an auto-encoder based approach (RAVE or Magenta-DDSP). But it also occured to me that rather than training those models, perhaps one could physically model the electronics from first principles - since it is a static system with known components (after the RF antenna stage), this modeling would have to be "simpler" than modeling a violin, for example (famous last words).

That brought me to the D-Lev, whose components are all very well described in the manual. I imagine that this would be a good starting point for such a software-based system. But is there such software already ? Any pointers on how one could go about it ?

Thank you !

This question is related to: Theremin World - Topic: Modelling the theremin wave in software.

Let's Design and Build a (mostly) Digital Theremin! (by dewster)

"It’s strange that the spiral shield didn’t have any effect. What was the pitch of the spiral?" - ILYA

One was a very loose spiral: http://www.thereminworld.com/forums/T/28554?post=224158#224158

This one was tighter:

And I did some sparse grid testing here: http://www.thereminworld.com/forums/T/28554?post=222469#222469

Can't say it was all a dud, and can't say that I've exhausted the experimental space, but the initial gains seem so marginal that the trade-off of plate closeness and reduced sensitivity probably isn't worth it. Even if the kit builder goes for rods, the "hot" end - indeed the entire coil - benefits from some air around it, which the coil mounts provide.

Let's Design and Build a (mostly) Digital Theremin! (by ILYA)

It’s strange that the spiral shield didn’t have any effect. What was the pitch of the spiral?

In theory, yes—the thinner the conductor, the less chance for eddy currents, but that would also decrease its capacitance.

Therefore, it makes sense to simply reduce the spacing between adjacent conductors (in the case of a spiral, this would be the pitch), while maintaining a high ratio of conductor width to gap width.

Alternative solutions include all sorts of ‘snaking’ patterns, combs, concentric open-ended rings (connected to each other), and so on. Whatever you can imagine. It’s purely a matter of aesthetics.

Let's Design and Build a (mostly) Digital Theremin! (by dewster)

"To avoid a shorted turn, the shield needs a break—or better yet, a structure that prevents eddy currents from forming. Fractal patterns look pretty good for this. You could start with a ‘comb’ design, similar to the one used on the EW Pro board." - ILYA

Could you provide examples of such patterns? I've tried plates made of spiral turns, plates with a big hole and slot, nothing seems to significantly reduce the "shorted turn" effect, and all reduce the capacitive interaction to some degree. Since eddy currents are circular, I imagine a conductor of any significant area would support them? So I'm guessing the conductors on the plate antenna would need to be really thin, like less than 1mm wide? Probably easier just to sufficiently distance the solid plate.

"We also seem to have touched on the chain of capacitors located inside the coil—a short segment of thin wire, rising coaxially from the ‘cold’ end toward the ‘hot’ end, could act as the capacitor."

True, but it can be a very controlled amount of distributed capacitance yielding a distributed voltage and distributed 90 degree phase by using the real deal components in series.

Let's Design and Build a (mostly) Digital Theremin! (by ILYA)

Just a case of déjà vu, though.

Let's Design and Build a (mostly) Digital Theremin! (by ILYA)

4. Positioning the "hot" end of the coil closer than approx. 2/3 the coil diameter to the plate antenna hurts Q by acting like a shorted turn.

I feel like we’ve discussed this before.

To avoid a shorted turn, the shield needs a break—or better yet, a structure that prevents eddy currents from forming. Fractal patterns look pretty good for this. You could start with a ‘comb’ design, similar to the one used on the EW Pro board.

5. The capacitive divider should probably consist of a string of capacitors down the center of the coil where the electric field is essentially zero.

We also seem to have touched on the chain of capacitors located inside the coil—a short segment of thin wire, rising coaxially from the ‘cold’ end toward the ‘hot’ end, could act as the capacitor.

Let's Design and Build a (mostly) Digital Theremin! (by dewster)

New Coils & Mounts

I wanted to get at least the pitch side out of the AM broadcast band, which would require a much larger coil than the current 1mH. The bottom of the band is 530kHz, and an 8mH inductor would resonate around 400kHz. Past experiments revealed the following:

1. Wire smaller diameter than 32 AWG hurts Q.
2. Wire smaller diameter than 32 AWG is significantly more difficult to wind.
3. The aspect ratio of the coil should be somewhere between 1 and 2.
4. Positioning the "hot" end of the coil closer than approx. 2/3 the coil diameter to the plate antenna hurts Q by acting like a shorted turn.
5. The capacitive divider should probably consist of a string of capacitors down the center of the coil where the electric field is essentially zero.
6. The coil in general should be reasonably kept away from anything metal, or even wood as it tends to be slightly conductive.
7. 3D printed coil forms can help guide the winding process by making the Z-step slightly larger than the wire OD.
8. Concentric items printed with the same Z-step can produce significant locking friction when fitted together.
9. The pitch coil should probably be the larger of the two coils.

Toward this end I've printed a bunch of candidate configurations, and have finally arrived at something that I'm fairly happy with. Both pitch and volume coil forms will be 70mm in diameter. Coil mounts on both ends will provide ~2/3 * 70mm = 45mm clearance on both ends. The mounts will have mounting holes for the AFE on one end and for possible attachment to a plate antenna on the other. The AFE will mount on the "cold" end of the coil mount. The width of the mounts should allow them to be slipped into the controls & tuner hole.

Above: the full thing rendered in OpenSCAD. This is the volume coil, the pitch coil will be somewhat taller.

Above: Old 1mH pitch coil on the left, new 4mH volume coil on the right.

Above: the old pitch box (left) and the new volume coil assembly (right).

Above: a view of the tube which will hold the capacitive divider string (left), and the AFE cover (right).

I'm printing the mount ends with 0.333mm Z-step. The 25mm overlap with the coil forms is printed with this same step in order to encourage interlocking. The winding section of the coil form is printed with a Z-step slightly larger than the wire diameter, and this is a simple thing to do in the slicer software. The coils and mounts will likely be secured via a couple of nylon M3 screws drilled and tapped through the OD of the overlap region.

TODO: Build AFE and make sure that end mounting on the coil form is OK (the magnetic field might interfere). Also check new AFE filtering (still working on that too).

Lydia Kavina: recommendations for theremin composing (by ILYA)

"often just start my compositions on a piano"

I usually “think” on the guitar, and then I can (with difficulty!) transfer something to a MIDI editor.

"I want to hear it!"

OK
If I gather the courage, I’ll perform a solo part on the theremin and send a video to Peter Thermen for the next annual “Tereminology” festival.

Trouble Hearing Theremin on Zoom (by Trymeinchesss )

This is super random but the Zoom live track is a great interface. I don’t own one but had the chance to use one over the weekend, and it was great.

Lydia Kavina: recommendations for theremin composing (by Trymeinchesss )

I am more of a producer than a composer, (I love orchestras but I wouldn’t want to compose for them because I don’t have access to a full orchestra to hear my creation) but anyway, like ILYA said I often just start my compositions on a piano, and then later move them to the instrument and sound I want. While I do agree that the article is somewhat uninformative, there are certainly limitations on what can be easily played on the theremin but those shouldn’t hinder your process. Worst case scenario, hand that part over to a cello or another synth

Also, when you decide to show people your stuff ILYA, I want to hear it!

Lydia Kavina: recommendations for theremin composing (by ILYA)

My humble view, (and I’ll admit upfront, I’m not a composer myself, though I’ve got my own “Four Seasons” lying around, which I haven’t shown to anyone yet), so, in my humble opinion, you should compose without tying yourself to specific instruments. Whatever wants to come out, just lay it down well. Everything else is the performers’ problem. The only exceptions are those particularly specific cases where you really want to hear a part played specifically on the theremin.

Lydia Kavina: recommendations for theremin composing (by guillaume)

It’s about nothing. ‘Water is wet, and the sound of a violin is triangular.’It would be more interesting to hear, for example, about such subtleties: which passage is easier to perform — a linear one, like C-D-E-F-G, or a leaping one, like C-F-D-G-E?

Wouldn't that equate to limiting your composition to what some of the players can't achieve, rather than what you think should be played?

The advice given of the singing test (or indeed whistling!) in my humble view is the best advice for what's 'easier'. It'll give a solid sense of what speed is too fast, and what leaps are challenging (specially in the context of the accompanying harmonies you're developing)

IEMs (by trymeinchesss)

I have been using the Zero:2 In ear monitors, (IEMs) and they have changed music for me forever I highly recommend than any music enthusiast pick themselves up some IEMs, because the sound quality is better than any headphones or earbuds remotely in that price range. The music is so much more clear and focused. Small details you have not noticed in the track like buried layers or tiny panning changes become apparent. I figured I should share my experience with those also interested in sound. While many can not notice the difference between and MP3 and a FLAC, an IEM will have a noticeable impact.