Build Project: Dewster's D-Lev Digital Theremin

Posted: 11/23/2018 10:27:40 PM

From: Minnesota USA

Joined: 11/27/2015

I used a piece of what I thought was junk mahogany for the wood cabinet ends (it came from Menards, a place not known for fine wood), but once I put some stain on I regretted wasting it on this.  It was gorgeous.

For the mic stand mounts I tracked down the adapters that Dominik uses on his Subscopes.  These are a lot cheaper that the 3-screw type, and they only require a single tee-nut to be installed from the inside center of the enclosure base.  Another side benefit is that they have a coarser thread on the top end, so when you need to remove your theremin from the stand you don't have to rotate the whole thing so many times.

The PVC extension arms in the picture below are free to slide in and out of the enclosure up to 11" without interference.  The left and right tubes are aligned, so if one is fully in, the other must be extended.  Since I play seated, I'm constantly fidgeting with the antenna placements.

My plan is to have a detachable tuner in some sort of compatible housing that sits on top of this enclosure.

Posted: 11/23/2018 11:52:37 PM

From: Minnesota USA

Joined: 11/27/2015

I had to think long and hard about my antenna/inductor/driver placement and how signals were routed near the antennas.  From the moment I finished winding the two inductors I was committed, if only on this prototype, to make them part of the extension arms rather than mount them behind the plates.  I did this for several reasons:

1) they were pretty and I wanted them to show
2) I didn't want bulky antennas with stuff behind them
3) and most importantly, mounting them horizontally suited the concept of having a plane of separation between high and low impedances feeding the antennas.

As you look at voltages and currents through an inductor in series with a capacitive plate, in a resonant condition, you find low voltage, high current at the drive end of the inductor (Lo-Z) and high voltage and zero current at the plate (Hi-Z).  The low impedance drive end is quite immune to anything capacitive nearby like your hand, while the high impedance end is very susceptible to capacitive effects, which is exactly what you want. 

As an example, if you are a "wide stance" player (Larry Craig?) and you want your Etherwave antenna further away from the volume loop, try this: move it out, but take the 3 series inductors mounted on the board with it. The longer extension wire that you run from the main board to the inductors will be low impedance and barely affected by any proximity, but the antenna will behave normally.

Anything conductive (or insulating dielectric for that matter) that enters the high impedance end, which is almost entirely an electric (E) field, will influence it.  Cables, metal panels, equipment stands, and the steel plate in your head are all capable of altering the resonant frequency of your LC circuit.  You can partially compensate for these things by adjusting your pitch control (if you have one), but the fact remains that these stationary objects all represent fixed capacitances that are paralleling, and ultimately reducing the influence of the variable capacitance created by your carefully trained wiggly fingers.  It seems prudent to keep this parasitic junk out of the pitch field, because even though you can compensate to maintain your resonant frequency, you have desensitized the pitch field to any small delta C.

So, it seems logical to try to keep everything that is high impedance (in our case the plates, rod, or loops, and the "hot" end of the series inductors) separated from cables and anything metallic that is low impedance or has body capacitance.

This leads up to the way I decided to arrange the antenna plates, inductors, and the antenna driver board.  The driver boards are located in the extension arms nearest the enclosure, and just outside the influence of the inductor fields.  Each driver board connects to the what is the low-Z end of the inductor, and the opposite end of the inductor, the hi-Z end, connects to the antenna plate through ball swivels. So far so good.

There is a little complication with my setup on the D-Lev because there is a sense line connected  directly to the antenna, which will be Hi-Z, that runs to a 1pF/100pF capacitive divider that reduces the high antenna plate voltage down to a logic-compatible level.  Dewster hangs his 1pF cap right on the antenna and runs it back to the antenna drive board which contains the 100pF divider cap.  All of his paths are short, and the sense signal can be separated from other elements such as the inductor.

I wanted to run this sense line back through the center of the inductor, but that would be passing a Hi-Z wire, with high E fields and high sensitivity to external influence, back to the driver board, with the divider located entirely on the driver board.  This is breaching the idealized Hi-Z to Lo-Z barrier a bit, and it drags the antenna pitch field a little more down the horizontal tube than I would like.  I do have options however.  I made extension boards that would allow me to insert the 1/100 divider some distance inside the inductor, and since the divider represents an impedance step, it becomes a trade-off between trying to keep the Hi-Z length of sense wire short but at the same trying to avoid having to insert too much of the Lo-Z side of the divider into the inductor where it starts to capacitively couple to the "hot" end.

It's a prototype.  So far it works - the drivers work, I have zero and quad waveforms, and I can change my routing if need be.  Isolation between the inductor winding and the centrally routed sense line is greater than 30dB and about the same if I route the wire outside, so I'm not going to worry about it.  It's all going to change anyway.

So here are some pictures of how one of the arms goes together for now, starting with the ball swivel that I made for the volume end:

The end nylon thumb screw can be loosened to allow the socket for the steel ball to rotate 360 degrees on the axis of the extension arm.  With this added axis of motion the locus of possible positions for the metal shaft connected to the ball will comprise more than a full hemisphere. 

There is a spring inside the PVC end piece that makes contact with the ball swivel in any position:

The brass wire in the center of the inductor tube is the sense wire.  It is quite long now to keep the driver board away from the inductor, but I expect it may be shortened later.

This is the similar ball joint and inductor arrangement on the pitch end:

Posted: 11/24/2018 12:31:22 AM

From: Northern NJ, USA

Joined: 2/17/2012

Roger, you've done an amazing job, and so quickly!  Kudos!  It's nice to see the coils, they make it appear as though there is something seriously science-y going on (and there is!).

Do you have some means to keep the coil sense wire more or less coaxial / equidistant from the inside walls of the coil?

Posted: 11/24/2018 1:29:58 AM

From: Minnesota USA

Joined: 11/27/2015

The board connector is positioned to hold it centered, and it is fixed on the swivel end, but I still have the wire centered in a block of foam anyway.

Posted: 11/24/2018 1:44:23 AM

From: Minnesota USA

Joined: 11/27/2015

And here is what the D-Lev looks like at this point with some plates mounted.  The swivels allow for a wide range of positions including quite a bit forward of the extension arms toward the player.

This is all I have for the moment.  I need to work on a main board layout and pcb, but I'm waiting on some 2x22 sockets before I can build it up. This board will support the LCD display and should consist largely of routing to pin headers, but the density might be a problem for my home brew boards.  The toslink and the EEprom should also be on this board.  I guess I could start on the tuner too.

Posted: 11/24/2018 1:50:03 AM

From: Minnesota

Joined: 3/9/2013

This is amazing to see come to life.  Very anxious to hear the result.  Thanks for doing this.

Posted: 11/24/2018 9:25:07 AM

From: Theremin Motherland

Joined: 11/13/2005

what an amazing implementation of dewstermin there is!

Posted: 11/24/2018 10:24:28 AM

From: germany, kiel

Joined: 5/10/2007

Great! Especially the ball ends are unique details. May i ask where you did get the ballheads? Or - if self build - how they are fabricated? 

Posted: 11/24/2018 10:43:21 AM

From: Minnesota USA

Joined: 11/27/2015

Hi Dominik,

The balls themselves are 1"  soft steel ball knobs with a single 1/4-28 threaded hole.  They are intended to be used on the ends of machine levers and the like.  They are polished, but they are machined instead of having been ground to a fine finish like a ball bearing, so they are not perfectly spherical and tend to have slightly tighter spots as you move them around.  I have successfully flame-annealed hardened steel bearings and drilled them out before, but it takes carbide bits to do it.  And even when I was able to drill a hole, tapping was impossible.  In that case I just ended up brazing a shaft into the hole.

The sockets were made of PVC pipe that was bored out internally to provide a tight fit.  There is a slightly smaller ID in the bore at the very end of the socket so that the ball pops into place and is retained.  The pitch antenna plate turned out to be too heavy for the friction provided by the ball socket, and if you study the photo of that end you will see a metal ring that I had to press on over the end of the socket to tighten it up.  I should have left more PVC wall thickness surrounding the ball.

Posted: 11/24/2018 4:55:39 PM

From: germany, kiel

Joined: 5/10/2007

Thanks. I just knew plastic ball knobs before. Yeah, i got the impression that yours looked like hardened ones and maybe automotive accessories. Aha! Good idea. Found brass ball nuts as well now. The metal ring on the pitch side is looking nice either. 

You must be logged in to post a reply. Please log in or register for a new account.