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

Posted: 11/15/2020 9:36:03 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

Negative Capacitance - Update

I have more insight into what was going on with my previous post on this issue:

The oscillator uses a 4mH coil, and the unpowered D-Lev plate also uses a 4mH inductor!  As I move my hand nearer to the plate I'm tuning the D-Lev plate to the same frequency as the oscillator, hence the hump.  Mystery solved!  Well, at least it seems less weird.  I wonder if shielding could be accomplished this way?

Posted: 11/19/2020 3:21:22 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

I see that the Open.Theremin had MIDI all the way back in 2017:

https://github.com/MrDham/OpenTheremin_V3_with_MIDI
http://www.gaudi.ch/OpenTheremin/index.php/opentheremin-v3/midi-implementation

So I'll be pouring over that code soon.

I see that polycarbonate has a dielectric constant of 1.9, but the same table tells me PVC is 1.7.  So I'm shopping for thin wall PVC tubing with a 2" or so OD and a thinner wall, which should reduce self-C and save a bit on weight.

[EDIT] Oops?  See below...

Posted: 11/19/2020 5:29:37 PM
Buggins

From: Theremin Motherland

Joined: 3/16/2017

Doesn't pvc have 4.0? 


Code:
Material			Dielectric Constant (ε)
Vacuum				1.000
Dry Air				1.0059
Foam Polyethylene	1.6
Fluoropolymers		2.0
Polypropylene		2.1
Butyl Rubber		2.3
SBR					2.9
Silicone Rubber		3.2
Plexiglass			3.4
PVC					4.0
Glass				3.8-14.5
Distilled Water		~80


Posted: 11/19/2020 6:07:14 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"Doesn't pvc have 4.0?"   - Buggins

I thought it was 3, but this table says says 3.4:

https://www.kabusa.com/Dilectric-Constants.pdf

It also says polycarbonate is 2.9 to 3.  Amazon has some thin wall 2" polycarbonate tubing that would cost ~$2 per coil:

https://www.amazon.com/gp/product/B000OMHIHM/

I wonder how difficult that is to saw?  I'll order some and see.  Clear coils would look kinda neat.

[EDIT] Here's the table I was looking at initially:

http://www.deltacnt.com/wp-content/themes/va/pdf/99-00032.pdf

My mistake was looking at powdered solids and such.

Wall thickness may be more of a factor than the type of plastic.

Posted: 11/19/2020 6:35:14 PM
Buggins

From: Theremin Motherland

Joined: 3/16/2017

I'm using polypropylene water pipe tube 32mm diameter, but it's thick (5-6mm wall). E const for ppl is 2.1
It will not only serve as frame for coils, but as well will hold antenna.
What C will add 5mm PPL frame?

Posted: 11/19/2020 11:42:45 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"What C will add 5mm PPL frame?" - Buggins

Who knows, but my spreadsheet, based on Knight's papers (which in turn are admittedly based on significant speculation regarding self-C), shows very little self-C change with mild variation in dielectric constant (PVC, Teflon, etc.), as well as very little self-C change with wall thickness.  I think this is probably deep in diminishing returns territory.

I converted the dielectric constant PDF into an *.ods (LibreOffice) spreadsheet and sorted the solids [LINK].  Maybe we should be making the coil forms out of foam rubber (1.1)?  Frozen Deuterium (1.3)?  Do they make Formica tubes (1.3)?  Seriously, polypropylene (1.5) looks like one of the best choices as long as it's mechanically OK for your application.  But once you pick some sort of plastic tube you're probably golden.

Posted: 11/22/2020 7:38:48 PM
Buggins

From: Theremin Motherland

Joined: 3/16/2017

Thank you for useful table.

Posted: 11/25/2020 2:34:11 AM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

P3 Progress

I haven't been playing P2, nor working on P3 nearly as much as I would like.  The ergonomics of P2 aren't the best which is a disincentive, and I can't seem to bring myself to cut the microphone stand down to make it more playable (burned / lessons from the past - don't make a destructive move unless you can see a clear path all the way to the end - which can be paralyzing, but never again).  This morning I awoke and vowed to stick all subsystems in individual cardboard boxes and just move ahead already, but got derailed stuffing the Lincoln Memorial antenna plate boxes, which actually went rather well:

The thicker (36 gauge: 0.005"?  0.127mm?) roll of aluminum I bought is both thin enough to be adequately deformable yet thick enough to hold its shape well.  I used blue painter's tape to hold it down, but then ran across some old double sided foam tape in my junk box which is now securing the top ends real good, and is probably all that's necessary.  Got a little carried away improving the coil securing hardware, and still need to address the insufficiently tall volume coil with an extension - the dimensions require a 90mm or so former length of 1 1/2" schedule 40 PVC - the plate side windings need to be a sufficient distance away so as to not be damped by the plate looking like a big shorted turn (thanks ILYA!).  No AFE's in there yet, I'll probably use my old hand wired ones from P1.

With P2 in particular, but even with P1, I didn't experiment around with the macro ergonomics enough.  This time I'm going to keep everything as utterly flexible as possible, and only commit to a case design and dimensions after I've played the best configuration I can come up with for an extended period of time.  P2 was a step backwards, and I don't want to make that mistake again.  You (or at least I) really can't think these things out in your head, you can give it your best guess, and you can come up with new configurations when the old ones aren't panning out, but you can't know where anything is going or where everything will end up.  For better and worse, it's wide open experimentation with electronic musical instrument controllers, which live / die on their ergonomics (if they aren't stillborn due to poor electrical / software engineering, or poor physical design - a thousand ways to die).  But I still have to ask myself things like: are boxes with U-shaped antenna plates sufficiently optimized (micro ergonomics)?  It probably never really ends, but it hopefully reaches a point where you've sufficiently explored and ruled out all of the obvious alternatives.

Posted: 11/26/2020 1:39:12 PM
pitts8rh

From: Minnesota USA

Joined: 11/27/2015

Eric,

What were the problems with the ergonomics of P2 again?

Posted: 11/26/2020 3:28:47 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"What were the problems with the ergonomics of P2 again?"  - pitts8rh


Above pic of P2 for reference

P2 Problems:
1. The whole thing is a bit too high, which I could fix by sawing down the mic stand center post.  So not a P2 issue per se.  Maybe I should try wearing platform shoes?  Due to the other issues below I've decided against the sawing, at least for now.
2. The exposed plates are really fragile and I'm constantly accidentally rotating them by brushing against them with my clothing.  Again, not exactly a P2 issue as the plates and their supports could be beefed up.  And I could certainly move P2 out of the high traffic pinch point between my PC and my workbench.  But the experience has soured me entirely to exposed antennas of any sort which just hang out in space.
3. Now and then I get a little off-center of the pitch plate.  Perhaps my peripheral vision is going to hell, but it leads me to believe that the best place for the tuner is directly above the pitch plate, something I'll be testing in P3.
4. The ACAL knob pushbutton is just too close to both axes, making ACAL a much trickier endeavor than it was for P1.
5. I don't think the coils should be anywhere near the controls, mostly for reason #4.  Coils belong behind the plates, which better concentrates the fields.
6. The plates shouldn't be positioned on parallel horizontal axes.

When I watch Theremin videos, it seems many address the pitch antenna somewhat frontally, with their volume hand more off to the side, and this is the way I believe I want to play too.  With my right hand a little off-center to the right, and my left hand more off-center to the left.  Kinda like P2 with the tuner moved above the pitch plate, and with the volume plate moved further to the left and lower, the coils behind the plates.

[EDIT] The P3 guts & all printed parts are finally a go (now we just have to get the developer to be more of a go):

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