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

Thank you for that info / clarification André!  I had a feeling the Claravox was grounded via the power supply, but nothing in my notes / pictures from when I worked on one indicated this.

In this thread [LINK] and below I show a picture of the new EW power supply, which has a grounded 3 prong AC plug.  Unfortunately I didn't test it for ground continuity through to the barrel connector.  I wonder if this is the same supply used for the Claravox?  The Etherwave played fine without any additional grounding, which leads me to believe the AC ground goes through.  That, and it makes sense.

Yes, this is the exact same supply I have with my Claravox.

"Yes, this the exact same supply I have with my Claravox."  - André

Thank you André!  I would have bet that was the case, use one custom (ground through) supply for as many products as possible.

The Litz Advantage

Quite interesting graph from "Litz Cable for HF Single Layer Coils" by Alan Payne, 2021:

The above compares litz vs. solid wire, both of the same outer diameter, close wound as a single layer solenoid.  For compliance and other reasons, the Theremin sweet spot is the range below the AM band (<530kHz) down to maybe 200kHz, and this just happens to coincide with the peak benefit that litz wire can impart to the reduction in AC resistance.  I assume Q over this range would then be doubled, going from ~200 to ~400.  The trick would be to find litz wire that has an outer diameter of 0.15 mm or less, as too much thicker would make for rather large, unwieldy coils.

Roger's / Rob's P3, SN #5

Thought folks might enjoy a few snaps of Roger's latest creation for Rob Schwimmer.  I took these while it was in the shop for a minor repair job:

Above: Lookin' good!

Above: Audience view.  I like the logo size and position, rather like a stage piano / synth.

Above: Player view.

Above: Guts view.  The removed top shell is supported by two padded swing-out legs - Roger thinks of everything!

Above: The power supply box that sits on the floor.  Love the logo!  Aviation connector on left, power LED in center, power rocker switch on right.  I think aviation connectors are really the way to go here, barrel connectors aren't secure and they scarily can't guarantee voltage, polarity, or even AC/DC.

Above: Power supply box showing fused C14 type AC entrance.  Really pro labeling job!

Above: Power supply box guts.  Not just beautiful on the outside!  Note the ferrite bead on the DC leads, and the braided nylon tube on the AC leads.

Above: The Mean Well switching power supply inside the power supply box - a brand you can trust (or so I have read).  It has a trimmer pot on the left (not shown) so I assume you can over-volt it a bit if necessary.  The D-Lev typically draws around 0.5A @ 5V so this is way overkill, but not necessarily a bad thing.

I think it's best to place the switcher far away from the D-Lev guts, as this solution does.  The opportunity to route AC ground through it is there too.  One might also make this box do double duty by mounting Mute and ACAL momentary footswitch buttons on it.

Underhanded

For some reason in my head lately I'm comparing how the player addresses the pitch rod as how one pitches a baseball or wields a tennis racket.  To wit:

1. Many take an overhand approach: hinging of the elbow sets the hand distance to the antenna.  The big downside here seems to be the need to hinge the wrist quite a bit as well in order to keep the hand orientation and its gestures roughly constant while addressing the rod, and in the far field where the hand is back at the body this looks like a pretty tense pose.

2. Some take a backhand approach: the back of the hand addresses the antenna, and the rotation of the shoulder / twist of the upper arm sets the distance to the antenna.  I would think that this somewhat precludes complex fingering techniques?

3. With the D-Lev pitch plate I take an underhand approach: my forearm remains parallel to the floor, and minor hinging of the shoulder and elbow set my hand distance to the plate.  I don't use a fingering technique per se, but I do use the hinging of all of my fingers together to do small pitch steps, which are a constant type thing regardless of the pitch being played because my hand naturally remains perpendicular to the plate.  My vibrato is via twisting my forearm, and there is barely any wrist action at all for anything.  Early on I used the "OK" type index finger and thumb together deal, but even that felt too tense, so my hand is just open and 100% relaxed now.

I'm not saying that the #3 underhanded approach is for everyone, but it works well for me (though I don't perform or play anything fast or intricate).  And for those who have injuries / limited mobility (not the reason I'm doing it BTW) it might be worth giving it a shot.

I believe the more that one can relax while playing, the less chance there is for playing related injury.  I know I've said this a zillion times, but analog pitch fields are so tight that there is an inherent tenseness to playing them.  And analog volume fields tend to be so loose that all the hand and arm movement necessary to get them to articulate properly seems like it could also lead to injury.

Chebyshev

Super useful & insightful article on Chebyshev approximation of functions: https://www.embeddedrelated.com/showarticle/152.php

I really could have used this when I was instead spending weeks / months searching for these coefficients via brute force in a spreadsheet.  No mention of coefficient resolution / rounding / truncation in the article though, which I might have expected given one of the examples was on an 8 bit machine.  Still, the general observations are spot-on, particularly this:

  "The only embedded systems that require precision floating-point computations are desktop calculators."

In the D-Lev I ended up using the lowest precision calculations I could get away with in order to minimize real-time and code space use.

"The only embedded systems that require precision floating-point computations are desktop calculators."

Well... I should probably apply it to my calculation of "log(x)/log(2 to the power of (1/12))" in Open Theremin with MIDI.
Sometimes it keeps me awake...

So if you calculated log(x)/log(2^(1/12)) on an 8-Bit integer Atmega328 than one can only conclude you don't know how to calculate and save calculation time ....

Chris

Thanks to help to keep me awake...

Well... this floating point calculation happens to work enough for the needs but I know for long that I should have a look at it. 
And, most important, see what is the impact on playing, if any...