Build Project: Prototype III - D-Lev Digital Theremin

Posted: 1/18/2021 6:06:54 PM
pitts8rh

From: Minnesota USA

Joined: 11/27/2015

In an ongoing quest to pervert Dewster's D-Lev design into something that he no longer recognizes, I am now working on the fourth iteration of a prototype (hence the PIII title...wait, what?) that further deviates from his vision.  He is going to rue the day he ever let me in on this.  Like the movie director that completely loses control of his or her visionary film to an uncooperative editor, I bring you D-Lev PIII: Return of Conventional ErgonomicsDirected designed by Alan Smithee.

Inside jokes aside, what started out as a quick breadboard build of another D-Lev theremin that would allow easy access for testing has blossomed into a new enclosure design. I wanted something that had all of the boards spread out and accessible so that several different types of inductors could be easily swapped and tested. At the same time this would give a good opportunity to try having the tuner built right in the front panel of the enclosure (my EW "Pro" style of enclosure for my first real D-Lev has the tuner on a separate stand). As it turns out this seems to work pretty well.

D-Lev Prototype II goes straight to DVD

What happened to D-Lev II (a.k.a.RPII)?  Well after spending probably a couple hundred hours designing and printing parts to build a few, I've back-burnered it in favor of a simpler cabinet that is practical and potentially easier for someone else to build.  I was very happy with the some of the quick assembly features in the RPII, but it still had those curvy parts that add complexity and scare off DIY builders.

The pitch arm is active and has a spring-loaded connector that allows it to plug into the housing and lock with a quarter-turn twist.  The angles of the pitch and volume antenna are friction-adjustable to allow them to be set correctly for different angles of the body (see the tilt adjustment below).  On this prototype the sides and back are walnut and leather, and the front is maple machined with a slight compound curve.  The bezel on the back just below the rear attach point provides a window for all of the connectors, and in fact the front panel would provide the support for all internal electronics.  The front panel could be separated from the housing by removing the two screws on either side of the bezel, and the front panel and all electronics could detach by removing a single narrow ribbon cable.

I like a lot about this design including the closer volume antenna when the body is tilted at an angle.  But I'm actually getting a little tired of the vertical cabinet style of my "Pro" D-Lev (and the Claravox too, and it hasn't even shipped yet).  I may eventually finish this one because it is so far along but probably not the five that I printed all of the parts for.

Enter the RPIII D-Lev:

This has a sloped panel to accommodate the LCD and tuner displays and to get away from having to crouch down to reach the front panel controls. And by reverting back to an Etherwave-type of horizontal cabinet it also dispenses with the pitch arm needed to gain antenna spacing when using a vertical cabinet.  I think it goes back to my early lust for an Etherwave Pro, but tastes and lusts change.

Something like this (keeping in mind this is a breadboard):

Actually the horizontal form-factor enclosure kept popping up in my thoughts, but the thing that seemed to keep killing it was the amount of finger space for the two columns of four encoders that serve as soft key controls on the display.  Several knob layouts were tested to stagger or slope the alignment of the encoders, but nothing felt right compared to two columns on either side of or below the display.  The height requirement for this encoder array had prevented me from delving into a horizontal enclosure until I gave up completely on the pitch-arm concept and decided that I had to make it work.

Finger spacing for gripping the encoder knobs plus the diameter of the knobs themselves was taking too much vertical height, so I decided to cut both down.  On the bench I usually just use the knurled shafts of the encoders themselves without knobs; it gives a good grip even at 6mm diameter plus the small diameter gives a good feel for the encoder detents.  After trying some small diameter 3D-printed knobs I think the solution is going to be either narrow printed knobs or molded rubber sleeves.

To maximize the space available for the knobs without placing them on either the front or top panels (neither of which work for both seated and standing playing positions) the sloped panel was again chosen (it was on the very first P0 prototype), and this has the added benefit of deviating a little from the simple boring rectangular box appearance of the Etherwave.

Another problematic issue for easy construction (I'm thinking of the DIYer again) is what to choose for a front panel material.  Encoders, switches, displays, and pots like panels around 3mm thick or less.  While thin baltic birch plywood works if you cover it, thicker wood requires routed cavities on the back side around components and modules. 3D printed panels take time and have appearance issues, and metal requires machining and still needs a coating or an overlay.

So I'm back on board with using bent 1/8" cell-cast acrylic for the entire front-top-back panel attached to a wood base with finished hardwood ends. It's hard to improve on this for a balanced combination of professional appearance and ease of construction. Anyway, that's my current storyline until I change it.  I think my first P0 prototype was on the right track, once I make the cabinet smaller and get rid of the pitch arm:

A little better acrylic bender

I think that flat or bent acrylic has so much potential for theremin panel or enclosure construction that it deserves to be described in some detail.  The acrylic could also be used simply as a non-metallic but bendable base material, with the surface finish being textured vinyl, leather, wood veneer, wallpaper, or yes, as a last resort, even speaker grille fabric!

To bend 22" long acrylic panels a larger and higher wattage hot-wire unit would be needed.  The panel in the photo above was bent over my knee after heating with a piece of nichrome wire stretched on a board, but going forward something more accurate and repeatable would help greatly. I looked at various commercial hot-wire units but they didn't really didn't have the bend-radius versatility that I was looking for.  You can get a water-cooled 24" quartz strip heater for $100, but you can't control the exit aperture to narrow or widen the heat zone.  Plus they use a quartz element that I expect is either unavailable as a spare or it exceeds the cost of the whole heater. And you have to haul out a bucket of water when you want to use it.

Instead I built a new adjustable 0-24VAC 40-amp power supply and a 24" aluminum and wood, air-cooled strip heater that uses a twisted pair of 20ga nichrome wires for the element.  I should point out that these actually aren't benders.  They are only the heaters, but once you soften the material along a line you can easily make bends over anything that has two planes at the correct angle.  In fact most of the hinged benders that I have seen are pretty sloppy and in fact obscure some of the cues that you need to obtain the precision to make the acrylic parts that actually fit the wood parts.

The twisted nichrome wire is spring loaded to keep it stretched when expanded.  At temperature the wire is about 10mm longer than at room temperature:

BTW, since this is air cooled it can only be used intermittently for safety reasons.  There are air gaps between all wood and hot aluminum, and the wood does not get hot.  Nevertheless, you don't leave something like this turned on when not in use (I have a darkroom timer to turn it off). The wood was needed to provide a non-marring surface for the acrylic, which must be bent with the protective paper removed.

The angles are polished on the sides facing the wire to lower the aluminum's emissivity, but the heat that the angles do absorb is at least partially dissipated by air convection around the rest of the material.  The height of the wire and the spacing of the two angles is adjustable to allow some control over the beam of IR that is projected onto the acrylic.  The top edges of the wood platens are in the shadow of this beam.  Thus far this is providing very uniform bends of either very sharp or large gentle radii.  After some trial and error it seems that the formula for success is to keep the wire quite far from the plastic and allow more time for the plastic to heat somewhat uniformly throughout the thickness.   Four minutes at 400 watts seems to be a safe combination so far.

So what's the feeling on acrylic/wood for theremin enclosures?  How about ergonomics in general?

And if there are any left-handed players watching, what specifically would you want to be reversed for a left-handed theremin?  Antennas would be of course, and to keep the LCD in front of the player that would be swapped with the tuner.  But how about the elements within the tuner itself. Would it be okay the way it is or should the circular scale and the volume LEDs be reversed as well?

Posted: 1/20/2021 8:53:11 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"But I'm actually getting a little tired of the vertical cabinet style of my "Pro" D-Lev (and the Claravox too, and it hasn't even shipped yet)."  - pitts8rh

Ha!  I couldn't warm up to the slanted vertical orientation much as I thought I would either.  One problem for my P2 is the volume coil is right at gut level, which makes body proximity rather touchy for volume axis auto-calibration and play.  I'm experiencing this with P3 too but it isn't quite as bad, and there are more options to explore in terms of plate placement.

I haven't gone through it with a fine-toothed comb, but it probably wouldn't be too difficult to make a lefty option for the tuner purely in SW, using the existing tuner PWB.

Roger, now that you've got some Melodia playing under your belt, what's your feeling re. a left plate rather than loop?  What about maybe making that plate horizontal?  I think the loop is mainly so the hand can get very near the loop without actually touching it, a feature that isn't required for digital Theremin volume side operation.

Posted: 1/20/2021 11:40:00 PM
pitts8rh

From: Minnesota USA

Joined: 11/27/2015

I guess except for the tuner's 7-segment display dot everything is symmetric, and the dot will be hidden anyway.  That sounds like a good thing if you can do it.  I've also wondered if it would be possible to have PCAL and VCAL swap positions when the system PV reverse is enabled. Seems like a logical thing, if possible.

Regarding the Melodia plate, I think it's interesting in that it can seem almost interchangeable with the loop, although that could be dependent on the player's volume technique.  I like the integrated look of the end plate but I could foresee that it might cause trouble if you are an up/down hand flapper.  Something that I have noticed is that if you dip too deeply into your volume cutoffs you can pop out the other side and get some double notes - not a good thing. This is why I have started playing above the end of the case.  It gives me the sharper cutoff that I prefer and if I start to drift while playing I bump into the case.

As far as having a flat plate for the volume side, I've never had any problem with that as long as the plate has a flange bent over on the player's side to allow for edge playing.  I keep various shapes of metal and plastic clamps in my music room and I try these things from time to time.  I don't see any difference between the volume plate and a loop on the D-Lev as long as you make a few tweaks.  The legacy advantage of the loop for dipping into cutoff without hitting metal isn't really an issue with the D-Lev because you can put the volume knee just about anywhere you find it comfortable. 

But I'm always thinking that the best approach for either of these antennas is a socket that can take any antenna that the player prefers.  I intend to make an interchangeable end plate for the RPIII just to try, but I've recently developed a left hand problem and I'm not sure if it has been aggravated by the Melodia. For now I'm changing hand positions quite a bit to try to minimize stress on it, and I may have to go back to a loop theremin for a break.

Posted: 1/21/2021 4:09:05 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"I've also wondered if it would be possible to have PCAL and VCAL swap positions when the system PV reverse is enabled. Seems like a logical thing, if possible."  - pitts8rh

There are quite a few vmod & pmod knobs that should probably swap as well to fully accommodate lefties.  Might be best to fork the SW load in order to do this.

"But I'm always thinking that the best approach for either of these antennas is a socket that can take any antenna that the player prefers."

I agree, though with (my usual, monotonous) reservations.  External antennas are so fragile and fiddly, and invite ESD.  I was constantly brushing against the P2 plates, almost knocking it over like every time. 

And, beyond some basics, I tend to view optional physical adjustments as a net negative, as any little change can easily throw off muscle memory.  Sorta like having the neck of one's guitar or violin, or the keys of one's clarinet, subtly change between playings.  Breakdown and setup of one's Theremin shouldn't change basic ergonomic dimensions even slightly (unless that is the desired goal, which it won't be after one plays it enough).  Maybe I'm a princess with a pea, but just changing the D-Lev stand height 1" throws me off for a while.  I think I'd rather adapt to fixed, slightly sub-optimal ergonomics, than have everything shift like desert sands between gigs.

"...I've recently developed a left hand problem and I'm not sure if it has been aggravated by the Melodia. For now I'm changing hand positions quite a bit to try to minimize stress on it, and I may have to go back to a loop theremin for a break."

Oh, sorry to hear that.  The repetitive small movements of the volume hand could easily lead to injury.

Posted: 1/21/2021 6:49:05 PM
pitts8rh

From: Minnesota USA

Joined: 11/27/2015

"I agree, though with (my usual, monotonous) reservations.  External antennas are so fragile and fiddly, and invite ESD.  I was constantly brushing against the P2 plates, almost knocking it over like every time. 

And, beyond some basics, I tend to view optional physical adjustments as a net negative, as any little change can easily throw off muscle memory.  Sorta like having the neck of one's guitar or violin, or the keys of one's clarinet, subtly change between playings.  Breakdown and setup of one's Theremin shouldn't change basic ergonomic dimensions even slightly (unless that is the desired goal, which it won't be after one plays it enough).  Maybe I'm a princess with a pea, but just changing the D-Lev stand height 1" throws me off for a while.  I think I'd rather adapt to fixed, slightly sub-optimal ergonomics, than have everything shift like desert sands between gigs." -Dewster

Adjustments:
They would only be there for a player to set up and experiment with to find a preferred position before locking down.  After that it's a repeatable plug-in at assembly time.  The exception would be the volume loop which I plan to make like the type that Dominik uses on the Subscope:  only one end plugs in so that the loop can be rotated to any orientation. And repeatability for the volume side isn't really much of a problem anyway.  But any adjustable plates would need to be keyed for both precise assembly and to counter the inevitable imbalance that might cause unexpected rotation.  Rod pitch antennas are always consistent!

Muscle memory: 
I think that playing different theremins with different pitch fields has made me come to realize that once you set the zero-beat (low note) radius correctly that the brain has an amazing capacity to compensate for non-linearities on the fly as long as they aren't too extreme.  For the longest time I've had consistent difficulty with certain jumps (almost always downward) because I was trying to let muscle memory do too much and failing in the process.  I don't know exactly when it happened but things got much better when I let my ear stay in control at all times.  It means slowing down a little in training and it makes either pitch-preview or barely audible bread-crumb trails during transitions almost a necessity, but things seem to be going much better now.

Even the same theremin is slightly different every time you sit down and the only real solution to hit the notes is to rely on feedback.  I am amazed that anyone can play an RCA theremin; unless the one at the Pavek museum here is messed up it is amazingly non-linear.  If i can ever get back there I'll have a chance to see if I've learned anything, and I'll probably get humbled again. 

Posted: 1/25/2021 7:02:01 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

Roger, I'm wondering if one of your issues with the RP2 / EWPro ergonomics is the narrowness of the body?  You have a "wide stance" Subscope which I presume is comfortable for you, and maybe that's somewhat driving the width of your RP3?  I mean (I guess) maybe it's more than just antenna distancing?

What's your feeling on increasing the front to back horizontal dimension (over the ~6" or so of the EW) to make things more like the T-Vox?  Use that dimension to move the pitch antenna up and away, volume antenna lower and nearer?  Kinda like a wide RP2 angled back a lot more?

With my P3 (as with P1) I'm not seeing a ton of height difference needed between the plates, and a similar difference in terms of front to back distance, which could conceivably be accommodated by some moderate tipping up of a rectangular case like the T-Vox.

Posted: 1/25/2021 8:21:04 PM
pitts8rh

From: Minnesota USA

Joined: 11/27/2015

After rotating around through the Subscope, Melodia, and now back to the D-Lev I'm finding that even though I favor a little wider spacing I'm not that hung up on any one.  I think the Melodia and the Etherwave both suffer from low pitches being modulated slightly by the volume hand movements so I would say that their spacing is too close for what they are.  But my D-Lev Pro has almost a magical wall of isolation at the body of the theremin, so I'm going to make PIII with a standard Etherwave spacing up to an inch or two longer.  I've decided that I don't really want a close volume antenna as long as I can flip the hook-shaped one up and over if need be.

I don't think I want to go any more than an inch deeper than the breadboard enclosure that I've posted, and that extra space is only to allow for the cover mount bushings and such and to give a little more sloped panel space. I have to stay within  the constraints of a 12" wide piece of acrylic, and I'm almost there.  I don't want to have to buy 2'x2' cell-cast sheets just to get a couple extra inches because they're not that cheap to render a third of it unusable. I'm all about going compact these days - with the long case there is plenty of room for everything including nearly full-size inductors, so this is really efficient even if it isn't anything ground breaking for theremin design.

Posted: 2/4/2021 2:25:23 AM
pitts8rh

From: Minnesota USA

Joined: 11/27/2015

PRESENTING....

The ClearaVox Digital Theremin!

Actually this is my third installment of Dewster's D-Lev theremin intended for use as an easy-to-access-and-modify shop unit.  The big deal here is that I made a first attempt at bending up and machining a 1/8" acrylic cover/panel, and despite a lot of scars and some sloppy bends I'm getting some things figured out.  It's hard to see but there are cutouts for the tuner LEDs, encoders, and both the 7-segment and LCD displays.  The intent is to have replaceable clear windows for the displays with the housing made of colored opaque plastic instead of the clear as shown.

I used clear acrylic here because I have quite a few sheets to practice on before attempting to use one of the few colors that are on hand.  It turns out that bending a 12" wide sheet as was used on my very first prototype is a lot easier than bending the 20" sheet used here.  It works better to bend at the lowest practical temperature and if possible to put the heated side of the plexi on the inside of the bend instead of the outside.  Warpage after cooling has been a big problem, but keeping the bending temperature as low as possible has helped this a little.

The wood jig that is holding the acrylic panel in the milling machine also serves as the forming pattern for making the hot-wire bends in the acrylic:

 

The form can be rotated to put any of the four faces horizontally flat if there should be a need for holes or engraved text on any side other than the sloped front.  In the first picture above the "MUTE" text was engraved in the top as a marker for a capacitive touch-pad or IR proximity switch. I've always thought that it would be a good idea to have the mute switch located where your hand is already muting the volume just by reaching for it. Whether or not the capacitive switch will work here remains to be seen - there's a good chance that it or the volume circuits won't like the close proximity.

After slipping everything together temporarily I sat down and played it for nearly an hour with wires still hanging out.  I'm really liking this enclosure style and already thinking about what color panel and what type of hardwood ends to put on the real one, which may be the next or maybe a third iteration. Like the SpaceX Starship rocket, you really have to build and crash a few of these things to get them right.

Posted: 3/6/2021 1:42:02 AM
pitts8rh

From: Minnesota USA

Joined: 11/27/2015

New boards for my D-Lev PIII prototype (a.k.a the ClearaVox)

When I last posted pictures above of my horizontal D-Lev mockup enclosure I was using the same pcbs that went into my Pro (the Etherwave Pro sorta knock-off cabinet).

There were several fairly major changes made to the board designs to eliminate having to make cables and make things more compact.  I don't mind making ribbon cables but anything else with crimp-on terminals is pretty tedious.  The biggest change is that the MAIN FPGA board had two separate harness-connected encoder banks whereas this new MAIN ENC pcb has been enlarged to accommodate integral encoders, or almost.  The encoders have their own individual boards that are elevated on header pins and plug into the main board, which also allows for correction of the height disparity between the LCD display and the encoder mounting plane.

This new main board also includes Eric's pwm dimming for the main LCD display backlight, which is working well. Other changes to other boards are pretty minor, mostly driven by cabling issues.

Here is the latest batch of boards from JLC.  The three largest boards are the new MAIN ENC boards, and the rest are as follows: center row from upper left - 1) 4x encoder pcb, 2) a long narrow SMD analog front end board for my PII project, 3) 4x4 panelized encoder boards, 5) tuner boards.  And in the upper right corner below the main board we have the 5) standard discrete AFE3 boards, and 6) the new dual channel audio board.

Below is the new dual channel audio board.  This has analog amps with pots on top driven by two piggybacked spdif digital-to-analog converters (these are purchased) on the bottom.  This single assembly has two separate digital inputs, each with a left and right audio output for a total of four outputs to feed two 1/4" stereo jacks for pseudo stereo main audio and pitch preview.

This is the back side of the new MAIN ENC board, and through the reflections you can barely make out how the encoder boards are mounted to the acrylic panel and how they plug in to the main board.

These next two pictures shows the pitch and volume inductors.  The leads are excessively long and folded because I don't want to cut them too short for when they go in the real cabinet.  The 3D printed ends for the PVC cores look almost like they should have ball bearings to allow the little drums to spin .

This is a wider view of what this thing will look like someday.  Complete with the ubiquitous industrial roll of toilet paper.

And a little closer view.  This is my hack and slash test platform...  If I need new holes I make new holes.

This photo shows a pretty good view of the analog board below the tuner with the piggybacked decoders on the bottom. This is a lot more compact than using the decoders in their metal boxes with RCA cables going in and out. 

That's all I have for this evening.  New boards and some current boards with minor mods are already in production and are expected next week.

Posted: 3/6/2021 12:49:35 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

Beautiful (as usual) Roger! 

Pop quiz:
1. You're "riding bareback" with no ESD ICs on the AFEs?
2. I see Dith=2 on your pitch axis, do you find that necessary?
3. What's the electrical connection inside the printed volume loop support?
4. Where are the audio outputs (they seem to be hidden in the pix)?
5. What are the bottom facing LEDs on the audio board for?
6. Why does the audio board have a missing cable on the right?
7. Can your audio board drive headphones?
8. So it seems everything except the AFEs and coils are mounted to the clear panel?
9. How is the tuner being held in place?
10. How is it powered?

TIA!

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