"I don't even know if we have a French word for it over here !" - Mr_Dham
Before Jeff said it, I didn't even know we had an English word for it over here! ;-)
Our town of Boonton recently celebrated its Sesquicentennial, which was my first introduction to this odd anniversary terminology: [LINK]
Leakage aux Folles
I noticed that the Bollamin cried like a baby when Jeff was performing ACAL, so I made ACAL initiate a pre-mute when using traditional volume sense (farther=louder) which saves a step / everyone's nerves. The user then has to manually take it out of mute post ACAL. There is no mute for "modern" mode (I keed) because far=muted already. Yet another reason to reverse the volume field! ;-)
In preparation for the D-Lev being used in unattended situations, I expanded the field swapping modes (P less than greater than V above on the SYSTEM UI page):
0 : normal
1 : pitch follows volume field
2 : volume follows pitch field
3 : pitch and volume fields swapped
Along with this there is a new knob Leak[0:7] on the SYSTEM page that implements a continuous very slow asymmetric ACAL option. This is useful for "sculpture mode" where one wants the null points to stay way out in space, very slowly adapting the presence of bodies, and much more quickly adapting to the absence of them. This mode can also be used to examine noise levels in the room, and for "demo" type situations like in a store, so people can walk up to it and start making noise without knowing anything about ACAL and such. As you might expect, it requires a reversed volume field to properly function. Alternatively one could have the pitch field control the volume (new mode 2) or volume control can be disabled (constant).
To give users some control over the pitch field gestural bandwidth, there is a new knob Trak[0:7] on the P_FIELD page that sets the pitch tracking filter octave shift, with limits of 100Hz max and 10Hz min. So if you are playing an A4 (440Hz) and have Trak set to 2, two octaves below this is 110Hz, so the tracking filter is hitting the max of 100Hz. When playing an A3, 2 octaves below this is 55Hz, which is within the limits, and so the tracking filter cutoff frequency (an 8th order low pass filter) is 55Hz. This kills hum, and also helps to attenuate internally generated interference. On the volume side this filter doesn't track, but is fixed at 100Hz. Gestural bandwidth seem to be much more important on the volume side, so it should probably be constant. Environmental noise modulation of the volume (not that I've ever seen or heard it do that) is likely much less noticeable than modulation of the pitch.
And with that, the SW is done! Or at least there are no remaining UI slots! I've updated the librarian to match these changes, need to compile to Windows and update the D-Lev web page and manual.
Desoldering - All I Really Need To Know I Learned In Kindergarden
I installed a voltage regulator the wrong way on one of the kits, and it was a bear unsoldering it and cleaning out the holes. Luckily the PWBs are extremely durable and can take a bit of rework without damage. I ordered one of those iron-and-soldapullt-in-one thingies and it both arrived today and I had the extreme misfortune of having an immedate need to use it (don't ask):
Some reviews say it doesn't get hot enough, but the heat level seems fine to me. The problem is the suction isn't nearly as good as my big blue plastic soldapullt (I hope whoever came up with that ingenious name got a huge bonus) and you need both heat and suction to get the job done, even when the PCB hole is only full of solder and the wire or component lead has been removed.
Turns out I don't need either to do a top notch desoldering job. My secret? Inertia! I heat the offending solder up with my regular soldering iron and toss the board down on my workbench top - all the hot solder jumps right off and leaves only a thin solder plating behind. When I was a wee soldering lad I'd throw it on the floor, don't know how I learned it. Sometimes you have to add solder to get it to work, which is a little counter-intuitive, but not so much if you think about it, as it all has to be liquid, and more thermal mass helps keep it liquid all the way to your workbench / floor (this goes for using a soldapullt too). You can even hold one end of the board and tap the other end kinda hard on your bench and get similarly excellent results without the thing bouncing around the room. Just watch your eyes! Close them when banging or wear protection, believe it or not you really don't want molten lead in your corneas, trust me on this.
Roger told me that a hot plate is a good way to remove parts like connectors from a board en-mass for reuse. I haven't personally tried that but I'm sure the day will come.
You Can't Tuna Fish
The initial wave of 10 kits is progressing. Ran into an issue and resolved it, but it involved rework on the main boards, all of which took about a day. Now I know why the LCD backlight is so dim on the Bollamin - the transistor is backwards! I "assumed" TO3 transistor pin-outs were standard, but it seems not and I should have checked. I had no clue backwards transistors would even work - they do indeed after a fashion, but behave like their current gain (Hfe or beta) is quite low. The transistors in the AFEs are backwards too, but at least they are socketed (a quick fix). The backlight is nice and bright now, and I redesigned the backlight PWM dimmer drive circuit a bit since I was in there anyway:
You basically want a constant voltage at the transistor base, which puts a constant voltage across the emitter resistors, which draws a constant current through the collector, which gives some regulation. R1 & R2 divide the 3.3V output of the FPGA pin to 1.65V, this gets dropped one diode Vf of 0.7V, so we have roughly 1V across 47||47, or 42mA. It's actually lower because the 47||47 is reflected back through the base by the transistor beta, so somewhere around 2k to 4k in parallel with R2. We're powering the nominally 3.3V backlight circuit off of 5V to give the transistor switch some headroom (and to unburden the FPGA 3.3V regulator) - it seems to work well between 4.5V and 5V, which should allow for any supply droop.
All the tuners are printed and assembled and tested, 6 of the main boards are printed and assembled and tested (one each are currently in the Bollamin):
Still have 4 main boards and all the coils and AFEs to go which might take another week or so. Printing the last knob set today, and all the FPGAs have been pumped. This is a little too much like work! ;-) Looking forward to doing one at a time, here and there.
Guitar Face
One positive side effect of the typically overly sensitive (IMO) analog Theremin pitch field is that it severely limits players ability to emote via writhing bodily movements on stage. Now if it could only somehow rein in Indie Voice...
[old_man_yells_at_cloud.gif]
Upstairs Downstairs
How does the Theremin / player ground loop even work when you're upstairs in a wooden house? I think this is at least half of what I'm seeing in the way of disturbance when playing upstairs - my body doesn't have nearly the ground reference it does when standing downstairs on the cement slab floor. Upstairs the Theremin itself is almost floating in air yet grounded via the electrical outlet. Weird.
Nearer My Plate To Thee
I've been doing video show-and-tell chats with the D-Lev kit buyers, and during the course of the meetings I find myself beating these uniformly wonderful and interesting people over the head with "the plate spiel". To me, plates are integral to the design because they have many engineering and playability advantages over rods. I think I'm probably coming on too strong when advocating for them, but I feel it has to be done and there is no one else to do it. Of course, the kits are for folks to experiment with however they like, which should be really interesting to see. But, just so no one feels left out, let me subject the good folks here at TW to "the plate spiel" too! ;-)
First, a tiny bit of physics. Theremins work by detecting the capacitance between the hand and the antenna. Capacitance is a function of two things: 1) the area of the plates, and 2) the distance between them. Increasing the area increases the capacitance, decreasing the distance between them also increases the capacitance.
Rod antennas make a lot of sense on analog Theremins because the near field is so compressed and non-linear. When the hand is far away it's interacting with the entire rod, and as the hand approaches it interacts with less and less of the rod, which helps somewhat to relieve the tightness near the antenna. But, because rods have rather low surface area, they aren't a great engineering fit for digital Theremins because they don't form the best possible capacitor with the area of the hand.
Rod antennas also have a "targeting problem" in the horizontal plane. The closer you get to the rod, the more precisely you must position your hand horizontally to address the antenna. Some players utilize this and play slightly off to one side of the rod in the near field to help linearize it. Some players likely alter their fingering techniques depending on how close they are to the antenna.
Plates form a more ideal capacitor with the hand, both in the far field and the near field. Because a digital Theremin can linearize the near field to match the mid and far fields, players can use the same fingering techniques anywhere.
A plate is a much softer horizontal target; correct targeting in the near field is greatly eased due to the width of the plate. Though of course vertical targeting is somewhat reduced over that of a long rod, but vertical alignment is generally about as easy to maintain as horizontal alignment, so this isn't a terribly significant issue.
Due to the tightness and the targeting issues, I imagine a lot of analog rod players simply avoid the near field whenever possible. This is unfortunate in the extreme because the near field is the best possible playing position! When the hand is near the antenna its capacitive signal is quite large, swamping the otherwise confounding environmental and body capacitance. Playing a Theremin from a distance looks impressive as all getout, but it becomes exponentially more difficult to control once the hand leaves the antenna and nears the body, where body position and movements start to dominate the total capacitance. So as a player you should want to play as close to the antenna as possible, and the field linearity and antenna geometry should accommodate you as much as possible here.
In conclusion: a digital Theremin with a rod has targeting issues and doesn't form an ideal capacitor with the hand. A digital Theremin with a plate has none of these issues, and so it allows you to use the same fingering techniques everywhere, including the exceedingly desirable near field "sweet spot".
On Plate Antenna Geometry
One might think a round plate would be the ideal shape for a plate, but I don't believe this to be true. With a round plate if you are off center vertically then the horizontal target width is reduced, and vice-versa. A rectangular or square plate doesn't have this interaction, and is therefore superior.
I also think the difficulties in vertical and horizontal targeting precision for the player are roughly equivalent, which means we should use a square plate. To get roughly the same surface area as the hand, the plate should be somewhere around 150mm on a side.
To increase far field capacitive interaction, we can extend two opposite, or all four, sides of the square plate back in space away from the player, forming a U or open back box type 3D shape:
I'm currently using the U type shapes in my lab unit, formed out of heavy aluminum foil:
The above is a rear view with the back removed. The foil covers the front and goes up the left and right sides. The coil is smack dab in the middle.
Rod antennas also have a "targeting problem" in the horizontal plane. The closer you get to the rod, the more precisely you must position your hand horizontally to address the antenna. Some players utilize this and play slightly off to one side of the rod in the near field to help linearize it. Some players likely alter their fingering techniques depending on how close they are to the antenna
Maybe it is the challenge of the transiion from one kind of antenna to the other. By playing slightly off to one side of the rod, I know that I consider the difference between closed and open hand more in term of surface rather than in term of distance between my finger tip and the antenna. And it has a lot of consequences in my playing technique.
Anyway, I am getting currious enough about plates to want to try them. Would it be only to get better understanding of what is involved between the changing shape of the hand and the antenna.
"Maybe it is the challenge of the transiion from one kind of antenna to the other." - Mr_Dham
I agree. And there's nothing wrong with players wanting to transfer / leverage their current skills on a new instrument, I totally understand that. It's just that the D-Lev offers new opportunities that I myself use, and I want others to be aware of their presence and value, even if they decide to go in more traditional directions with the kit (which it does support).
"By playing slightly off to one side of the rod, I know that I consider the difference between closed and open hand more in term of surface rather than in term of distance between my finger tip and the antenna. And it has a lot of consequences in my playing technique."
I believe this is the case for Roger too.
"Anyway, I am getting currious enough about plates to want to try them. Would it be only to get better understanding of what is involved between the changing shape of the hand and the antenna."
It's probably a lot (too much?) to ask of any player who has invested significant time and effort in developing analog / rod techniques to give plates a shot - but I'm asking! ;-)
It's not a Theremini, it's not a Harrison, and I'm not pushing plates for ideological or contrarian reasons - the physics of plates picked the D-Lev, but they offer some real playing advantages IMO.
That said, rods will work too, and I don't want anyone to feel like I'm strong arming them into something they don't want or like.
You must be logged in to post a reply. Please log in or register for a new account.
