Crazy (?) theoretical / technical ideas

Having just posted this, I can see its probably utterly useless and works to make linearity worse not better! ;-)

But if I delete this post, one gets a dead link to this page - so I will leave it for now and either correct it if I find a solution, or perhaps delete it later..

====================================================

Ok - up till now the TX/RX topology above perhaps wasnt too crazy.. But thats about to change! ;-)

I am wondering if the above antenna construction (top right corner) could do anything to improve linearity..

C1 is the capacitance between TX ("Shield") and ground which is fixed by the antenna construction (lets say probably about 1nF)

The antenna and shield are couped by their fixed proximity (say 100pF), The oscillator frequency will be seriously affected by this coupling - Increasing this coupling will lower its frequency (series coupling to the antenna inductor)

The antenna has direct coupling (likely to be lower than usual due to the shields effect) to ground (C4) and to the player (C3).

The player will be coupled to ground by (and this is the bummer) an inconstant capacitance C5) value.

The player will ALSO be coupled to the transmitted (TX "Shield") signal "leaking" from the antenna assembly (C2).

(In my prior simulations I have a capacitor between the oscillator output and the "shield" equal to the value of C1 - but this complicates explanation more that I can manage, although I think it will be needed)

The player becomes quite a complex entity with the above - Both C2 and C3 will be varying as position changes, and there will be (I think) a mixture of coupling effects - There will be the usual increase in capacitance as a result of increased ground coupling on the antenna as the players hand approaches it, causing a lowering of the oscillator frequency - but there will (?) also be an increase in capacitive coupling between the shield and antenna (effectively via C2 -> C3) acting to increase the oscillator frequency. (Where did this nonsence come from ???   ;-) !

The question is, can the antenna assembly be engineered to actually cause these opposing effects to implement linearity?

An alternative / additional idea is to have a couple of seperate ground shielded "TX" antennas oriented to provide increased "TX" coupling to the player in the near field:

(Actually, the only hope would probably be to focus these TX antennas to the far field)

I actually have no idea if any of this idea has any merit, either theoretically or practically - I may be thinking nonsense.

Fred.

"The question is, can the antenna assembly be engineered to actually cause these opposing effects to implement linearity?"  - FredM

I think the vertical oblong shape (rod, narrow plate, etc.) is kind of hard to beat as an all-rounder, though I'm sure there are antenna configurations and geometries that give more linear response for a given oscillator / heterodyning scenario.  To me, the really intriguing aspects of these types of crazy designs are the opportunities for (1) "one sided" directionality and (2) common drive radiating from all antennas.

Crazy Idea e^(i*pi)+1 : Player TX

Make the player radiate RF and use relative signal strength at the various antennas as control inputs.  Maybe use spread spectrum to get around any sharp tuning / selectivity issues.

Obvious problems: (1) no one wants to "put something on" in order to play an instrument, (2) there could be shock hazard issues associated with the direct coupling apparatus, and (3) you've basically got a bunch of AM radios.

"Make the player radiate RF" - Dewster

I think something like this was done for Pen and Teller (Theremin chair ?) - May be wrong - might have been something else .. I will see if I can find it..

What about PRBS RF on the player, bluetooth or other "coupling" taking this PRBS directly to the recievers to compare against the E-Field coupled PRBS?

Nah - this aint just crazy, its plain silly! ;-)

Spend hundreds of £ on improving something that was as good as it can get way back in the 30's, and cant get any easier to play unless we (or evolution) modify the player!

;-)

MIT Document re TX RX Etc re Penn+Teller

"Spend hundreds of £ on improving something that was as good as it can get way back in the 30's, and cant get any easier to play unless we (or evolution) modify the player!"  - FredM

Ain't that the truth! 

Nice paper Fred, thanks for the pointer! I see that their sensors also uses transconductance and synchronous detection.

[EDIT] Over on the "Beginner" thread, Pearl mentioned the AirPiano, something I've never heard of.  From the videos it looks like there may be significant delays during play.

"Pearl mentioned the AirPiano, something I've never heard of.  From the videos it looks like there may be significant delays during play." - Dewster

The "AirPiano" is, IMO, in an entirely different "class" from instruments like the theremin or violin or other free-pitch musical instruments - in fact, IMO its in a different "class" to "normal" musical instruments - I see it in a "intellegent controller" class - the same class that a laptop computer falls into when being used to trigger playback of samples or sequences from its keyboard or mouse or whatever..

I personally have a strong dislike for this kind of "instrument" (or at least the use of this class in a performance setting) - it is "personal" and probably irrational, but I find performances where a "musician" is standing over a laptop hitting keys that trigger some sequence or noise a real irritation.. I was quite heavily involved with (and even played with them on a couple of performances) a band called "Kangaroo Moon" oh, Many years ago ( I was first introduced to the digeridoo by them, got lessons from them and mastered it, and was also learning Djembi from master Kenyinda Coco Makala, which was at the time my main instrument)  - I loved (and still love) their early stuff (Their album "Bagpipes on the beach" is one of my all-time favorite albums) .. They all played their instruments back them, they were wonderful musicians - nothing was canned, and they didnt need any canned backing, they were good enough and could play live.

But they got into the dance scene, the flutes and violins and didge gave way to pre-recorded sequences triggered by some (IMO) unskilled PC key pusher - I think they may have found their way back to real music now, but the music they produced in that phase was not IMO worth listening to.

This "controller" class of instrument includes controllers that can be used to "really" play (A weighted master keyboard can be used to perform a Rachmaninoff piano concerto and also be used to trigger any MIDI sequence or whatever) but also includes instruments like the "AirPiano" which is IMO almost useless for real playing, and can really only be used to trigger sample playback - You couldnt do a piano concerto on it, or even play a moderately complex piece of music on it unless its pre-programmed and planned. (to me, a musical instrument is an instrument that a musician could pick up and play - get a musical idea / tune, and be able to get that "out" without difficulty, be able to be with other musicians and not need to program a sequence in order to join in)

The whole instrument / controller / effect differentiation started to get lost IMO probably in the 60's, when the recording studio became seen as a sort of  "musical instrument" in its own right... To me, thats fine and "correct" - It IS a "musical instrument" in terms of producing music.. The problem for me is in the area of live performance - I want to see (and hear) musicians perform, I have no real interest in going to some venue to watch them play back pre-recorded stuff they did in some studio - I can do that - I can buy their CD and push the "play" button !

Fred.

"I see it in a "intellegent controller" class - the same class that a laptop computer falls into when being used to trigger playback of samples or sequences from its keyboard or mouse or whatever..

I personally have a strong dislike for this kind of "instrument" (or at least the use of this class in a performance setting) - it is "personal" and probably irrational, but I find performances where a "musician" is standing over a laptop hitting keys that trigger some sequence or noise a real irritation."  - FredM

I have the same reaction to these sorts of things.  Give me a playable instrument or give me death.  And the next time I see someone triggering a complex sequence with the flick of a controller it will be too soon.

But the distances he's getting with LEDs and a crappy Arduino are impressive (averaging can accomplish all kinds of magic, too bad about the latency it brings with it).  And the different notes at different elevations is something I wouldn't think of doing.  I don't see a way of controlling the velocity of individual notes though (unless it's the axis perpendicular to the "keys"?).

Crazy Idea elevendy-twelve : Gyrator "LC" oscillator

Make a low frequency Theremin by replacing all the tank inductors with gyrators.

Been playing with this in LTSpice for the last hour or so and, due to op-amp delays and such, the operating point seems to be fairly touchy with respect to drive.  And of course you don't get the crazy voltage boost you get from real, manly-man, beefcake inductors.  Oh well.

The gyrator is on the right, the equivalent inductance is C3 * R4 * R5 = 108.9 mH.

Via the Colpitts drive & sense on the left, this resonates mainly with C4, which gives f = ~84 kHz.

If you could get this anywhere near as stable as a real LC you might could use it as an easily tunable (via R4 or R5) local oscillator.

Hello Dewster ;-)

Desperately trying to restrain myself from posting - I hate having this name filling all the latest posts ;-) .. But I cant restrain myself, LOL ;-)

Yeah, gyrators are lovely - in theory! - I have never got on with them though - Oh, there are places (graphic equalizers for example) where they are the only way to go in terms of size and cost - but...

I remember that someone (I think it was Art Harrison) at some point set out to eliminate pesky inductors by using gyrators in theremins.. Probably most theremin designers and tinkerers have tried to get rid of L's and replace them with gyrators, or replace oscillators with RC ones... I went this route at first and got my fingers badly burned - I learned that a beefed-up antenna voltage is worth the hassle of having inductors! ;-)

One day, someone might pull it off - but I suspect that the route to an inductor-less theremin is probably going to be digital - spread spectrum stuff where voltages are low but the inevitable interference is eliminated.. I dont think there is any hope of achieving a reliable analogue theremin front-end without L's, and if one must go inductor-less then you can probably get as good results using a TS556 running at 15V and biasing this for maximum voltage triangle wave on the antenna, and giving it feedback to provide linearization - I thought I was onto a winner with that, but tests showed that it was still just a toy with almost no prospect of ever being anything more than a toy... I think the same may be true for all analogue inductor-less designs.

Used purely as a local oscillator, the idea should work - But if one is doing the tuning at the LO (something I try not to do) and using an LC for the antenna (variable) oscillator, then I see little point and lots of headache in the approach - There are lots of ways to tune an LC local oscillator.

So my vote is that this idea qualifies as crazy! But perhaps the really crazy thing IMO is that so many (all?) of us are likely to explore this route at some time in our "journey" ;-)

Of course, I could be completely wrong! ;-)

Fred.

"I remember that someone (I think it was Art Harrison) at some point set out to eliminate pesky inductors by using gyrators in theremins.. Probably most theremin designers and tinkerers have tried to get rid of L's and replace them with gyrators, or replace oscillators with RC ones... I went this route at first and got my fingers badly burned - I learned that a beefed-up antenna voltage is worth the hassle of having inductors! ;-)

One day, someone might pull it off - but I suspect that the route to an inductor-less theremin is probably going to be digital - spread spectrum stuff where voltages are low but the inevitable interference is eliminated.. I dont think there is any hope of achieving a reliable analogue theremin front-end without L's..."  - FredM

Oh man, we all end up going down the same roads it seems.  There aren't an infinite number, only a few make sense - and this one is (thankfully, rather clearly, right out of the gate) a dead-end!  Got the idea for Theremin gyrators while reading Self's "Small Signal Audio Design".  (Can you tell I'm in the EQ section of the book?  He's an incredible author.  Doesn't "hand things down from on high" like a lot of technical writers - thanks for recommending it Fred!)

Spread spectrum capacitive sensors usually employ somewhat highish voltages (perhaps the highest achievable without getting too wild about it?) though reaching the voltage levels that a simple LC is trivially capable of can be difficult.

LC so easily wins in Theremin applications because the tank stores a slug of swinging energy, which is added to incrementally each cycle as drive, and subtracted from each cycle for sensing (and Q loss).  Livio's Colpitts shows this clearly from a schematic standpoint, with a tiny 10pF separating the tank from the drive/sense.  In a capacitive sense the LC is a high power, highly frequency selective, transmitter and receiver all-in-one, and you have to get up pretty early in the morning to have any chance of outwitting it. 

[EDIT] LC is also FM, which is hard to beat because most environmental interference is AM.  And it's periodic HF, so heterodyning gives you audio or increased digital resolution.  I forget - why do I keep looking for alternatives again?  Oh yeah: inductors have thermal stability, magnetic pickup, and bulk size issues; emissions are problematic; multi-sensor arrangements in close proximity to each other don't play nice; and mains hum is difficult to deal with in the FM domain.