TECHNICAL HIJACK!

Quite often, some innocent thread gets taken to technical extremes.. This thread has been created as a dumping ground for such discussions... Anyone is free to use it, but being the most guilty of hijack, I will probably use it most!

Please dont hijack this thread! ;-)

--------------------------------------------

Escape from Theremini Review Thread:

"Funny, I've been contemplating this conundrum myself and don't have a good answer.  Though the main "interferer" that high antenna voltage overcomes could be the internal noise of the oscillator?" - Dewster

Interesting idea.. But I still have problems...

Its the simulation side, the fact that I get (for similar LC values) the same kind of responses when I simulate with low antenna voltages (oscillators like the UNO / Theremini) or with high antenna voltages (Series LC like yours or my more complex ones ;-) - I am talking here about single stage LC, not complex parallel tank / series resonant configurations.

Ok, I dont trust the simulations down at the fine change in C levels one gets at the far-field (and I certainly dont trust my models of active components enough for issues such as the internal noise when in a sensitive circuit like a CMOS oscillator)  - but when one puts a spreadsheet together which is determining frq based on LC, and uses a table of realistic capacitance values derived from antenna / hand capacitance, there is no difference to the curves for different drive / antenna voltages.. One a purely LC basis, there should be no difference.

The fact that there is a difference must come down to real-world factors - interferers, oscillator noise as you say, those kind of things...

(I certainly dont regard this "problem" as any cause for doubt about capacitance being the mechanism by which the sensing operates - when one has adequate antenna voltage, then one gets the results you expect for the degree of capacitance change caused by a player, no more, no less.. Its only when antenna voltage is low that I have found greatly less change in oscillator frequency in the far field than what one should see - and also greater change close to the antenna than one should see.)

But is there something else? ... I am at a loss really - I know its crazy talk, but I wonder about electron 'density' in the skin of the antenna.. Or is it some ground 'dipole' effect - lower voltage giving the 'ground antenna' more influence, and thereby increasing  non-linearity ?

But at a simpler level, if one looks at the electrical activity in a theremin environment, and the electrical fields effectively 'moving' the 'ground reference' (things like power cables effectively modulating low level fields) perhaps it could just be that far-field "mush" simply drowns out any usable signal (or the mush is so bad that filtering is required, which attenuates useful signal in the far field) - but higher voltages on the antenna overcomes this and makes the far field more usable.

I really dont know!

"If they are heterodyning, then TP_PRO is likely a counter or PWM output from the DSP, the frequency of which is set at the first step of calibration, where you get away > 4' and it counts down for 5 seconds." - Dewster

Yes, this seems likely.. May also be the reason that PO signal is sent to the DSP - It can be directly read then, and the PRO frequency simply based on this (add 1kHz or whatever) - particularly as they probably couldn't know from the TIP signal whether the PO frequency was higher or lower than the PRO signal without doing several adjustments! - 5 seconds to count the PO frequency, write the required PRO frequency to the counter / PWM.. Simple ;-)

Fred.

Could it be that lower voltage swing designs simply have more capacitance in parallel with the antenna, swamping any delta C?

The problem with the UNO and Theremini oscillators is that the antenna "sees" the full capacitance that it is parallel with, but the inductor only "sees" the reduced (series) capacitance for resonance.  Unbalancing the capacitances ameliorates this and provides a larger voltage swing.

"Could it be that lower voltage swing designs simply have more capacitance in parallel with the antenna, swamping any delta C?" - Dewster

Yes, this is true - series LC allows for smaller parallel capacitance and therefore greater delta C..

And I would rest with that if I hadn't seen sensitivity / linearity improve on parallel LC oscillators simply by increasing the supply voltage on these.. But oh, it could be so many factors.. A 4069 based oscillator like that in the theremini, driven from 12V vs driven from 5V will have higher antenna voltage - but loads of other factors which could influence performance will also change.

Not really worth bothering about when series LC oscillators work so well and are so simple to implement - just annoying to have this uncertainty about the reasons why the work so  much better! ;-)

Re C2:  I doubt that increased voltage for ESD purposes is the reason! - If voltage at that point ever rises to >= 50V the IC will have fried anyway, and if it doesn't fry (as in, ESD does not exceed the IC input specification) the voltage on C2 will be clamped to the rails.

Fred.

Found livio's video of FET LC vs. 4069 LC (with plate antenna, 4069 adapted to 2.7MHz).  His discussion at YouTube is regarding noise rather than pitch field size:

It would be interesting to make an LC oscillator with easily adjustable voltage swing and do an investigation of what the heck is actually going on.

"Re C2:  I doubt that increased voltage for ESD purposes is the reason! - If voltage at that point ever rises to >= 50V the IC will have fried anyway, and if it doesn't fry (as in, ESD does not exceed the IC input specification) the voltage on C2 will be clamped to the rails."  - FredM

So obvious when you state it that way!

Hi Dewster,

I really dont know what to make of that video, but the bottom line is that I dont trust it! For one thing, there is no data given regarding relative filtering on Open.Theremin vs Theremino - if the audio pitch and related displays are taken purely in terms of audio frequency derived from processing the direct PO from in the Theremino, and the heterodyne output from the UNO, well - I would be more suspicious of the theremino (certainly in terms of latency) and think the UNO pitch looks far more "real".

I really dont see that video as "evidence" either way.

Having said the above, I have no doubt that the BF862 oscillator is less noisy than the 4069! - What I cannot tell is whether the difference actually has any significant (or even measurable) effect on anything.. So far I have not seen anything that leads me to think it does! - Does the video ACTUALLY show anything related to oscillator noise? - I doubt it!

But as you say, that's just one "side" aspect (noise); -- it may not relate pitch field size or linearity in any way.

Just to emphasize - I think in terms of analogue - direct from oscillators to audio via heterodyning.. And with this topology I think phase noise on the pitch (and/or reference) oscillator needs to be quite bad before it can be heard at the audio - a single audio cycle is effectively the "average" of quite a substantial number of HF cycles (1kHz audio with HF at ~250kHz will average 250 HF cycles - more for lower AF Frq, more for higher HF Frq.)... I suspect the same is true for topologies like the UNO which use heterodyning, but with direct topologies like Theremino, some sort of deliberate averaging probably needs to be added.

And to me this is the crunch - with digital theremins, one can display a perfectly 'clean' signal simply by filtering out any 'noise'.. The filter is likely to be there in some form in all such implementations .. But unless one specifies the filter / averaging parameters (or matches them to ones competitor against which one is comparing your product) one can make your product look far superior, even if its inferior.

The problem with averaging a large number of samples (and / or aggressive filtering) is latency -

Please note - I am NOT saying which is "superior" or that any deliberate deception is occurring, or even that the video cannot be taken at face value.. All I am saying is that, IMO, there is not enough data given for a viewer to make a judgment.. Or certainly not enough for me to make a judgement!  ;-)

Fred.

Of course what I write below is embellished. (-‘

Fred I have followed your ideas and wish I could better understand the technical side of things. When you put at the top of your posts “Permission is granted for use of this material for non-commercial use.” I imagine you are politely asking the good people on the Internet not to copy your designs, made public by you, and then profit from them? That seems well intentioned until you start assisting that other guy who cracked open the box and began exposing our proprietary technology and now shares it across International boundaries, I am not talking about Richard Snowden. Do you ever watch YouTube videos where the “stupid” kid breaks the law and records it so the World can watch it on social media? If the box cracker thinks he is invisible… we know where he lives, that he has a net value of more than $1,000,000, which is why he can take those nice vacations. If our product that is being violated does not come out of the red in sales he will be paying some salaries around here.

The Boy Scout Motto is “Be Prepared” or The American Motto “Sue Everybody”

T

[OT - NOT "Technical Theory" - More about ethics, copyright, and such subjective matters] - All my following comments are my opinion only, and may well not be shared by TW management or anyone else - I AM NOT A TW MODERATOR!!! I am merely a TW member with no other "affiliation" to TW ( I emphasize this because some people have emailed me under the misconception that I "represent" TW in some way - I DONT! )

" I imagine you are politely asking the good people on the Internet not to copy your designs, made public by you, and then profit from them? That seems well intentioned until you start assisting that other guy who cracked open the box and began exposing our proprietary technology and now shares it across International boundaries," - T

1.) Copyright does not apply / is not relevant on reverse engineered products, it would apply if, for example, one was to copy a page from a service manual - but DOES NOT APPLY TO IDEAS! .. To get protection for a new idea or method or procedure, one needs a PATENT! (and in order to get a patent you must disclose the idea fully enough that anyone "competent in the art" can understand it)

2.) Absolutely nothing "disclosed" by "that other guy " is "proprietary technology" - there is nothing that isn't in the public domain, and, quite frankly, nothing which hasn't been copied from somewhere / someone else!

3.) There is nothing to stop anyone from taking the ideas in my "copyrighted" material and incorporating this into their designs, and selling these - The ONLY thing that copyright prevents is the copying of my DOCUMENT and DRAWINGS etc and re-publishing this.. As in, if someone was to copy what I have posted here and put it in a .pdf or include it in some book or article for which they charged a fee, And/Or they were to extract a section of my copyrighted material and publish / distribute this out of context,  THEN and ONLY THEN would they be in breach of what I, the author and copyright holder, have specified. (and to be honest, there would be F*** All I would do about it if my conditions were breached - But its there just on the off-chance that I land a publishing deal on my book - Publishers aren't happy if the author has unconditionally given away rights to material that is included in a book - they might choose to defend their authors material if it is copied illegally, and even do so retrospectively)

4 "that other guy who cracked open the box and began exposing our proprietary technology and now shares it across International boundaries"

So you are worried about this amazing technology falling into the wrong hands? Well let me put your mind at rest - There is NOTHING in the theremini which would be of the remotest interest to anyone except us nerdy theremin engineers! - There is NOTHING that is even remotely at any technological cutting edge.

Your words "our proprietary technology" puzzle me a bit - Who is the "our" ? Unless you are a Moog employee, the illusion of "our" could not apply (I say illusion because everything revealed is common circuitry in the public domain) in terms of Moog's "proprietary technology" -

Who else could "our" be referring to - some entity you feel part of, obviously - The U.S.A. ? - No, I dont think there is anyone who could frame the theremini in that context, surely ???

Lets spell it out - the theremini is being sold internationally.. If there was any "proprietary technology" of any interest to anyone, there are establishments which could strip it down to near molecular level, and gain full access to every aspect of its construction and workings - Less than 10 miles from me there is a facility equipped with tunneling electron microscopes used to reverse engineer IC's!  But the theremini wouldnt need this - a couple of weeks (or less) analysis in a facility in China, and the Chinese could be producing theremini within months. (anyone copying the PCB ARTWORK directly, or copying the shape / form / baguette - any of the 'artistic' aspects, could be in breach of copyright - but if the circuitry was laid out on a different board, and the physical form was different, and the code in the DSP was written to perform the same function but wasnt a copy of the code, then there would be no grounds to sue)

5.) I note your offense at "our proprietary technology" being exposed, but kind of wonder why you endorse the Wavefront theremin in the light of this righteous indignation - Gene Segal designed this theremin almost entirely from "ripping off" Bob Moog's EW design, and used his high-end EW Clone to directly compete with Big Briar (Bob's) theremins, much to Bob's helpless anger and frustration - I assume its ok in your books for Americans to rip each other off, just so long as the business stays in America, but god help anyone who dares to disclose "proprietary technology"(even technology which is crap and in the public domain anyway) to anyone else! LOL !

Rhe only people interested in "getting inside" the theremini are those who want to gain information that SHOULD be provided to potential customers PRIOR to them parting with their money, and those with an interest in theremin technology...

And here is where we have different perspectives. I do not believe that hiding technology serves any good purpose (IMO, the only technology which should be hidden is technology that should never have been developed - technology from the death industry - and this technology is not constrained by concience or borders, its available to those with enough petro-dollars, regardless of whether they're facist's or state-sponsored terrorist armies or whatever)  . I also think Richard Snowdon is a hero for his actions - sadly though, as all people who act on their conscience eventually discover, matters often get a lot worse before they get better, and often they dont get better as the result of a good action (certainly not in the short term, if ever).

So dont worry! Revealing the theremini "secrets" wont impact the coming WW3 one iota ! ;-) And wont even impact theremin development one iota - because theres absolutely nothing new or of use in it!

The only "impact" of any "revelations" made here will be to inform potential purchasers of what they might be buying, and to give technical information possibly  useful to persons who buy the theremini, and to increase technical awareness of issues related to theremins and particularly digital theremin implementations.. In all, I see only benefit to the theremin community from these revelations!

Fred.

"I really dont see that video as "evidence" either way.

Having said the above, I have no doubt that the BF862 oscillator is less noisy than the 4069! - What I cannot tell is whether the difference actually has any significant (or even measurable) effect on anything.. So far I have not seen anything that leads me to think it does! - Does the video ACTUALLY show anything related to oscillator noise? - I doubt it!"  - FredM

Yes.  I'm not saying I distrust livio, but I have reservations about that video as well because it isn't clear enough what is actually going on.  A CMOS oscillator can be made much more sensitive than the one in the UNO.

"That seems well intentioned until you start assisting that other guy who cracked open the box and began exposing our proprietary technology and now shares it across International boundaries, I am not talking about Richard Snowden."  - Touchless

IANAL, but AFAIK hardware peeping and yacking about it is completely legal in the US.  A good example: the IBM clone industry is based almost entirely on reverse engineering IBM hardware.  It gets murkier when you start disassembling software.  And if you want to keep proprietary secrets you forfeit legal protection of them via patenting.  Patents are supposedly win-win in that society grants a time-limited monopoly in exchange for the exposure of a trade secret (i.e. trade secrets are bad for society and should be discouraged).

"If our product that is being violated does not come out of the red in sales he will be paying some salaries around here."

Or maybe they could try, I don't know, making a better product?  Oh, don't listen to me, I'm hopelessly old-fashioned! ;-)

"Or maybe they could try, I don't know, making a better product?  Oh, don't listen to me, I'm hopelessly old-fashioned! ;-)" - Dewster

All sorts of issues come up on this matter - But as I see it, Moog had a huge advantage and head start - if they had produced a good (or good enough) theremin at the theremini's price, they would be at near zero risk of incurring loss - Disclosure or non disclosure would make no difference.

The only theremin that has ever made substantial profit for anyone is the Moog Etherwave Standard, and this theremin was fully disclosed by Bob Moog - Oh, some people (at least in Bob's opinion) "violated" his design and competed with Big Briar / Moog at the high end using Bob's EW design (Wavefront being an upgraded  EW in a big wooden cabinet, and which, from accounts, really got Bob angry!) but in the main, Moog benefited greatly from the full publication of the EW and the fact that as a result this instrument became the standard.. I am quite sure that Gene Segal would have reverse engineered the EW and used it in the design of the Wavefront  even if Bob had not published it - and there is nothing Bob could have done about it.

"Patents are supposedly win-win in that society grants a time-limited monopoly in exchange for the exposure of a trade secret (i.e. trade secrets are bad for society and should be discouraged)."

Sadly, patents are now a win-lose (and therefore, eventually, a lose-lose) - Only the big bullies and lawyers can now win, everyone else loses - and this does increase the tendency to non disclosure.. But these days, "trade secrets" (certainly in terms of electronics / circuits etc) are easily hacked.. So really, all that a small innovative company can do is to establish a reputation and win loyalty by producing great product and being seen to be active in advancing some niche market..

This is what Moog HAD! - Sure, Bob's designs (Which he freely published back in the old days) are the backbone of almost all analogue synth modules (and some theremin) designs available today.. But ask anyone in these niches whether they would buy a MFOS synthesizer or an Moog synthesizer if they could afford either, and they would buy a Moog. Moog was at the high end, and was only interested in this end, and only fought competitors (primarily ARP) at this end - But even when Moog (under different management, without Bob) dipped into the lower cost instrument market, they had the good sense to use Bob's designs and produce instruments (The Prodigy for example) that did not trash Moog's reputation.

Moog Inc is now, IMO, producing low cost rubbish which will trash their reputation.. Moog (while Bob was the leading designer / developer) had always been the major innovating company amongst a myriad of cottage businesses *(certainly for theremins) - the "reference" to which others aspired. Alas, this reference is gone.

*(there were bigger synth manufacturers like Roland, Korg and Yamaha who primarily targeted a slightly different market but nearly wiped out analogue synthesis and caused many years of hardship for us .. Bob spent some of these years as Big Briar who produced theremins not synths.. Most engineers in this area went off and got real jobs designing controllers for washing machines or other useful stuff like that - I returned to medical electronics ;-)

Fred.

Hi Dewster,

Ive been looking at offset heterodyning more closely, doing my own simulations and even playing with some oscillators.. I took your spreadsheet analysis at face value at the time (cannot find them now) - but I am having real doubts..

To me, it looks like offset heterodyning makes matters far worse at the far / bass end.. that the standard null tuning gives the best far field result.. Ok, I am not doing this numerically - I am using the difference frequency directly as audio - but AFAICS this should apply also if one is counting the period of heterodyned signals.

At the near field, there is improvement, and this improvement increases as the offset frequency increases, up to the point where the offset frequency becomes too large and causes non-linearity in terms of excessive spreading at the high end. (actual numbers depend entirely on sensitivity and oscillator frequency span - the larger this is, the 'better behaved' offset seems to be and the bigger the offset frequency one can add to linearize the near field)

The low end however looks to me like it gets severely crippled even with moderate offset.

I started looking at this to explore if it was possible to use offset with an analogue theremin - the idea was along the lines of using PLL's and dividers to shift say a 600Hz offset down to an audio difference frequency of say 16Hz, and switch the dividers to allow register switching so the offset became audio at 16,32,64,128Hz after heterodyning.

But everything I did, every sim and spreadsheet I ran, showed abysmal results in the far field / bass end.

So I am really puzzled - I dont see how it works, particularly with low voltage antennas!

Fred.

Actually, it just occurred to me - one doesnt really need spreadsheet or sims to realize the "problem" ... Take a conventional analogue theremin - the bass (far) end gets spread out, so changing the VFO frequency by say 40Hz to 80Hz requires a greater distance movement than 80Hz to 160Hz, and much greater distance is required from say 20Hz to 40Hz (assuming no linearization or oscillator pulling is involved)

Now if we add 120Hz offset (to the reference oscillator), for the same distance required to move (the difference frequency) from 20Hz to 160Hz (3 octaves) we would go from 140Hz to 280Hz (1 octave) and the higher the offset frequency, the less change (in terms of distance / octaves) we would get.

Am I going crazy? It all seemed to make sense before, but I cannot remember how / why it made sanse, and cannot now see any way it could make sense! ;-) (for the far field anyway)

As I see it, any offset added must reduce "effective sensitivity" in the far field.

-----------------

Brings me back to my earlier solution to linearity - drive the VFO into a BPF, get a voltage proportional to VFO frequency, and make the reference oscillator a VCO so that "offset" is added (or actually subtracted - VCO frequency drops as VFO frequency drops - but in a 'shaped ' way) as the the VFO frequency changes.