Possibly a "new" theremin topology, and links to related background

"Yeah! ;-) .. Must admit that I am deeply unimpressed by the layouts! - That PSU card is just shocking! - I can understand the thin tracks and lack of ground planes on the other boards, but the PSU should be chunkier even if it only supplies low current ... (also, minimal etching is more environmentally friendly).. That board could be half the size, packed with copper, and be forward to clear the pitch antenna so the extra copper's effect was reduced."

At least, they are double sided... ;-) 

I wasn't shocked when I saw these circuit boards for the first time, it was a rather familiar feeling because they resemble much the German Hifi and TV equipment of the late seventies (Dual, Telefunken, Grundig, etc.) 

"At least, they are double sided... ;-) " Thierry

Yes, you are right! - I am way too critical anyway.. comes from always designing stuff to fit in impossibly small enclosures with extremely critical EMC and safety requirements.. That side of my life as a designer is now over - SMD and those who can work with it made my PCB design "skills" utterly redundant! - My last Thru-hole PCB design for that world was fuel monitoring sensors fitted in Euro tunnel trains.. that was more than 15 years ago!  I simply cannot do a SMD PCB!

I spend WAY too much time perfecting even my prototype boards - Its stupid and crazy - My Skywave boards were rushed and a complete mess, but they worked fine - If I hadnt had the pressure, I would easily have spent as long laying those boards out as I spent doing the whole job!

Fred.

Recently (in a now deleted post) the following comment was made:

“It has been openly stated in the forums the Moog instrument you are now copying does not meet the definition of a theremin,”

I wish to correct the above.

1.) I am not copying any Moog (or other) theremin.

2.) Whilst there is a common factor between my “upside down” topology and the Moog 91 (a factor I was not aware of until this thread), the method of sound production is entirely different.

3.) The common factor is that both designs have their antennas operating at a fixed frequency.

4.) The Moog 91 derives a control voltage from antenna capacitance, and uses this to control a conventional analogue synthesiser ‘engine’ to produce the audio.

5.) With my design, I am simply swapping the ‘location’ of the fixed and variable oscillators – Rather than the reference oscillator being ‘external’ to the antenna and the pitch (variable) oscillator being connected to the antenna, my “reference” oscillator is connected to the antenna, and my pitch (variable) oscillator is “external” to the antenna.

6.) With my topology, sound is produced through heterodyning in EXACTLY the same way as was implemented by Lev, and as is used in the Moog EW, and every other “conventional” theremin.

7.) With my design (as with all conventional Theremins) the mixer is a “sub-circuit” for which one can choose whatever topology you want – I could put exactly the same mixer into my design as is used in the EW, or I could put a tube mixer in, or I could put a 4Q multiplier in, or I could use one of my own designs.. It is my belief that with heterodyning theremins, it is the mixer which is the primary mechanism by which the "tonal qualities" of the sound is determined, and I intend  this to be a rather special mixer giving the ability for far greater waveform adjustment and harmonic variation with pitch than any other to date.

8.) “does not meet the definition of a theremin,” is utter nonsense anyway – Who gives a s**t about some non-existent “definition” – all the musician wants is something with the sound they like.. The 91 is an instrument loved by many - so it doesnt heterodyne - SO WHAT ?

I have had my head up my backside over this matter for too long, and felt there may be some "heterodyning magic"  - The truth is that IT DOESNT MATTER if the sound a theremin produces is by subtractive synthesis or by FM or by digital wavetable synthesis or by heterodyning - IF the AUDIO waveforms are the same, they will sound the same - And it is possible to produce IDENTICAL waveforms with many different technologies!

[any doubt about this? - Well, if you record a theremin with a digital recorder, when you play that sound back, its what could / would be produced by a high resolution digital theremin! - Apart from hearing an analogue theremin played live, every theremin you have ever heard has a sound that COULD be produced digitally!]

It is, however (IMO) , EASIER to produce the same waveforms by using the same technology as was used in the "original" - but there is no "mystical  component" to the theremin sound - its all just synthesis of one kind or another.

IMO, its not the sound thats the major difficulty with other technologies, its the interface - Digital is now at a level where resolution and smoothness problems can probably be overcome - And it can probably deal with linearity correction and other matters where analogue was once the only route -

My topology is analogue and heterodyning - perhaps a dying breath for this "class" of theremin - an attempt to design a 21st century theremin that can compete against the new digital generation before they all go into extinction (which I think is quite likely)

Fred.

Finally, I intend to publish this design freely - Here if I am not hounded by jealous disruptive  idiots, otherwise elsewhere.

Ruslan,

I just posted a huge reply, and it dissapeared!

Damn annoying!

Will get back ASAP.

Hi Fred, I managed to open it. If you do not mind, I'll get it back...

Hi Ruslan,

For what I am hoping to do, the primary advantage is that one has all antennas on the theremin operating at a fixed, synchronous frequency - this allows one to add as many antennas (for other functions) as one wants, and also allows one to have as many theremins (of this type) operating together in close vicinity, without any interactions..

I think ideas about polyphony are a bit silly - but my main interest is in having a 5 antenna theremin - 4 of these antennas will be built into the volume loop, to provide X,Y and Z sensing of the volume hand position (to implement volume and formant control or other 3d control) and the 5th would be a conventional pitch antenna.

Aside from the above, by locking the antenna frequency, all sorts of other "ghost tone" type problems go away - I am extremely sensitive to tiny overtones produced from pitch - volume oscillator interactions, and have never managed to construct a theremin which had assynchronous pitch and volume oscillators that satisfied me - for this reason I used the reference oscillator with volume antenna resonator on all my theremins.. this, however, adds the same 'complexity' as my "upside down" topology - one cannot tune the reference oscillator, as this is shared by both pitch and volume - so tuning must be implemented at the antennas.

One other major "difficulty" with the conventional theremin topology, is that the sound produced is intimately linked to whatever one does at the front-end RF side.. If you modify the linearization (through oscillator coupling for example) this impacts on the sound (worst case being that oscillators lock too soon, meaning that one loses bass, or gets the EW "stutter" at the bass end).,

What one wants (IMO) is to be able to have independent control over linearity / sensitivity, and have independent control over the sound.. I think that one of the reasons for the limited sound "pallet" available to conventional theremins is somewhat to do with the difficulty of balancing the conflicting requirements.

With my proposed topology, all the RF stuff is "fixed" - one does not need to mess with it either to achieve linearity or to influence the tone.. This was also true with the 91 theremin - but with my topology, the sound can be produced through heterodyning, whereas with the 91 its not, and it cant be.

My topology will alow absolute control over linearity, and absolute control over sensitivity / span (number of octaves available in the 60cm playing field) - It will allow conventional heterodyning mixing and even controlled oscillator coupling which will not affect linearity, and it will allow easy register switching if an apropriate mixed-signal heterodyning "mixer" is employed.

In my head, I have a plan ;-) One conventional "voice" from a analogue mixer based on a clone design of the Lev mixer, and one mixed-signal voice which can be register switched reletive to the other - There is a CV available internally which I plan to use to enable altering the sound proportional to frequency (morphing waveforms) -

But whatever - what this topology does, I believe, is provides the best "front end" possible for a theremin - From this one gets a HF reference frequency, a variable HF frequency, and a CV for the pitch section, and can do what one wants with these without in any way impacting the linearity or span or whatever - think of the "front end" as a controller.. with output signals that can feed whatever "genuine" or "other" theremin circuits you choose - the essential fact being that these signals are the same (and more) as one has in any conventional theremin.

Fred.

 I should just say that "swapping" was a simplification - what I do is to derive the control for the variable pitch oscillator from the "error" or "correction" signal used to keep the variable oscillator connected to the antenna, in lock with the crystal reference oscillator -

My problem is that I dont really know what to call these ;-) .. The oscillator connected to the antenna is a variable oscillator - but it doesnt vary! ;-) .. It would vary if the control loop (PLL) was disconnected.. If we forget about linearity / span etc, and look at a simple configuration:

one has a variable oscillator (oscA) whos natural frequency varies as a function of antenna capacitance, and one has a fixed crystal reference oscillator (oscX) and one uses a PLL to lock oscA to oscX, from this PLL there will be an "error" voltage (cVe) used to adjust oscA's frequency and keep its frequency constant as changes occur to the antenna capacitance..

Now we have an identical oscillator (oscV) which is not connected to an antenna, but has a fixed capacitor there instead, and we drive this with the error voltage (cVe) derived from oscA's PLL.. (in fact, it would be this cVe inverted, but ignore that ;-)

If everything was perfect, oscV will run at the same frequency that oscA WOULD be running at IF oscA was not locked to oscX.

ok - if youve got that...

If we now insert a "modifier" between the PLL (cVe) output and oscV's control input, we can do all sorts of things.. We can amplify cVe (increase the span) we can attenuate it (reduce the span) we can process it with a log or exponential or squaring circuit -, or mix these functions to taste (effectively create a modifiable tranfer curve through whatever means we want - I am probably going to implement this by morphing several ramps - linear and exponential for example - and using these in a frequency-control PWM - my maths weakness causes me some problems in this area, but the functions are reasonabley simple), and therebye change how oscV responds to cVe, and therebye have complete control over linearity...

 My major problem is that I cannot build anything at the moment - I really need a competent partner somehere who can build stuff and work with me on projects.

Hi Ruslan - did a long detailed posting but it just vanished .. real annoying to lose my last posting.. oh well -

 "what is the practical advantage of swapping the location of reference and variable oscillators?"

"Swapping" is an oversimplification.

Advantages: As many antennas as one wants, as many theremins close together as one wants, no interactions..

Complete independence between RF front end and audio production - one does not impact on the other.

Full user inearity adjustment, and span (number of octaves in 60cm) adjustment.

"Does your proposed approach allow to make the schematic more stable to changes in temperature and humidity?"

The topology probably has some advantages in this respect, but its really only good design and the right components that determine this. *having said that, because the variable oscillator is voltage controlled, it is MUCH easier to monitor temperature and apply a correction voltage to mitigate against any thermal drift

"Does this topology allow to measure the capacity of a person, get it in numerical form, make computer calculations and make adjustments to the frequency of the variable oscillator to control the range and linearity? "

Not in "numerical form" - its entirely analogue

" Does it give any more control over the timbre, compared with the traditional approach?"

Infinitely more!

I will give a deeper discussion in the next few days - my "thesis" reply I posted in answer to your questions just vaporised, and I am so disheartened by this that I must now go and take some anti-depressants! (only joking ;-)

Fred.

RUSLAN!  ;-)

You got it back !!!!

I had the gun to my head, but thankfully thought id have one last look at TW!

I am extremely grateful, thank you!!

;-)

Yeah,

I love Windows XP (compared to all the later versions) which locks me to Explorer 8 or other browsers I dont like.. I have loads of essential SW which wont run on later Windows..  TW messaging doesnt seem to like E8.. Occasionaly I will use Chrome to deal with the TW inbox.

Then I have a BIG problem with spam - I believe that someone is deliberately signing me up to spam lists, I get at least 200 spams a day, and 80% of them are utter filth.. Its not just normal spam - someone is signing me up to porn sites and the like.

So its time for me to sort it all out - get a PC exclusively for on-line stuff, and change my email address - But what with all the other stuff going on, I just havent got to this yet..

As for Facebook - I cannot use it - mail / notifications just get directed into a folder I never look at anf just delete every so often without even checking..

Sorry - I dont know why SpamFighter is blocking you - I will go into it and see if I can sort that problem out.... Last couple of emails I had from you crashed Outlook - Litterally, the PC froze when trying to invoke Word, and I had to do a system restore to an earlier date because Outlook wouldnt open after that -

Fred.

 "Sounds amazing, but extremely complicated." - Its not that complicated (fabulously simple compared to my prior designs ;-) - In truth, my primary focus isnt the theremin - its other controllers using theremin concepts - One of the reasons this came about was because in designing these other instruments I saw the topologies potential for a conventional theremin control unit.... The other instruments are where I believe any money might be... Which is probably why I am happy to give the theremin ideas away! ;-)

Oh, just as an aside - Another advantage of fixed frequency is that one can implement a SMPS (switch mode power supply) locked to the master frequency - this way the whole instrument can be powered from a low voltage (8V to get a good regulated 5V) which powers the oscillator, and this drives the SMPS to give +/- 15V internally..

Fred said: "6.) With my topology, sound is produced through heterodyning in EXACTLY the same way as was implemented by Lev, and as is used in the Moog EW, and every other "conventional" theremin."

Fred I think the complete description of your design is excellent and this one statement above for me is what captures the Theremin Spirit, you have a technical hybrid. It is when this fundamental is missing that I say call it something else, I have nothing against digital itself when used in the properly named applications.

Rusian your use of battery power is an excellent idea as I have found it to be the cleanest form of powering the extremely sensitive theremin oscillator mixer section of a conventional solid state design. In one of my theremins this osc section is separated from the rest of the circuitry as there is only a 2 ma current draw and so the batteries last for years. Also this eliminates added heat sources, at 300 Hz my design can remain on the same musical note over a 10 degree room temperature change, that is all day at my home. This is accomplished through the use of minimal components and designed with circuit balance.

Interesting Coincidence?  I have often stated that 922 kHz is what I found to be the ideal theremin pitch frequency for linearity. It is interesting from this thread the Moog 91 crystal freq is 922 x 2 ?

That is also Michael Faraday's birthday. 9.22

Christopher

Fred wrote:

Oh, just as an aside - Another advantage of fixed frequency is that one can implement a SMPS (switch mode power supply) locked to the master frequency - this way the whole instrument can be powered from a low voltage (8V to get a good regulated 5V) which powers the oscillator, and this drives the SMPS to give +/- 15V internally..

Philips did the same in their color TVs in the late eighties: They combined the SMPS with the line and HV stage to economize one transformer. Everything ran on the 15625Hz line frequency. That was the first generation of TVs where either everything would work or nothing at all because a failure in whatever circuit section would shut down the line oscillator, disabling the complete power supply. The latter wouldn't stop trying to restart which gave the typical tic-tic-tic-tic-tic... even if only the audio ic was defective.