Let's Design and Build a (mostly) Digital Theremin!

Posted: 3/9/2014 8:14:51 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

livio, I think you are smart to use closed inductors, and the one you picked is quite nice compared to most.  Have you ever tried putting a second one between the antenna and the tank?

I'm not married to CMOS or anything, if there is a better driver to use I'll definitely use it.  I'm going to play around with Spice some more.

Posted: 3/9/2014 11:54:36 PM
FredM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as WaveCrafter.com . ...................................

Joined: 12/7/2007

"Fred, on top of a general sensitivity increase, another reason to use L2 in just about any Theremin oscillator is to have another low pass component between the oscillator and the environment." - Dewster

To be honest, ive never added any series antenna L for this purpose - its an interesting idea.. My focus with using antenna L's has been for linearization and increasing antenna voltages - Low pass wasnt something that I ever felt was strongly needed (was more interested in high passing the antenna to get rid of LF problems)

But actually I can see a lot of merit in your topology - Probably the ideal would be a HPF (antenna series capacitor or perhaps something a little more complex) followed by a LPF (series inductor) to give a bandpass.. As I see it, this is all in place with your oscillator design.

I have played with ceramic resonators in series with the antenna and had some interesting results - Unfortunately, that was back in quite early days, when I didnt understand the series / parallel topology, so what I was trying to achieve never worked - also I never understood ceramic resonators well - Now that I think I perhaps understand both a bit, its been something I wanted to play with again - In particular, using a few ceramic resonators with Ls and Cs to produce a BPF at the antenna.. I now dont think im likely to bother though - my focus has changed.

Oh, BTW - I agree with your comment about adding a littlefuse or similar ESD protection to the input of whatever active part is used for your oscillator - the 70 or 90 Volts clamped by a discharge tube could still do damage. I tend to use low capacitance schottky diodes to the rails as my standard protection for device I/O, but anything that clamps voltages to safe levels is fine, and the littlefuse parts fit the bill.

Fred.

"- Tried a single inverter 74HC04, 4069UB, both work at 3.3V with the Bourns coil."

Thats interesting! - My simulation wouldnt run with a single inverter - so it seems the real parts have more gain.... Oh - Actually, if you were using buffered (B) parts this would probably make sense - my model is unbuffered (HCU).. I used this as I figured they would more accurately approach your ideal opamp's behaviour - simply replaced the opamp with 3 unbuffered inverters, then added two more and the circuit waveforms were almost identical.

Posted: 3/10/2014 12:54:55 AM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"Probably the ideal would be a HPF (antenna series capacitor or perhaps something a little more complex) followed by a LPF (series inductor) to give a bandpass.. As I see it, this is all in place with your oscillator design."  - FredM

I can't tell if I see more or less LF instability now when I use a capacitor in series with the antenna.  I think I see less when I use a tiny (~10pF) capacitor, but of course the entire signal is smaller.  Everything kind of goes to hell when your scope is triggered way out and zoomed in on a rat's nest.

"Yes it starts back after any stall and it never stalls, also with many square meters antennas. Try it and you will see."  - livio

That's good.  A non-recoverable stall in a Theremin would be bad.

"Also in the 2 to 3 MHz band there are some nautical stations but with a power incoparably lower. Maybe some 100 or 1000 Watt not 500 MegaWatt."

livio, I'm curious.  You give good advice when you say stay out of the AM broadcast band - have you have bad experience with this?  Not saying your oscillator is bad, just wondering if AM broadcasts have given you trouble that you can discuss here?  I mean, is your advice general good practice, or is it from personal experience?

Posted: 3/10/2014 2:41:52 AM
Chobbs

From: Brooklyn,NY

Joined: 12/1/2009

Hey Dewster -if youre free over the next couple of weekends- you should come by the NY theremin Society Residency Events at Pioneer Works in Red Hook, Brooklyn.   You're so close and it be great to meet- maybe bring along some of your 'projects' for a little show and tell. Love to see what you have been up to.

Posted: 3/10/2014 3:57:07 AM
FredM

From: Eastleigh, Hampshire, U.K. ................................... Fred Mundell. ................................... Electronics Engineer. (Primarily Analogue) .. CV Synths 1974-1980 .. Theremin developer 2007 to present .. soon to be Developing / Trading as WaveCrafter.com . ...................................

Joined: 12/7/2007

"I can't tell if I see more or less LF instability now when I use a capacitor in series with the antenna.  I think I see less when I use a tiny (~10pF) capacitor, but of course the entire signal is smaller.  Everything kind of goes to hell when your scope is triggered way out and zoomed in on a rat's nest." - Dewster

LOL ;-) Tell me about it!  I worked for years in a tiny cramped lab with wires and equiment growding available space, and sneeking to use the kitchen table for critical tests.. These instruments are a bugger to work on!

I introduced the small series antenna capacitor right back in my earliest days when I was playing with my first (digital) theremin ideas - because I could SEE the 50Hz and other LF crap with my logic analyser - It was big and ugly, but most of it vanished when I put the cap in, (I then noticed HF and "digital" noise from the display etc) - but  all of it vanished when I had sensible antenna voltages.

As to whether its REALLY needed, I couldnt say... But if one has say 500kHz top VFO frequency, having a roll-off at say 450kHz and 550kHz could be nice..

With analogue heterodyning theremins, sensitivity isnt a real big issue (unless your oscillator frequency is real low) - One can match the Ls and Cs etc to get the span you want - One doesnt want huge sensitivity, so a series antenna capacitor is no big deal.. So im in the habit of including it.

With direct digital data aquisition (not using heterodyning) one starts out with your hands tied behind your back, IMO - You need the greatest PERCENTAGE change in oscillator period that you can manage - Frequency change in kHz (or even MHz LOL ;-) is irrelevant, its purely the % which is relevant.. So you are forced to increase the sensitivity of the antenna to as high as possible if you want to maximise resolution and minimise latency -

With heterodyning, the losses in sensitivity caused by a series C are a lot less of a problem.

I will never forget the panic and horror I had when my Epsilon controller misbehaved so badly - Everything had worked "fine" as seperate modules - but when it all came together it was complete rubbish - I was saved because the event where it was going to be premiered was cancelled! - But since those days the series C and me will not part easily! ;-)

Fred.

Posted: 3/10/2014 7:57:51 AM
livio

Joined: 2/2/2014

"You can hear the effects of the noise in this video..." - Livio

"Did you keep the tank capacitors about the same?  By this I mean: is the sensitivity and antenna voltage swing very low like it is on the Open.Theremin oscillator?" - Dewster

The frequency shifts you hear in the video are produced by the 4069 noise that changes continuously the input voltage of the input P and N  mosfets. So the frequency produced is not stable.

Then when working at long distances our linearizing software amplifies by a very large rapport the frequency changes and the little instabilities can be heared as in the video.

The antenna is not important in this, removing the antenna the noise is always present, because generated by the oscillator components. 

I have seen your original oscillator in your video and it appears perfectly stable at 1 meter with a frequency generated of about 1 Hz, But with 1Hz it is impossible to hear little frequency changes. Can you please test it making the zero beat at 1.5 meters and then, nearing the hand to 1 meter, produce a note of about 100 or 200 Hz. With this note you should hear the noise effects.

"livio, I'm curious.  You give good advice when you say stay out of the AM broadcast band - have you have bad experience with this?  Not saying your oscillator is bad, just wondering if AM broadcasts have given you trouble that you can discuss here?  I mean, is your advice general good practice, or is it from personal experience? - Dewster"

The 4069 (unbuffered) oscillators used in some installation like the first "Tiktaaliik" (1998) was working very good in my little town (Bollengo). But, when going at the museum in Torino we had many problems. The skeleton was sometimes moving by itself. (Torino is a big town and there are many RAI medium waves transmitters. RAI is Italian Radio and Television organization)

Our old 4069 oscillators frequency was about 900 KHz. In the middle of the Medium Wave band. The problem was completely corrected changing the oscillator frequency (if I remember well the frequency was lowered by about 100 KHz, or maybe a little more...)

In the 2005 we changed all the Tiktaalik electronics with a early version of the Theremino System with CapSensors modules working at 2.5 MHz and now the AM broadcasting problems are completely solved. And also the skeleton movements are more fluid and stable.   

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

Your research on best circuit configuration is really important. I will help with new simulations and ideas.

The DewsterV2 is not convincing me, because of:

- too much components (1)
- components (like Operationals and Dual in Line or SOIC inverters) that impose long and mazy PCB tracks (2)
- too much inductors (3)
- lack of the input isolating capacitor (4)

(1) My experience says that in the front-end circuits all must be minimal. With few and very short connections.

(2)  Using operational amplifiers or CMOS inverters, the PCB tracks become mazy and long. With many and long connections (more than some mm) the circuit becomes a radio receiver, can be unstable and can even stops to oscillate because of the stray capacitances.

(3) I could be wrong, but seems to me that the phase shift produced by the second inductor, could be best obtained with capacitors and active inverting. Yes the secon inductor produces a low pass effect but the same integration can be done in software (or firmware) and if done in software we have more control on it. Without the second inductor the number of hardware parts is reduced, the lenght of the PCB tracks is reduced, the noise is reduced etc.. All the functions that can be done in software must not be done here.

(4) The input isolating capacitor is important, maybe you could raise it to about 33 pF (if inductor is 330 uH) but not more or the 50Hz becomes to be a problem in some difficult ambient.

Posted: 3/10/2014 11:53:08 AM
livio

Joined: 2/2/2014

Here you can download a simplified version of your circuit with the following advantages:

- component count reduced

- antenna voltage increased from 50 Volt p.p to 140 Volt p.p.
  (i do not prefer to use high voltages but in this simulation you can see that increase the voltage is easy)

- sensitivity more than doubled

- only one inductor

- resistance to stall similar to your DewsterV2 (all those circuits stop to oscillate with a 10k in parallel to the antenna and all - including our CapSensors - can oscillate with more than 50 k)

The stall is not a problem. All those circuits are auto-restarting from any stall.

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

I do not know yet what the active element you want to use, I've heard of CMOS inverters but here I see an OpAmp. In all cases it would be better to use the components provided by our libraries. It took me years to collect them and make them available for everyone. 

http://www.theremino.com/en/downloads/uncategorized/#ltspice

After downloading our libraries, should use the OpAmp5 (from the "OpAmps" folder), and then right-click and choose an appropriate "Spice Model", a Level2 is too simple to properly simulate these circuits.

Or if you prefer to use a CMOS inverter you should select a 74HCU04 from the folder "74HC UB and CD". In this folder you will find the best CMOS inverter simulations (the more similar to the reality) in some slightly different versions.

 

Posted: 3/10/2014 12:05:53 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"Hey Dewster -if youre free over the next couple of weekends- you should come by the NY theremin Society Residency Events at Pioneer Works in Red Hook, Brooklyn.   You're so close and it be great to meet- maybe bring along some of your 'projects' for a little show and tell. Love to see what you have been up to."  - Chobbs

Interesting!  I had no idea this was going on.  It would be great to meet you & Dorit Chrysler & Rob Schwimmer - and the Theremini?  I'll definitely keep it in mind!

Wish I had a working prototype to bring but I don't, just random minor circuits laying on my bench at the moment.

Posted: 3/10/2014 12:09:33 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"I worked for years in a tiny cramped lab with wires and equiment growding available space, and sneeking to use the kitchen table for critical tests.. These instruments are a bugger to work on!"  - FredM

It must be hell to be married to a Theremin researcher!

Excellent summary of your experiences Fred!  I very much appreciate it, and 100% agree where it overlaps with mine.

Posted: 3/10/2014 12:18:12 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"I have seen your original oscillator in your video and it appears perfectly stable at 1 meter with a frequency generated of about 1 Hz, But with 1Hz it is impossible to hear little frequency changes. Can you please test it making the zero beat at 1.5 meters and then, nearing the hand to 1 meter, produce a note of about 100 or 200 Hz. With this note you should hear the noise effects."  - livio

I know what you are talking about.  With the configuration in my video there is significant noise when my hand is far from the antenna, but it is difficult to hear at such a low frequency.  The noise is exposed by subtracting a large constant number from a large changing number that has a small amount of noise.  I'm not sure if you would be able to hear it or not given the scenario you describe, but it might be on the edge.  Unfortunately I don't have the setup working at the moment, as my bench is occupied with oscillator stuff.

"The skeleton was sometimes moving by itself."

I hate when that happens!  ;-)

Thank you very much for describing the interference issues you've encountered!

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