Theremin Circuits Scratchpad

I had an email recently from someone wanting to understand coils / transformers, particularly with regard to how the Lev oscillator worked. I prepared the following for my reply, but feel it may also be useful to others, so am posting it:

A useful Online calculator related to the following can be found at Pace Magnetics

The above shows the PHASE of signals in/out - in relation to windings and phase dots on schematics.

I found this crude test oscillator I built years ago to test SRF of inductors - results are only "ball park" .. dont use this if accurate measurement is required.

This is based on the EM theremin design by Bob Moog - It is a wonderfully reliable oscillator, and clipping an inductor between +Vout and TST one will usually get oscillations at aproximately the SRF of the inductor.. I think it worked at least down to about 30kHz and up to about 2MHz if I remember correctly

I put the transistors in sockets, as it is possible to blow them up if one connects something stupid - I also put sockets to allow a capacitor to be connected across the inductor.

A simple, basic but useful piece of kit for theremin developers... May also be useful for those laying out EM theremin circuits (plenty of space near the test connectors to put a tunable inductor, fixed inductor and capacitor, without any top-side links - just move the ground track).

Built on 0.1" pad board.

Pads G1 and G2 are ground, VP1 is STABLE +12V, VN1 is STABLE -12V, Out is to 'scope / frequency counter, VP2 is +V supply to inductor, TST is to other side of inductor. Note that the TST lead is an "antenna" - it needs to be short.. The OUT signal will be low level unless into a high-Z input - the 100k is in series to reduce the effect of the measuring instrument, but this isnt ideal.

Layouts viewed from component side (top).

Some ideas on Buffering

This is posted at Element-14 Theremin group

I will not be visiting or contributing to TW much in 2013, and am unlikely to be replying to any threads.

Please feel free to contact me by email.

My email address is shown on my avatar.

it is:

fred

at

fundes

dot

co

dot

uk

"The OUT signal will be low level unless into a high-Z input - the 100k is in series to reduce the effect of the measuring instrument, but this isnt ideal."  - FredM

Could you maybe put a resistor in the collector leg of Q2 and use that to drive a buffer?

Hey folks. I just found and posted about a cool little phone & Mac app for simulating circuits on the go.  Check the front page for details.

Thats an interesting app, Jason!

I have been thinking about getting a cheap android tablet, and looking at what electronics related apps I can find - mainly need scientific calculator, spreadsheet and that sort of stuff - but a schematic entry tool is another thing I would like.

I dont think simulation would be any use to me - too slow to be practical (my dual xeon struggles with the stuff I throw it) but I can see it would be extremely educational.

No Jfets by the look of it - so even my simple Lev oscillator, mixer and buffer circuits couldnt be run.

Truth is that simulation, like playing the theremin well, is not easy - Yeah, we can all make "woo-woos" - But it takes a bit more to be a PP or LK.. And anyone can run a simple simulator with limited parameters and components - but for anything except amusement and elementary education, it will be useless.

Fred.

"I found this crude test oscillator I built years ago to test SRF of inductors - results are only "ball park" .. dont use this if accurate measurement is required."  - FredM

That's a neat and useful test circuit idea Fred, thanks very much for sharing!

I'd like to get a better handle on the inductors and capacitors I use in my digital Theremin AFE, and so have been shopping for LCR meters.  Seems there is a fairly new IC set that makes them fairly trivial to construct (ES51919/ES51920 from Cyrustek - the datasheet is an education in complex impedance measurement) so ~$200 USD and some careful shopping can get one the Mastech MS5308, UNI-T UT312, etc.  I'm tempted to get one of these meters, but they don't do things like tell you what the SRF of a coil is, how the capacitance changes with applied DC, how L or C change with frequency (the test frequencies are set at powers of 10 with 100kHz the top frequency) etc. so I'm sure one of these would be useful, but maybe not $200 worth of useful to me.  Any suggestions as to what to get?

My little Peak "Atlas" LCR meter (LCR40) has served me extremely well - its only downside being automatic frequency selection, which means that with large inductances it switches to 15kHZ (rather than the 200kHz for smaller inductances) - But this is typical of most lower cost instruments.

To get what you are looking for one really needs to do things the hard way - As in, swept sine oscillator and 'scope.. There are some amazing Digital frequency synthesis oscillators available now, with high sine purity and wide frequency coverage - Many Chinese offerings for example this: DDS Digital Signal Generator with Sweep Function * 5Mhz Range *   And use of this with a 'scope or rectifier + DMM, and a calculator, will probably give all the data you need..

I am a bit spoiled in this respect, having a basic ATE to play with.. Although I tend to use my LCR40 for gross inductance and capacitance measurements, and my HP3325A sweep generator and 'scope to get the SRF..

I also have a low cost Chinese capacitance meter (Newcason XC6013L) which is really good for the price (<£30) - with some adaption (modifying it to take a 'scope lead, and to compensate for the leads capacitance) it can actually measure down to 1pF (well, actually reads down to 0.1pF, but I round this to the nearest pF), and is usable for ball-park measurement of antenna and background capacitance.

Fred.

Thanks for your reply Fred!  You're helping me focus my thoughts.

The LCR40 looks really nice, but for $127 + $13 shipping = $140 USD (plus duty?) I could toss in a few more tens of bucks and get the Mastech MS5308 (which I may actually do).  Now that I think about it I really just need a quick reading on the value, secondary resistive characteristics, and perhaps how it behaves at a frequency or two, and any in-depth analysis done as you say with signal source and scope.

I use DDS in the digital Theremin as the DPLL NCO that drives the tank as well as the NCO that converts the pitch number into audio.  Ironically, one of the main problems with DDS is that it always minimizes output phase error (e.g. the output square rise/fall event happens on the system clock edge closest to the ideal location).  Therefore the output phase noise isn't white, with attendant phase jitter frequency "spurs" rising up from the noise floor, destroying SNR.  In my digital Theremin this behavior creates "sticky" points in the response, though this is simply and effectively dealt with via noise injection downstream of the phase accumulator.  (One of the best overview papers I've found on DDS: http://lionel.cordesses.free.fr/gpages/DDS1.pdf and http://lionel.cordesses.free.fr/gpages/DDS2.pdf).

One way to do precision DDS is to dither the square output to break up the spurs, and feed that to an analog PLL (used as a low pass phase noise filter).  Another is to feed the DDS accumulator ramp to a sine lookup table, run the output to a high speed DAC, then low pass filter the result.

There is an interesting discusion on this EEV forum about the Mastech MS5308, you may find this useful.

Edit >> Quite a few pages of discussion about LCR meters..

And what looks like a really good price on the 5308: Here