Super Simple Design Idea

Hi Thierry..

Yeah - Even I am (despite having worked with them before) getting too old to keep up with FPGA's etc.

What puzzles me is why people are so 'afraid' of coils! - One simple adjustable inductor which can be bought for about $1 from Mouser (use one winding of the 42IF106 IF transformer), allows a simple LC oscillator to be constructed.. This simple oscillator will kick the s**t out of any RC oscillator costing twice the price!

Still need to build my compensated 555 osc.. This may be something which could interest you, Jo.. If one keeps the supply voltage stable, and use NPO capacitors, the TS555C is quite stable over +/- 10C..

http://www.thereminworld.com/Forums/T/28556/new-theremin-circuits

"If you want to build something without coils, follow the digital ideas which FredM and dewster are developing here. " - Thierry

Actually, even "digital" theremins often have an analogue front end (AFE) which incorporates coils..

Lots of good reasons for using inductors - one which sometimes gets missed is that with RC oscillators, the voltage of the waveform at the antenna is limited to the supply voltage - and any local RF signal source has a strong chance of messing up this signal - Using an inductor one can have voltages on the antenna anywhere up to several hundred volts - a lot less susceptable to influence from local low voltage interfering signals! 

I should also say that this is Dewster's project exclusively - I am quite out of my depth, and am not persuing digital theremin design myself - Analogue and mixed signal are the theremins in my 'comfort zone' - LOL.

Hey now, no one is too old to take up digital design - or analog design for that matter!  I came upon pure digital a little over a decade ago and now am in semi-retirement (@ 53, yikes!).  A determined engineer can design in any medium he/she sets their mind to.  After having done a fair amount of analog design (mostly in the lay hobby realm, but also professionally) I find digital design to be incredibly fruitful and rewarding.  I feel as if I'm actually building something when I'm coding / simulating in verilog (because I am).  FPGAs are a gift from the gods.

"Hey now, no one is too old to take up digital design"- Dewster..

Oh, I spent many years doing almost purely digital designs.. perhaps its more about real mastery - there is just (IMO) too much data to assimulate for one to be a master in both analogue and digital.. just keeping up with the latest developments in FPGA's is a full time job!

Analogue, on the other hand, is much more stable - radical new devices do not pop up daily, and one has a set of building blocks which improve, but rarely do radical new ones appear - and parts do not tend to dissapear as rapidly either.

BUT - for me, the component packaging is also a major factor - Even with analogue, I am forced to use a few SMD components - I can barely see the bloody things! I need to actually prototype and make changes to the P/T boards - simulation only gets me a circuit giving the kind of waveforms etc that I want - fine 'tuning' is down to actually listening and tweeking the circuit, then running the modified circuit through the simulator again to check for tolerances etc..

And SMD's make prototyping a hell of a lot more difficult and expensive - I cannot place or solder these to the board myself, so must design (or buy)  adaptor boards and get the SMD components mounted on these boards.

So I have as few SMDs as I can get away with.. This rules out many MCU's FPGAs and other components.

Fred.

It's not that I'm afraid of coils or too old for digital (heck the only electrical engineering course I ever took was in digital circuit design).  I'm doing this purely for the satisfaction of having designed one that actually did something, using mostly parts I already happen to have lying around.  I do plan to build a couple of coil theremins when the time comes, hopefully even eventually wind my own coils. This is just a step in that direction building upon parts and circuits I'm already comfortable with.  If I an get a tone out of it and control the pitch and volume at all I'll be happy with it.  If I can play Summertime on it I'll be ecstatic, but given the feedback I've gotten so far, I'm not expecting that.

"I'm doing this purely for the satisfaction of having designed one that actually did something, using mostly parts I already happen to have lying around"

ok - what parts do you have lying around? ;-)

" hopefully even eventually wind my own coils "

IMO, there is little / no advantage in this - other than that they can look impressive!

Fred.

Fred said: "IMO, there is little / no advantage in this - other than that they can look impressive!"

You are absolutely correct Fred, I always assume big coil winders had Tesla Envy. My research has proven to me I want High Q in theremin LC circuits not high coil forms. My theremin research began with coil winding until I discovered the 796 kHz IF Transformer.  Coil winding is more of a religion, so I do understand it!

Edit: I am open to know what big coil winders experience in using them besides their beauty.

ok - what parts do you have lying around? ;-)

I've got all the transistors I'd need for the oscillators and amp section, a box full of assorted resistors and capacitors, the pots my design calls for, pretty much everything but the antennas, speaker, enclosure and knobs and maybe a diode or nine depending on how some things work or don't work when I get it all put together.

I build effects modules and other little gadgets for guitars all the time, so I keep a running supply of typical discrete components I'm always needing.

" Coil winding is more of a religion, so I do understand it! " - RST

LOL! - Chris! You mentioned the "R" word first, this time! ;-)

When it comes to equalizing coils, high Q and low capacitance are, IMO, the fundamental requirements.. These can be achieved with big air coils, but equally can be achieved with smaller series connected coils on ferrite formers.

For those who dont understand this, I will explain:

It is the 'trick' of connecting coils in series which is where one gets the best of all worlds - the capacitances of these coils is reduced for the total capacitance of the final composite inductor - For example, the Miller (Bourns) 6304-RC coils (which I mainly use) are 5mH 5pF each.. Connect 8 of these in series and one has 40mH (series inductances add together), but the capacitance becomes 1/((1/C)*N) = 1/((1/5pF)*8) = 1/(0.2*8) = 1/1.6 = 0.625pF ... You will have difficulty winding a coil which gives 40mH with capacitance of 0.625pF!

Using 10mH coils, each with 8pF, one gets (for 4 coils) 40mH with 2pF - this is ok, not as good as using 8 5mH coils, but ok (and 1/2 the price of using 8 inductors).

Using one 40mH coil on a single ferrite former, one gets a lousy equalising coil, because capacitance is usually much higher. 

Fred. 

(in fact, the Miller / Bourns 630x inductors partly achieve their low capacitance by being wound as 3 seperate coils on a common former - using the same 'trick' I have described above)

"Fred said: "IMO, there is little / no advantage in this - other than that they can look impressive!" "

I was wrong about this. The advantage of large air coils is that one can achieve high inductance (required for equalizing antennas, particularly when operating a lower frequencies) without suffering the severe variation in inductance caused by thermal behaviour of ferrite.

At the time I wrote the above I had not done much experimenting with well constructed air coils, and good quality ferrites (Bourns 6300 series for example) seemed to have comparable performance. Since playing with the Lev oscillator topology, I have wound several coils and transformers, both air type and ferrite, and constructed a better 'chamber' (modified Peltier fridge ;-) allowing me to better sweep the temperature of test circuits and evaluate the performance.

Good air coils are certainly superior to the Bourns 6300 series inductors when using NPO / GOC capacitors for the dummy 'antenna'. I have not tried the Hammond inductors. The difference in drift is probably not significant enough to warrant air coils in most applications - but if no other means of thermal correction is applied (choice of N or P TC capacitors in the tuning circuit for example) then there probably is more reason for air coils than just looks, particularly on high-end and tube theremins.

tomsoong brought this thread up and I cannot leave my error exposed without correcting it.. Lev used big air coils - natural due to the era.. But Bob Moog also used them on his best instrument, The EtherVox.

I suspect that air coils may be even more important with theremins that use tubes / valves, as perhaps temperature variations in these instruments is more extreme.. (?)

Fred.

"Good air coils are certainly superior to the Bourns 6300 series inductors when using NPO / GOC capacitors for the dummy 'antenna'. I have not tried the Hammond inductors. The difference in drift is probably not significant enough to warrant air coils in most applications - but if no other means of thermal correction is applied (choice of N or P TC capacitors in the tuning circuit for example) then there probably is more reason for air coils than just looks, particularly on high-end and tube theremins."  - FredM

That air core inductors are measurably more thermally stable than ferrite core is something I would expect, but I haven't gone to the trouble of measuring this as you have Fred.  Could you give some numbers here (delta Hz/C, ppm/C, etc.) for both types?  I'm quite curious.

I imagine thermal variation of an air core is most likely due to the expansion / contraction of the copper wire, which the coil former might help minimize.  Hence the use of ceramic formers with the wire tightly wound in ham VFOs.