Super Simple Design Idea

I'm by no means an engineer, but I've been wanting to design a simple coil-free theremin with pocket-sized circuitry and most or all of the on-board components be transistors, capacitors, diodes,and resistors (no coils, no opamps if possible, no thermistors, etc).

Is there any reason a design based around twin-t oscillators and a simple transistor amplifier wouldn't work?  I have built low-frequency twin-t oscillators for guitar effects and I really appreciate their simplicity and how clean the waveforms are.

If this idea is workable, it would be a fun design project within my comfort zone I'd like to try.  The ultimate plan is a pitch and volume pocket theremin, with one of the antennas in a self-supporting base on a flexible ~18" wire, so the whole assembly is shaped like a nunchuck, be powered by a 9-volt battery and have an internal amp and speaker.  You'd be able to take it out, stretch it out on a table and play it anywhere.

If you are not playing a joke, then I learned this expression early on in my research.

 “The theremin will dictate her needs as they must be tailored to; which may not necessarily be yours.”

Not being an engineer will give you a creative advantage as you will try things others would not.  You are off to a good start as you will have a nice tone. The question is how are you going to sweep this tone from 1 Hz to 3 kHz?    Build the volume control first and you can have at the least a mono-tone theremin. (-'

nunchuck? This is going to be a real nut buster, are you 12 yrs old?

Too many just talk theremin, so go for it!  You will have success if the theremin invited you.

You are off to a good start as you will have a nice tone.

This is encouraging.

The question is how are you going to sweep this tone from 1 Hz to 3 kHz?    Build the volume control first and you can have a mono-tone theremin. (-'

Correct me if I'm wrong, but I don't think I'd need to sweep the variable osc that far to get a nice range of beat frequencies against the fixed osc.

nunchuck? This is going to be a real nut buster, what are you 12 yrs old?

Just seems like an elegant solution to having a full-sized playing field out of a device that fits in your pocket. Replacing the 'arm' on the pitch antenna of Moog Etherwave Pro with a flexible cable.

I'd say go for it too, but you will likely return to the same well-worn path of coils.  LC oscillators are somewhere between the stability and pullability of RC networks and ceramic/crystal resonators.  RCs give you plenty of pullabilty, but stability is poor.  Resonators will give you stability in spades, but aren't pullable enough.  The only middle ground that I'm aware of is LC.  Stability is essential because you are magnifying frequency drift via heterodyning.

Then if you stay in the analog domain you will find yourself play the crying games of linearity, note/hand spacing, and tone.  This is why I chose the mostly digital route, where almost any problem can be divided and conquered (I know, famous last words, but so far so good).

Please keep us posted on your progress!

Just a thought: you might want to try one of those integrated transistor arrays (e.g. LM3046) for the oscillators, that way they'd all be at the same substrate temperature and might "drift together" to some degree.  Personally I'd probably go the op-amp route, they make a lot of nice low voltage, low power ones these days.

RCs give you plenty of pullabilty, but stability is poor.

I really appreciate this kind of advice.  When I get the whole thing working and it does start having stability problems, I guess I'll have to determine how bad they are and if there's any compact and simple way to try to control for them.

...you will find yourself play the crying games of linearity, note/hand spacing, and tone. 

Yeah this will be the part that hurts the most, but I guess there has to be some trade off for the approach I'm taking.  We'll have to see if there's any way to deal with this issue when I get to it as well.

Please keep us posted on your progress!

Will do. Since my first post I've drawn up a block diagram and a rough preliminary schematic with no component values.  Next step is to do the tons of algebra and trial-and-error to figure out all the values, then order the parts I don't already have and breadboard the thing.  Thanks for your input and wish me luck!

Just a thought: you might want to try one of those integrated transistor arrays (e.g. LM3046) for the oscillators, that way they'd all be at the same substrate temperature and might "drift together" to some degree.  Personally I'd probably go the op-amp route, they make a lot of nice low voltage, low power ones these days.

May look into that later.  Right now I have a bag of 2n3904s and stuff like that lying around begging to be used for something.  If you think an array would help with stability issues though, I'll take you up on that advice when the first twenty tries are totally unplayable.

Jo, like you I am no engineer and started where you are at about ten years ago. Everyday I explored possibilities and think I can and think I can. The theremin is like a lady I believe and she will tease you daily with just enough to keep you enamored.

It was ten years ago an uncle in the newsgroup sci.electronics.basics warned me about re-inventing the theremin with all the good designs out there. Well that pissed me off and so I did it my way. I learned a lot about theremins, some things that you cannot learn from any webpage.

All these years later I realize the uncle was not being negative but realistic, he was wrong about all the good theremin designs out there... but I think I owe him dinner.

"Thanks for your input and wish me luck!"

Good luck!

Also wanted to add that inductors, particularly the "linearizing" inductor in series with the antenna in most good Theremins, can boost the voltage swing at the antenna to many hundreds of volts.  This magnifies any capacitance change and increases SNR.  The capacitance changes you are trying to "measure" are in the femto-Farad range.

There are already a lot of bad and cheap (if not unplayable for the precision player) analog theremins with RC oscillators out there. I would not waste my time by creating another one.

If you want to build something without coils, follow the digital ideas which FredM and dewster are developing here. They are both competent engineers and they know what the say and do.

I for myself am too old to talk about FPGAs and Micro-controllers, so I keep with my coils and I will keep my mouth shut in the digital threads. Even Bob Moog when creating the Ethervox designed only the analogue circuitry himself and outsourced the development of the digital and midi circuit parts. These were realized by the German engineer Rudi Linhard, a renowned Midi expert.