Advice Request

Hey hey,
New guy here :-S
I'm wanting to make a theremin as part of my final year project at university. I spoke to one of my lecturers and he claims that making a theremin with valves wouldn't come up with great results due to the valve generating alot of heat and thus making the tuning be unreliable. I was wondering if anyone may be able to advise me on how true this is?
Also, does anyone know how hard/easy it is to make an analogue circuit that can give the user the ability to change the wave-shape of the signal generated by a theremin? (or would anyone be able to point me in the direction of any books/websites that may help)
Sorry to be a pain, but thanks to anyone who may be able to help :-D

Welcome to ThereminWorld efwa,
I *hope* you've come to the right place.

Must be the time for it, a couple of final project
posts (http://www.thereminworld.com/forum.asp?cmd=p&T=3570&F=3) have come up recently.

Unfortunately I'm just a player not a technician, and do not know what valves are in the context of the theremin, I hope some of our tech-head members do and will chime in.

If nothing else, keep trying the search here.
Despite the troubles database growth is causing us, we've kept everything said in the forums from day one. (and if you know our memebers, that's a boat-load!)

The Theremin is a fantastic project for a lot of educational goals. Please do us the favor of keeping us posted.
Good Luck!

Valves is a synonym for tubes.

There's some folk around here that know about tubes and have made some jolly good theremins with them.

Changing the waveform is as hard or as easy as you want to make it - a diode will change a waveform fairly radically. In fact I have a theremin that has just such an option, but I wouldn't call it an improvement.

Valves.. take time to warm up, heat can cause badly selected components close to the valve to give drift - essentially, capacitors and resistors have thermal charactaristics which are usually given as PPM/degree C (parts per million per degree C) which the value will change by.. Polypropylene film capacitors are in the order of -200ppm/C - one needs to calculate possible temperature variations, and what effect this will have on the circuits stability.. but you need to do these calculations for any circuit, be it valve or semiconductor.

Trimmer capacitors can have quite bad thermal charactaristics(+/- 500ppm/C), as can potentiometers(+/- 250ppm/C for Cermet, +/- 500ppm/C for carbon), so care must be taken in selection and placing of these.. 'Standard' metal film MF12 and MF25 resistors have +/- 50ppm/C.

Waveshaping: Best to look at this in combination with whatever hetrodyning mixer circuit you adopt.. (assuming you are going for a conventional Theremin) .. The Jaycar / EPE Theremin has one of the simplest designs I have seen, using biasing of the MC1496 mixer to give waveform and symetry controls - This is quite a neat solution, as the MC1496 is quite a good balanced modulator, and certainly beats the simple diode mixer used in the Moog Etherwave.. However, it takes a bit of care to set up.

You can do wave shaping either before, in, or after the mixer - the 'before' option is the most tricky.. There is only one Theremin I know of which uses this scheme, the Moog 201 - Waveshaping is done by selectively filtering the high frequency signals from pitch and reference oscillators.. say the pitch osc fundamental frequency is 200kHz, and goes up to 205kHz.. Filters tuned to 200, 400, 600.. kHz will effectively, when mixed, allow harmonics F,2,3.. to be mixed... and produce audio with these selected harmonics only...


Post mixer waveshaping is easiest.. get the audio (out of the mixer), produce a square wave with a zero-crossing detector, shape this wave as you wish, mix this/these back into the original waveform in whatever proportion suits your taste..

The Etherwave has an extremely simple squaring / distortion scheme where the gain and bias to the transconductance amplifier (VCA) is varied - and people seem quite happy with this (see HotRodEtherwav.pdf .. I cant find a link right now, but Google for it..)

Last - there is the VCO/VCF method.. From the audio output of the mixer, on drives a pitch-to-voltage converter.. Then use this voltage(via an exponential converter unless you use a phase-locked-loop system for your VCO) to drive a VCO, and you can have your VCO generate any waveform you want.. This voltage can also drive a VCF for filtering of this waveform.

There are so many links to Theremin design etc, that I dont know which to give.. But perhaps the best starting point is to read all the articles by Bob Moog, get your head 'round his designs..

Fred wrote: "The Etherwave has an extremely simple squaring / distortion scheme where the gain and bias to the transconductance amplifier (VCA) is varied - and people seem quite happy with this (see HotRodEtherwav.pdf .. I cant find a link right now, but Google for it..)"

I'm really not happy with this, especially for lower frequencies the tone of the Etherwave Standard is harsh and not at all beautiful, independent of the timbre and waveform settings.

That's why I have a Tvox Tour on top of my wishlist or, in case they will really not restart production, I'll have to design my own theremin.

Thierry wrote:"That's why I have a Tvox Tour on top of my wishlist or, in case they will really not restart production, I'll have to design my own theremin."

I heard (and saw) the Tvox played (by Lydia) for the first time a few days ago, and was extremely impressed.. But then, I think Lydia could probably make any Theremin sound awesome! Having said that.. The low tones were absolutely superb - I have not heard anything like them from any other Theremin.

With regard to the tone from the E-wave.. There are all sorts of things one could do to improve this without needing to completely redesign the instrument..

The simplest route I can think of would be to build the mixer of the Jaycar/EPE Theremin, and feed the reference and pitch oscillator signals from EW into this (remove the horrible diode mixer etc.. ) one could take the audio from the EPE board back to the VCA on the EW, or you could use the EPE VCA (probably slightly trickier)..

If one only used the EPE mixer / tone circuits, and fed the audio back to the EW VCA, you could leave the EW tone control, AND have the additional 'brightness' and 'waveform' controls of the EPE... This should give a huge range of possible (pleasing) tone variation.. Provided that one set the levels from the EPE correctly so that the EW VCA was not overdriven.

An even better route would be to do the above, AND/OR to fit a pitch->voltage converter, then add a VCF controlled by the pitch voltage.. The VCF roll-off frequency and resonance can be adjusted and any 'raw' signal filtered, to allow an output from sine (maximum attenuation of harmonics) to unfiltered.

I am [b]almost[/b] 100% sure that the Tvox uses a VCF on its output - it had (to me) the sound one gets from an under-driven 24db/octave ladder filter, with cut-off set to about an octave below the Fundamental, and the Q quite low.

I use very subtle LPF on most of my recordings. It's good if you want something to sound "muffled" of course, but very gentle filtering and resonance can enrich a track of audio to no end. Pretty sure I'm going to get a VCF for live use in much the same way.

"I use very subtle LPF on most of my recordings. It's good if you want something to sound "muffled" of course, but very gentle filtering and resonance can enrich a track of audio to no end."

The main reason why sound through a fixed frequency LPF can sound "muffled" is that, as the audio changes (frequency WRT cut-off) spectral signiture of the sound changes - the brain (rightly) interprets this as an "object" 'muffling' the sound.

VCF's are often thought of in terms of a 'dramatic' effect - this idea may be the result of extensive use of envelopes changing the filter cut-off, and the peaking at the cut-off when Q is set high - The "Moog" sound..

But - If the VCF only tracks the fundamental, and Q is at a sensible level, then one can generate a smooth output waveform with required degree of harmonics, and the harmonic signiture will not change as the pitch changes.. This achieves the 'mellowing' of the sound, but removes the cues which cause the brain to interpret the sound as "muffled".

Back to the topic .... "Valve Theremins"

I am really surprised none of you directed this person to the Georgia Tech project! There's a link on the schematics page (GT Theremin).

The "bad news" part is that you will really have to try hard to find variable tuning capacitors like they used in their project.

You can buy most of what you need from Antique Electronic Supply, but it might be a bit outside your budget.

Check out other schematics on that page as well! There are lots of good ideas there.

Don

If you do a search for guitar effect (maybe start at www.harmony-central.com ) schematics, you will find a lot of ways to alter the timbre of a theremin.

Distortion circuits are easy to build, the simplest involving 2 diodes in parallel, in opposing directions, placed across the line output or in series with a resistor with the whole network across the line output.

Tone control circuits are also easy to build. Tone controls are basically just dedicated filter networks.

Spring reverb is pretty easy ... digital reverb is a bit beyond a term-end project.

Don