College students are a pain, right?

Hello folks.

So, three friends and I are in this vibrations and controls laboratory during our last semester as mechanical engineering undergraduates. We got the idea to build a theremin, you dig?

Sheesh, most of us have only recently heard of these intergalactic instruments, and now we are committing to building one--with ~two months to graduate!

I would love suggestions for kits or a schematic/parts lists to shoot for, based on the fact that goofy college students are having a go at this. I'm leaning towards the Theremax by PAiA, another kid is looking at the kit at http://thereminkits.com/, even the non-heterodyne, optical theremin for its simplicity (outrage?).

Also, if you had access to an oscilloscope, NI LabVIEW, and NI ELVIS, how would you put it to good use as far as performing signals analysis on your theremin?

Thanks!

The simplest Theremin circuit possible is a pitch-only design using a single 4093 cmos part.. one gate of the 4093 is configured as as a reference oscillator (one resistor and trimming variable resistor, and one small capacitor of about 22pF). One gate is configured as a variable frequency oscillator (2 resistors and a neon bulb for ESD protection).. Outputs from these oscillators drive another gate giving an ANDed output to achieve the equivalent of hetrodyning, and the output of this gate is taken to a simple RC filter to filter out the 'sum' component and leave an audio frequency difference signal, which is the output.

This circuit is at http://www.therasynth.com/html/a4093cct.html

About observing waveforms etc - you would want to look at the frequency from the pitch and reference oscillators, and tune these to be within audio frequency difference of each other.. for example, if the pitch oscillator range is 200kHz to 250kHz, you would want to set the reference oscillator so that its potentiometer spans (can adjust over) this range..

Taking the schematic above, trim R2 so that the frequency on 4093 pin 4 is not less than 180kHz and not greater than 250kHz when a short antenna is connected (one wants to get the equivalent capacitance for the player being furthest from the antenna). then adjust RV1 to be in the order of this frequency... It is a good idea to have another potentiometer between pin 3 and the trimmer RV1, and this would be the user control (it should be about 100k and set mid way to begin with).

Check the audio output with a scope - this output should be able to drive a Hi-Z amplifier.. If you find the amplifier loads the circuit, you could reduce R3 down to a MINIMUM of 470 ohms, and increase the value of both C3 and C4 .. C3 will determine the rejection of high frequency, C4 will determine the low frequency response.

One can trim the user control by putting a resistor in parrallel with it, or tailoring its value, to give a 'friendly' control range.

4093's vary quite widely in their threshold voltages, so this circuit can be a bit of a pain to adjust.. but apart from that the circuit is extremely simple and cheap. If you find the neon a bother, it can be left out - but you are advised to have the 4093 in a socket, so it can be replaced if zapped! (cost of 4093 is less than a neon).

Also - The ground must eather be connected to earth, or to some form of radiator (metal box for example).

I have now added an audio output amplifier and some minor refinements, and done a simple breadboard layout.. and pasted this at http://www.therasynth.com/html/a4093cct.html

This was in response to an email.. I include this here as it may be of interest to others:

>> Have you built this? Do you have a picture of a completed circuit?

I have built numerous versions of this circuit into other products, but do not have a "stand alone" version I can photo or demo.

>> You say the output can be connected to a High-Z amplifier, is a guitar amp high impedance? So, this output can be a 1/4" jack, which
can be connected to any amp?

The ideal would be to have an op-amp configured as a follower, before driving an instrument amp.. However, with component changes, it should just be capable of driving an amplifier directly

>> Where is the voltage source for this circuit? Is that 9 volts into the CMOS chip?

Yes, the circuit shown is a full simulation. 9V would be applied to the 4093 (pin 14 to +9V, pin 7 to 0V) - You should also be sure the supply is stable (no problem with battery) and have a 100n decoupling capacitor soldered between pin 14 and pin 7, as close to the IC as possible, with as little wire between the IC and the capacitor as possible.

>>I do not understand the distinction between the ANTENNA and C2, where you say that C2 is the capacitance of the circuit. Please clarify.
>>Suggestion for an antenna design?

C2 is not actually fitted on the board (or anywhere).. It is put on the circuit purely to allow the simulator to 'pretend' that an antenna is connected.

The capacitance of the circuit (capacitance between IC pins, tracks on the board, wiring to the antenna) and the capacitance of the antenna to 'ground' (the battery 0V, or - terminal) is usually in the order of 20pF, going up to about 25pF when a conductive object is extremely close.
NOTE: IT IS CAPACITANCE BETWEEN BATTERY - AND THE ANTENNA WHICH IS IMPORTANT.. IF BATTERY - IS CONNECTED (Via, for example, the earth on an instrument lead) THIS WILL IMPROVE COUPLING. I HAVE SHOWN A GROUND "ANTENNA" TO EMPHASISE THIS.

See my posting on the " I want to know everything about the antenna" thread for more information on antennas.

>> What are the blue arrows, pointing to R#(#)?

These are the test points on which measurements are taken (scope connected) - If you look at the waveforms (say the bottom right one, labeled "VARIABLE (PITCH) OSCILLATOR" you will see R2(2) above this waveform.. this shows that the waveform is what you would see if probing the point marhed with the blue arrow as R2(2).. Effectively, each component has reference 'terminals' - a resistor has terminal 1 and terminal 2.. with resistors, the order these terminals occur in a circuit does not matter - with diodes, for example, it does. The reference R2(2) is simply shorthand for Resistor 2 Terminal 2.

>> What is up with the variable resistor, RV1? Is the green arrow simply to indicate that the potentiometer is set to its half-way position?

It is a variable resistor constructed from a potentiometer, the pointer is the wiper, and, yes, as shown it is set to mid position.

A potentiometer has 3 terminals.. These are the resistor track ends, and the wiper.. connect the wiper to one of the track ends and you get a variable resistor. IT IS A GOOD IDEA TO DO THIS WHENEVER YOU NEED A VARIABLE RESISTOR - If you simply connect between wiper and one end of the track, the wiper going open-circuit will present a high impedence.. Tying the wiper to one end means that worst-case, the failure resistance seen will be the full potentiometer resistance, not a disconection.

As (I think)I mentioned, RV1 should be a trimmer, and another REAL (manually adjustable) variable resistor should be placed in series with it as a user control.

The most (only) tricky bit of this circuit is getting the values f