Ring Modulator question for guitar

[digi]

I'm planning on building a ring modulator based on this circuit:
http://www.musicfromouterspace.com/analogsynth/RingModulator/RingModulator.html

For the carrier input, I was going to use some op amps to make a wein-bridge oscillator, a schmitt trigger with an integrator to produce a sine, square, and triangle wave generator. I just wanted to know what a good frequency range would be for the carrier signal to be used with guitar (I know the output frequency will the be the sum and difference of the two signals, I just wasn't sure what would sound good for a guitar range). Also, anyone know what the typical peak to peak voltage is from an electric guitar? I'm sure I'm going to have to boost the guitar signal a great deal.

This is actually a project for one my labs in college (I'm an EE major), anyways any help would be much appreciated!

[ringworm]

I had found this info on HC it might be of use

PITCH (Note name)     FREQUENCY (in Hz)
Bass: E string             41.2    
Bass: A string             55.0     
Bass: D string             73.4     
Bass: G string             98.0   
Guitar: Low E string     82.4
Guitar A string             110.0    
Guitar D string             146.8    
Guitar G string             196.0    
Guitar B string             246.9       
Guitar E string             329.6    

[Mark Hammer]

The rule of thumb is that the lower the modulating frequency, the more "pitched" the sideband products will seem to be. 

This occurs for a number of reasons:

1) If you add and subtract a small number from a much larger number, what you get is still something close to the much larger number..

2) When the modulating frequency is high, the difference or sum sideband product will sometimes be outside the bandwidth covered by the circuit.  For example, a 400hz note modulated at 2000hz will have an audible sum, but no audible difference.  A 2khz pitch, modulated by 1khz may have a more audible difference than sum (because of lowpass filtering to keep noise at bay).  As a result, there can appear to be "missing" sideband products as one moves around the fingerboard.

3)  Because the sidebands are not proportional to the original, the degree of "unpitched" quality to notes played will change depending on how high the note played is.  If the modulating frequency is smaller/lower, then higher notes will seem to be closer to true pitch, while lower ones will seem more off pitch.  If the modulating frequency is high then sums and/or differences will seem to disappear, depending on how high or low the note being played is.

None of this is quintessentially "wrong".  Rather, it is a question of how you get a desired sound.  My own preference is to stick with modulating frequencies below 250hz or so, such that when I want a riff to sound like it's going up, it goes up.   Higher modulating frequencies are great for introducing a sense of randomness.  Me, I just aim for a "rubber band" texture and that's enough for me.

Since Ray Wilson's circuit assumes you are feeding it a pure and constant waveform from an oscillator, I will mention just a few more considerations:

1) A guitar produces a very UNsteady signal.  It is useful to precondition the signal in some manner that results in a fairly steady amplitude.

2) Ring modulation has its origins in use of simply oscillators with highly controlled harmonic content (often pure sine waves).  Producing sum and difference products of harmonics is what cymbals do.  If you want to actually hear notes, then you need to trim as much of the treble (harmonic content) off the guitar signal as you can to come closer to sine-wave conditions.  A nice steep 4-pole lowpass filter ahead of the input would help.  It'll make the result sound more musical whilst not robbing it of its weirdness.

[Paul Perry (Frostwave)]

How wide a freq range you want, depends on what you arfe intending to do musically..
personally, when making the Frostwave Blue Ringer, I found that some customers wanted to take the local oscillator down to 0.1 Hz.
And I personally think it is a mistake not to have the oscillator go up to at least 15KHz..
Which shows that the 'average' LFO circuit will not cut it.

A second point, is that I used a circuit (based loosely on the kind of thing that you find in an analog synthesiser, with a roughly exponential control to give an even spread octave wise as one tunes it) that gave a ratehr pure sine wave.
i wish now that I had put in a waveform control as well.

Good luck!

[digi]

Thanks for the response guys! I've done some simulating and ordered some parts, I just need to get into lab and play around with this thing.