building a CV oscillator

Started by levon, October 26, 2005, 10:12:00 PM

Previous topic - Next topic

levon

ok first post on the new board, ive made a few bits a pieces, a pro co rat, and 2 theremins. but when it comes to understanding what things actualy do im a bit lost, and apart from following the schematics/pcb guides im not sure what im doing when it comes to designing circuts,. anyway.... got a few questions

i have an effects pedal that has an expresion input. which operates between 0k- 50k resistance. and what i would like to make is an  oscillator, that has a pseudo-random wave form, and if possible a sine ,tri and square wave form.

now the questions...
if i made an sine wave oscillator that had an amplitude between 0 and 5 volts, and pluged the +ive and -ive into the expresion pedal, would that do the trick? would a simple oscillator like an audiable tone generator work if i decreased the pitch of the tone. therefore changing the speed of the oscillator?

for a pseudo random effect, the way i was thinking of doing it was use 4 or 5 square waveform oscillators all at different speeds and amplitudes, is there a better way of doing that?

i havent touched digital electronics at all, but would digital be the way to go rather then analog?

has anyone made anything like this before? got any links?, thanks
thanks for your help.

(doug harrison)

It could be that it needs a 50K pot instead of voltage. What kind of pedal is it, and what  does the expression pedal modulate?

If it needs voltage, then you could probably use any LFO.

R.G.

Doug is correct - if it wants a 0 to 50K resistance, it may or may not work at all if you feed it a voltage, and if it works, it may have an odd response to a voltage.

Quotefor a pseudo random effect, the way i was thinking of doing it was use 4 or 5 square waveform oscillators all at different speeds and amplitudes, is there a better way of doing that?
It's OK-ish, not truly random. It will have a regularity to it, even if a complex regularity.

The analog way to do that is to make a waveform generator that generates square and triangle and then shape the triangle into a sine, and also amplify up zener noise to large signals for a really random signal.  There is a lot of circuitry involved in this. A MAX038 waveform generator can do the tri-sine-square. You can do the many-squares oscillators with a hex schmitt trigger CMOS inverter.

It's much less complicated to use a flash memory ucontroller and have it put out the appropriate waveform. It can generate the pseudorandom inside with a repeat period of years.

There is a fair amount of info on this at GEO under LFO's. You'll have to look and see if it's what you need.
R.G.

In response to the questions in the forum - PCB Layout for Musical Effects is available from The Book Patch. Search "PCB Layout" and it ought to appear.

Paul Perry (Frostwave)

I thought I replied to this.. but if so, I don't know where!!
Anyway, I suspect that the effect unit has the 0-50K variable resistance as the bottom half of a voltage divider resistor pair, or else fed by a constant current source, to develop a variable voltage. So what you need, is a way to make a variable resistor, that is voltage controlled. One immediately thinks of a voltage to current converter (or a resistor!) driving a LED in a LED/LDR pair, but that won't go to zero... do any of the EEs think a power mosfet could be used somehow??
I certainly wouldn't feed a voltage to it!! For all I know, the input might go straight to an analog input of a micro & you risk blowing it (though, I suspect it woulod go to some kind of conditioning circuitry).

varialbender

I don't know anything, so don't listen to me, this is more of a question to people who know stuff. Couldn't you test the pedal with 0 and 50k resistance with a lovely meter and then build your oscillator to feed it the voltage that it would have been receiving? Of course, this solution would only work for this specific case (if at all) and it might be nice to have something more versatile.

How far from zero would the LED/LDR be?
What type of control is this varying anyway?
Couldn't the LED/LDR thing work sufficiently well, while just not having as big a range?

levon

its a crapy zoom multi effects pedal, and i can use the expresion pedal to control wah, ring modulator, pitch shift, and a few other effects that i dont see it as useful for.
i was thinking that a 50k expresion pedal would be the same as a 5V CV, but i guess thats not the case.
i also thought about an oscillating LED circut, but looks like that might be ruled out as well.

ill do some reading up about it and post back here.


thanks everyone.

Paul Perry (Frostwave)

Varaialbender, thanks for making me think harder...... on reconsideration, I think that it probably WOULD be OK to connect up a variable voltage, that covers the same range as observed in use. Because the impedance of  the effect input control circuit must be fairly high.
As to how low a LED/LDR would go, I think 500 ohm is easy, under 200 might be hard. Whether this is acceptable depends on the effect, I expect that there would be some effects that you might want to be able to go to zero for. I'd like to hear what people have got from their DIY led/ldrs.
Anyone recommend a cheap zoom or something with one of these mysterious inputs that is suitable for experimentation?

Gladmarr

This still doesn't get down to 0 ohms or anything, but it's an optoisolator that will go down to 200 ohms. http://www.fairchildsemi.com/ds/H1/H11F1.pdf This could be controlled by the Pseudorandom LFO generator from geofex.  It may get you where you're going.  Just throwing this out there, would running two of these in parallel get to a lower "on" resistance?  I think that would also result in a lower "off" resistance as well, so you may be causing other problems while trying to solve the initial problem.

(doug harrison)

Gladmarr-  Good idea, parallelling LDRs. If you use a rectangular LED, you could put one LDR on each side. Off resistance shouldn't be a problem; some LDRs go into the megohms when dark.

Or you could use a superbright T 1 3/4 LED. Face the LDR towards the TOP of the LED for maximum brightness.


Paul Perry (Frostwave)

You can parallel the LDR sections, but the LEDs should be in series (or else driven
by separate resistors) because, one will likely turn on earlier & hog the current.