Lend me your brains!

[soggybag]

I have built the ring mod below. It makes a great Ring Mod sound. The one hang up is that the carrier won't stop. It seems the 2n3906 (center) is used to shut off the oscillator made from U1C and U1D. My guess, is that the 2n3906 is used in a digital mode. With voltage at the base the transistor is off and the oscillator functions. When there is no signal at the base the transistor turns shorting C17 and stopping the oscillator.

U3A, U3D and U3C form a peak detector. I've included a detail image with some voltages that I get when nothing is plugged into the input. Obviously there is something wrong. Give me some suggestions. I've been looking at this for a few days now and can't seem to figure it out.

[soggybag]

bump

[earthtonesaudio]

First, I'm going to assume your regulator is not really a 5V type as drawn, otherwise none of the voltages make sense.

I'm not really sure about U3C, especially the diode/resistor in the positive feedback loop...

It does make sense to gate the oscillator off when you're not playing, and a negative voltage applied to the base of Q1, relative to it's emitter, would accomplish that.

My suggestion would be to replace U3C/D with a more conventional peak detector/precision rectifier stage with a lot of gain, add a diode to the output, and finally a resistor from the base of Q1 to ground.

[Eb7+9]

I have built the ring mod below. It makes a great Ring Mod sound. The one hang up is that the carrier won't stop.

one thing that jumps out is the fact that you've got the gain of your OTA strongly biased "on" when there's no input signal ... you might try increasing your bias scaling resistor (R11 4k7 - thus decreasing bias scaling) and increase the signal gain of the first stage (UA3) proportionately ... then, like Earth-tone says, also get the gating to work better on the ref oscillator ... still, any residual LFO signal will get strongly pumped out when there's no signal the way you have it designed ... in this approach you need total gating of the oscillator

I'm tempted to say you'd get better results if you went the other way, signal to the input of your OTA and modulate the bias with the oscillator ... with no input signal, even if the bias is modulated you should get less idling output ... in this case even if you only partially gate the oscillator you'll be modulating noise - that's probably the best you can do without resorting to more band-aids ...

as an aside, notice that many ring-mods rely on a balanced mixer to cancel clock output - relying on a  gate to get a clean output will tamper with the transient response of the ring-mod (ie., at the front-end of your signal waveforms) ... which means the attack will not sound the same by virtue of the turn-on delay of the gating ... not saying this is necessarily bad, just something to consider ...

[soggybag]

Thanks for the replies.

I think there is something funny at R18 and pin of U3C. This should come off of the regulator, which is 5V. So at this point I should be getting about 2V not 4V.

Maybe there's something wrong with the voltage divider at U3B.

As it is I get a great ring mod sound. The sound is really smooth. I think the low pass filtering is the secret to getting a good sound.

The oscillator is almost in audible when the lowpass filter is turned up. When it is turned down you can hear that the oscillator never is not shutting off.

On that point, when I first fire it up the oscillator stops. If I hit a note the oscillator starts up. After the note decays the oscillator continues to run for a couple seconds then shuts off. This continues for a little while. After which the oscillator runs continually.

[earthtonesaudio]

Hm... how is C12 supposed to discharge?

[soggybag]

Hm... how is C12 supposed to discharge?

I was wondering about that myself. I didn't quite understand how the peak detector worked so I let it go by. 

[soggybag]

I was testing things around the Regulator and I noticed that it was putting out about 8V where it should have been putting out 5V! So, I tried another regulator, and got the same thing. I'm not sure what's going on here?

I bought 20 of these LM78L05ACZ regulators from mouser, and have used them on several projects in the past without trouble. So I have to conclude that they work. In this circuit, the regulator doesn't seem to get hot, and everything else seems to be working correctly.

This is not the part listed for this project. The project lists TK1165OCT low drop out regulator. You can only this part at Digikey and they cost lot. So I opted for the parts on hand.

What could make the regulator I have put out 8V instead of 5V? When I tested. I didn't connect anything to the input or the output. But, the rest of the circuit is there. I would guess that the voltage divider R6 and R8 would provide a sufficient load. What am I missing?

[earthtonesaudio]

Maybe a ground fault at pin 2?

[soggybag]

Maybe a ground fault at pin 2?

I'm not sure what you mean. If we're talking about the LM7805 pin 2 should go to ground?

[soggybag]

Looked up the data sheet on the LM7805 and it turns out I had it in backwards, doh!

I used Eagle to create the PCB and gaussmarkov library. I used the 7805 part there. It must have different pin out.

It seems to be working now. There seems to be a delay before the oscillator shuts off and when you stop playing. This was predicted by astute minds above.

There also seems to be a problem with the oscillator. The range seems to go in audible on one end. In the other direction the oscillator seems to shut off about a 1/3 of the way before the end of travel for the pot. I can't seem to tune it below the open B string before it shuts off.

The sound is very much a Ring Mod. But, it sounded better before I fixed it! Not sure what was going on before, but the sound was very smooth and clear. If you could describe a ring mod as "transparent" and "clean" this would fit the bill.

[snap]

a 78L05 is not a 7805 !