Stutter Pedal update

[R.G.]

I updated the schemo at GEO. It shows how to hook up the same circuit as a series or shunt switch by changing one resistor value and one wire.

The shunt switcher can serve for variable depth stuttering by changing Rsource  to a 1M pot.

[R.G.]

OK, this is scary.

It struck me that a good switch is half a ring modulator. I got out the simulator and threw in another opamp to invert the original input signal, and another JFET switch, driven out of phase with the first and coupled both into the output buffer.  Result is classic ring modulator waveforms for a square wave carrier. The 2N5485 that I threw in because it's familiar happens to have very good characteristics for this.

I bet the JFET switch performance is not that good in the real world, but it's sure going to be a strange sound  8-)

[moosapotamus]

Too cool, R.G.!
Think I'm gonna have to give it a shot.
THX!
~ Charlie

[Peter Snowberg]

A while back I found a circuit snippit for an opamp stage that could have the gain switched between +1 and -1. I drew up a schematic that had the audio going through that +1/-1 stage and a 555 with variable pulse width/frequency to drive that. The 555 optionally went into a 4013 as a /2 to make a clean 50% duty cycle. Never built it though..... Functionally it sounds pretty similar this.

-Peter

[petemoore]

I knew somebody's get it...the new ring mod I've been hoping to see!!!
 Funny how things work out when everybody's vibin' on the same wavelength!!!
 Cool find...I'll bet you're pleased !!!
  Way to go RG  !!!

[petemoore]

I knew somebody's get it...the new ring mod I've been hoping to see!!!
 Funny how things work out when everybody's vibin' on the same wavelength!!!
 Cool find...I'll bet you're pleased !!!
  Way to go RG  !!!

[Tim Escobedo]

Hey, RG! could you explain the biasing again. I tried something like this a few times, basically taking the Boss type JFET switch and each time, I found it didn't quite work quite like I wanted. I was trying to get a combo ring mod/VCA type device. I'm interested in your resistor/diode biasing method.

[R.G.]

A while back I found a circuit snippit for an opamp stage that could have the gain switched between +1 and -1. I drew up a schematic that had the audio going through that +1/-1 stage and a 555 with variable pulse width/frequency to drive that. The 555 optionally went into a 4013 as a /2 to make a clean 50% duty cycle. Never built it though..... Functionally it sounds pretty similar this.

See http://geofex.com/Article_Folders/polarity_reverser/polarity_reverser.htmfor at least one version of this from 2001.
The trick is getting the switching transient down. The 2N5485 looks good on paper, but Mother Nature is full of surprises for us when we actually try things out.

I knew somebody's get it...the new ring mod I've been hoping to see!!!

Let's get the practicalities ironed out. As I mentioned, there's some work to do.

[Peter Snowberg]

The trick I had on paper was to use a transition catcher (pulse generator) made from an XOR gate and an RC delay to trigger a JFET to shunt the output. A second XOR was used as a delay to place the clock glitch right in the middle of the muted period generated by the transition catcher.

-Peter

[R.G.]

Ah. Yep, that will work - once again, if the JFET glitches shunting the signal are themselves small enough not to be heard.

I'm wondering if the H11F1 LED Photo-FET might not be the answer to several of these things. It's the only JFET setup I know that is really glitchless without a lot of other junk around it.

I'll go take a look at that.

[R.G.]

... so I go haring off to read about the H11F1 in some depth.

The H11F1 datasheet promises no charge injection at all - no, none, zip, nada. Which is great!

But it only offers 15uS on and off times. For good PWM, you want the on and off times to be under 5% of the full cycle, so the fastest you can run a PWM cycle is 300uS, which is 3.3kHz. Bummer. No analog tremolo there.

However, the ring modulator will work up to 3kHz, maybe higher if you can stand some signal loss. I'll see what I can do with that.

And of course, static or low frequency switching should be GREAT.

[Gus]

R.G.

 What if you differentiated the led drive and turned up the leading edge current but keep the average current in the led s safe area, Or use a sawtooth wave form?

 Is the rise time of the photo fet more due to the fets cap or the led drive?