H11F1 optocoupler with fast transients?

[austin]

I'm building a trem using the H11F1, right now configured between two buffers.  I was having popping problems with fast transient LFO waveforms (square, ramp, etc) and solved it by slowing down the transients.  Problem now is it's not quite as choppy as I'd like.  At slower speeds, it's okay, but fast speeds loose a lot of edge.  Anyone use this optocoupler with really fast transients?

[mikeb]

Perhaps it is the rapidly changing voltages of the LFO itself that might be inducing popping in nearby audio lines?

Mike

[austin]

Possibly... I have separate power supplies for the LFO and audio parts.  Large caps across the power rails.  The LFO and audio circuits are on separate breadboards... i'm learning eagle right now to get them onto a pcb.   Maybe the breadboard has something to do with it.  Other than the H11 chip itself, the control voltage is not physically near the audio lines.

[mikeb]

Ahhh ..... good thinking, all that. I can see why you're wondering about the device itself then. I've never used those (expensive down here) so can't help, sorry.

Mike

[zachary vex]

i've seen induced popping from vactrols before, so i wouldn't be surprised if you get it with a photofet.  try driving the led with less current (use greater series resistance).

[R.G.]

i've seen induced popping from vactrols before,

Just as with relays.

Every single known signal isolator is prone to capacitive coupling of control signals and shared ground noise if the ground is shared.

The closer the control signal comes to the controlled path - like in a semiconductor device, or a vactrol - the better the coupling is and the harder it is to keep the pops out. More importantly, the higher the impedance on the signal side, the easier it is to couple transients of any kind in, including particularly control signal discontinuities.

The solutions are always
(a) make the control signal change as slowly as you can. A full change of state over a couple of milliseconds is "instant" to the human ear, and does not couple in fast edges
(b) make the impedance on the signal side as low as you can. No, you can do better than that...
(c) think about where the ground currents flow and make sure that the signal ground does not share a common conductor with the LFO ground on the way back to the power supply. If possible, cut the grounds loose entirely, or try to couple equal signals to signal and signal ground.

[ExpAnonColin]

Z Vex-The problem is that with less current, the resistance range might mess up for worse tremolo action...

You could try raising the impedance to ground from the H11F1, on the signal ground...

Also, doesn't this have to do with the fet inside the H11F1 not being biased?  Do you have it biased?

-Colin