I'm driving a TC4420 mosfet driver IC with a PC817 optocoupler. The optocoupler tends to have slow rise and fall times and the datasheet for the mosfet driver mentions that slow rise times could end up double pulsing the mosfet. The datasheet does not show any numbers though.

Datasheet: https://ww1.microchip.com/downloads/en/DeviceDoc/21419D.pdf

 

The TC4420 has Schmitt-trigger inputs. Aren't they meant to stop that?

 

To avoid the "slow switching" of the opto you can add a Schmitt trigger stage from the output of the opto to the input of the gate driver. I don't know if you can deal with the total delay time of the opto-Schmitt-gate driver and if you have tight margins for dead time in your application. I believe the opto is adding 3-5 μsec of delay from edge to change in output. Can you really tolerate that? You might have to consider some parts with a 40 nsec delay like the following HCPL 7710/0710 or similar devices.

AV02-0641EN (broadcom.com)

As a bonus the Schmitt Trigger stage would no longer be required.

 

The PC817 optocoupler is too slow. You should use a "digital" isolator.
There are isolated gate drivers with everything in one IC.

 

See if this Chip will meet your needs .................
.
.
.

 

See if this Chip will meet your needs

I have used the 8271 many times. Also the 8273/4/5.