If a circuit uses a multi channel opt photo transistor type isolator, what is the isolation rating between channels on the output side.
Looking at stand DIP type packages with 2.54mm pin spacing (and pins about 1.2mm wide), I'm guessing that the gap of 1.3mm between pins provides much less than the rated isolation voltage of the device.

Would potting the output side in epoxy help to raise the rating (preventing flashover etc) ?

 

hi f,
All the opt-isolators I have used have the Input side on one row of pins and the output side on the other row,
ref this image.
So the isolation spacing is not 0.1inch minus the pin width, its around 0.3inch.
E

 

what is the isolation rating between channels on the output side

Doesn't the datasheet tell you?

 

If you need high voltage isolation between outputs (not input to output which is the normal concern) then you should use separate, single channel optos in different packages.
What is your application that you have a high voltage between outputs?

 

What is your application that you have a high voltage between outputs?

It is a control circuit for the FL5150 dimmer, where the GNDA on the device is floating with the mains input. Hence the need to isolate the input PWM.
The circuit is shown at https://easyeda.com/mriksman/fl5150-with-opto-and-opamp_copy_copy_copy_copy and employs a negative feedback to the op amp buffering the input to linearise the response.



My concern is that GND and GNDA are only separated by the distance between two adjacent output pins.

 

Doesn't the datasheet tell you?

The datasheet shows capacitance between channels, not voltage isolation.

 

My concern is that GND and GNDA are only separated by the distance between two adjacent output pins.

I'm guessing using two separate isolators, rather than a dual channel device, would be safer. But then you run into CTR matching issues etc.
I'm not familiar with how CTR differs between channels on the same device, vs between two instances of a single channel device.

 

Why do you need such high linearity?

If you drive the opto input with an op amp current-source, then the opto output current is fairly linear with respect to the op amp input voltage (Ltspice simulation of example circuit below).