Optocoupler: decoupling circuit parts sharing same ground and power line

Question

Consider a circuit contains a 5V supply that can provide quite high current, say 2A. Now, on one side of the circuit we have some control device, for example microcontoller, on the other - device that uses quite a lot of power, for example a high-voltage device powered with DC-DC boost converter which converts our 5V to 200V. But, our high-voltage device is controlled via 5V line, so technically we can connect it to MCU output pin with a resistor, to limit current consumption.

Our other option, which is actually my question, is to use optocoupler like 4N25M to split MCU and high-voltage device in such manner that the only things they share are 5V line, GND line and the optocoupler device.

Finally, I can now formulate the question - does it make sense to use optocoupler in such a circuit? When saying "make sense" I mean than when I use optocoupler between sides of the circuit that are completely isolated and doesn't use same power source - optocoupler can protect control device of being damaged in case of input pin of high-voltage device suddenly got connected to HV line (in case of damage for example). But, if two sides of our circuit actually shares same power supply, I am not sure if such usage of optocoupler makes sense at all.

Answer 1

If the step converter is also isolated to the level of insulation required adn is short circuit protected by design then it would not defeat the isolation of the optocoupler.

The bigger question is why would generated voltage become shorted and why would that damage a controller if one was careful to protect it without insulation but rather just high resistance and clamp protection.

Obviously the choice is yours as your needs are unknown.

However when I designed a 2nd source supply for proposal at Lexmark printers, when this model was being prioduced at 1k units a day, it required no feedback isolation from HV output to 5V PWM and 24Vdc input , it generated regulated 3 non-isolated outputs with sweeps from -10kV to 15kV and used a string of 10M resistors to convert to the 0 to 5V range and was fully protected from damage with diodes and measured by differential input with HV return line sharing leakage to stay within the CM input range. This was the Vcharge transfer controller to drum for laser printing and was accurate to 1%.

There are still many details to consider and a few obstacles to avoid for crosstalk of CM noise, inbalanced CM to DM pickup but there is no problem per se with what you suggested, except it may not be necessary.