Good day,

I am building an H-bridge (picture below). My opto couplers were ILQ74. I am using 560 ohm current limiting resitors for the opto coupler diodes, with an input voltage of 3.3V (from MCU). My calculations proved that even 100 ohm would be enough. However, it already happened to me twice that one of the opto coupler diodes brakes (with two different opto couplers). I measured the resistance over it and I get a value of 3 ohm (definitely broken). Could it be just a "bad batch" or something else? enter image description here

  • 2
    \$\begingroup\$ It could be because you're exceeding the peak reverse voltage when you reverse the polarity: datasheet bottom of page 1. Try putting a small signal diode like a 1N4148 in parallel with each one, but with opposite polarity. \$\endgroup\$
    – brhans
    Nov 22, 2016 at 16:36
  • \$\begingroup\$ I agree with brhans. Use a freewheel diode across the optocouplers like you do across the TIP107 transistors. \$\endgroup\$
    – DerStrom8
    Nov 22, 2016 at 16:42

1 Answer 1


I agree with the comments, you are exceeding the reverse voltage spec of the opto LEDs and this is most likely what is causing the failures you see. The LEDs in the optocouplers are rated to 3 V reverse voltage maximum, and you hit them with 3.3 V.

There appears to be another problem with your design; you have insufficient current flowing in the LEDs to ensure the output can saturate. Assuming a +V of 12 V then to saturate the optocoupler output transistors requires just less than 12 mA. The optical conversion efficiency is only 35%, so this requires a minimum of about 4 mA through the LEDs. Normally you would increase the current to ensure you met all conditions, and I'd suggest the minimum LED current should be 8 mA.

To get 8 mA through the LEDs using 3.3 V logic with an LED VF = 1.2 V you need an R value of 270 Ohms maximum.

The top pair of optocouplers can be configured as self protecting since they are only driven by the FWD and REV signals.

The bottom pair of optocouplers need protection diodes across them. The maximum source or sink current is about 16 mA so this should be ok for any microprocessor output.

I hope the schematic makes sense since I only included the LED circuit stuff.


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ I'll just add a diode at every LED, since the opto coupler LEDs that were blown were top left (PNP) and bottom right (NPN). \$\endgroup\$
    – Matthias
    Nov 23, 2016 at 6:29

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.