Tool/software:
Attached is a diagram/doodle of the circuit being used. Conditioning components are mitigated for simplicity but they all conform to datasheet recommendations for filtering and the like. The PCF8574A is being used to provide a positional ID to the MCU over I2C via a TI ISO1640DWR bi-directional isolator. The fault is observed when the various boards these components are mounted to come together forming the circuit as seen on the diagram.
With the Isolator and everything to the right of it not connected, the SDA/SCL lines are high - as one would expect being pulled up to 5V_1. Once the rest of the circuit is introduced these lines sink low permanently and can only be recovered by replacing the PCF8574A IC. The MCU does not need to be connected for this fault to occur. Simply attaching the board with the isolator (with it's associated 4k7 pull-ups) is enough to cause the lines mentioned to latch low. I suspect something is damaging the IC.
5V_1 and 5V_2 share the same 12V source and reference ground, but VCC of the PCF8574A and ISO1640DWR are provided by separate DCDCs due to design restrictions elsewhere.
Any ideas as to why this is occurring is appreciated immensely. Any further details, anything I've missed which may be important, I will gladly add.
Kind regards,
Darren
