Other Parts Discussed in Thread: INA210, INA213, INA282, INA283
We have come upon a situation where the INA210 suffers from unrecoverable damage in some, but not all, our produced units. When the circuit is damaged, one of two things happen: Either the output voltage matches the applied VCC voltage, or the output is short circuited to one of the input pins.
We have investigated three possible conditions that we suspected might have been the culprit:
1. Common mode voltages in excess of +26V - This might have been the case on an earlier design. However, this was improved a few months back, by changing the zener diode controlling an overvoltage protection thyristor from 24V to 20V. No bench-top testing since has produced common mode voltages in excess of 20V.
2. Fast transients - The new datasheet warns about transients greater than 20kV/uS, as they might trigger the ESD protection circuitry. We have searched all our transients thoroughly, and never seen any transients exceeding 2 kV/uS - a decade lower than the specified limit.
3. Negative common mode voltages - Our design seemed to suffer from negative common mode voltages far beyond the -0.3V limit set by the datasheet. We did a bench-top test using an RC-filter at the inputs to reduce this effect. With our current set-up, the greatest negative pulse is -0.24V at 880nS, but even at this level we are experiencing damaged ICs.
We are using Version A of the IC. It is worth noting that we are also using a INA213 in reverse parallel to measure reverse current with a different resolution, and we have not seen a single failure on this IC.
As far as we can tell, we are within all boundaries set by the datasheet, and completely at loss of any new ideas. Our production line is halted until this issue is resolved, and any help on the matter would be greatly appreciated.
