We are using bq27210 in an opamp powered configuration as per figure 1 of SLUA387 example. The battery is a 12S NiMH pack (10-18.5V), feeding an appropriate divider. The voltage seen by the gauge is in the 2.3 to 4.3V range. Most of the time that works out well, but in some cases we end up with a gauge that stops measuring analog inputs correctly- it would start reading current as 100x too high or too small for example. Oddly enough it would get stuck on a low side one day and then bounce to high side the next day. It would not return to normal once that begins. On one of the boards the BAT pin is stuck at 0.9V, seeming to indicate internal damage. Are there additional precautions that should be taken around the gauge in such a configuration, or any known issues discovered since then? The opamp is powered by a much higher rail, so it could potentially swing above the 7V abs max for the gauge. We still have 1K/0.1uF filters on VCC, BAT and SRN/SRP pins. Schematic fragment is attached