AnsweredAssumed Answered

AD8422 latch-up (PCN 14_0158)

Question asked by JKK on Jun 8, 2016
Latest reply on Jun 14, 2016 by brian.pelletier@analog.com

We are developing with AD8422ARMZ in amps and have noticed what looks like a latch-up issue with "balanced" positive and negative rails. The only discernible symptom of the problem is that the input terminals go to one of the rails. Since reading the PCN, we have identified that the batch of ICs we are developing with fails within the effectivity date.

Our remedial action so far has been to replace the ICs within the effectivity date with samples outside the effectivity date (date code 1551) and so far this has proved to be a successful remedy. The slightly concerning issue is that this has occurred very repeatedly on all our chips and occurs with the inputs tied to functional ground via low impedance (so not just at high impedance as indicated on the PCN).

My questions are:

  • will there be an uprev to the datasheet (as indicated on the PCN) and when is this likely to be published?
  • how confident can I be that the problem is well under control with appropriate date code management of this part?
  • it is slightly unclear as to whether the device complied with JESD78 before the PCN. Can this be clarified please?
  • At power-up our inputs may be at high impedance (the input impedance of the AD8422). Do we need to tie down with high-value resistors? This would degrade our input and CMRR spec.

Outcomes