Ulrich:
We had 2 controllers on 2 prtotos fail- on one NTC pin went to 27 Ohm to GND pad.
On proto1, the controller that we hoped to fly off with - had NTC pin and REF go low impedance.
On proto3, the controller that we hoped to fly off with - had NTC pin go low(30) Ohm
Of the total 3 or 4 chip failures, NTC shows up 3 times. REF twice.
I think we can hypothesize that putting 1 uF on REF perhaps causes REF pin drive too much current...but how is that possible?...but assume it is possible.
We find .1uF won't get REF top correct level, but min .22uF will.
Assume we use that now on.
NTC?
There is nothing on this pin. We even disconnected the Opto form this ; we only have 30.5k or so from pin 9 to GND.
And on startup, this pin is not even active except perhaps a short( low impedance).
We have ruled out ESD,. soldering temp, overheating etc..
@ possibilities: somehwo NTS does not work well with resistor?...sounds strange.
#2 is REF on changing level, drives higher current through NTS current source?...also sounds improbable.
In any event, took up all our time to dif into the root cause.
Yet, we did apply VBULK to 200V...RUN happened, some PWMLs ...but did not go into a steady state....we are considering if VBULK somehow couples to NTC?...
SOMETHING is still not right.
Any guesses from you will be highly appreciated.
robin
