bigo said:
We're using a Telecom TC4422cpa mosfet driver chip and until recently we
didn't have any problems. What's now happening with some ICs is that with
the input held at zero, the drivers try to turn on together every so often
so that the current can reach as high as 50ma, but the chip somehow regains
control. Has anyone else observed this behaviour?
jjj
These parts are sourced by different vendors and are mistakenly
treated as jellybean replacements.
Make sure their hasn't been a source change.
A recent shift from TC4427 fabbed by Microchip, to MIC4424 fabbed by
Micrel, showed abnormal performance that no amount of increased
decoupling or input and output protection could remove.
This resulted in both outputs of the 4424 going simultaneously high,
without measurable input signals, for a short period. Input signal was
highz current feeding a 1K5ohm pull-down resistor. Output drives were
buffered for pull-down only, and schottky/zener protected.
Returning to the original part, from the original vendor, was the last
step taken to solve this problem.
There are many precautions advised by the vendors when using these
parts, and they should all be followed. It won't solve the problem
described, however.
In many applications, this behaviour would not cause abnormal
operation sufficient to warrant investigation. In the case described
it was only recognized after unexpected non-destructive shoot-through
current was noticed, in the course of other testing.
RL
