Dear LucaV,
Thanks for your reaction.
1.I have only that scope picture which seems to me that there is a pulse coming down from 13V to about 9V and ends to 0V at 12 msec. The same line is seen about 13V lower up 10 12 msec. See also the 2 pictures
2. In the meanwhile I measured the input voltage of the bridge during the malfunction. See the picture (AC coupled scope). Is that good?
3. I introduced a delay time in the PWM program of 250 nsec (the minimum possible). But I found no change in the behavior
Does this help in find a solution?
Kind regards
Jan Lichtenbelt