Other Parts Discussed in Thread: UCC27714, UCC28951
Hi Team,
My customer is having a UCC28950 issue with long on-times on the D FET causing high current and eventual FET failure. They tried increasing the slope comp and it did not help, so they believe it might be due to noise in the current sense cabling. It only happens at very light loads when the controller is right on the edge of burst mode.
They hypothesize that the controller is about to change states from continuous to burst mode when noise causes it to lock the “D” output for 2 clock cycles. It can only be replicated with their customer's load which is fairly noise. Have you seen this issue? Any initial thoughts on what could be happening and how to mitigate it?
Please see waveform below:
-CH1: transformer current
-CH2: Vds of B FET after going through a UCC27714
-CH3: D FET drive out of 28950
-CH4: Vgs of D FET after going through a UCC27714
Thanks,
Antonio