We are using the UCC3808N-1 in a mature product, which have been produced for +15 year with no significant issues. Now suddenly in medio 2020 we began to get failures on fielded products, where the UCC3808N-1 has failed. The UCC3808N-1 stops outputting pulses on OUTA/OUTB. On some ICs it will also stop outputting a clock signal on RC, but on others the clock may still be running.
The issue seems to be primarily on ICs with the ID codes: UCC3808N-1 (87B, WF63). But also seen on (96B, W80H) and (8AB, W89F). We have not yet seen any returns on repairs with batch (95B, W80N) or (8AB, W89J).
We have made som minor adjustments to the design afterwards to improve resilience:
- Increasing the gate resistors R8/R9 to 33R (from 10R)
- Reduced C1 from 1nF to 470pF (to ensure timing resistance >10k Ohm)
But have seen a failure with the changes implemented.
VDD to the IC is ~14V +/- 0.5 through a 100R series resistor limiting the clamp current.
Frequency is 278kHz and pulse width 2.3us
Questions:
- Can you provide any hint towards the root cause based on the failure mode? (No pulses on OUTA/OUTB)
- Are you aware of any changes, deviations or counterfeit cases related to the IC batches?
- Can you provide insight into how to decode the lot/date of the ID codes?
Best regard,
Tommy
