Hi All,
We've been using a TPS54620RGYR for a design for some time, and are evaluating a series of new external power supplies that should be better suited for our overall design. Ideally, the only change in the circuit seen by the TPS will be an increase in voltage from 15V to 16V. In the existing design, input power comes from either a CinCon Electronics TRG100A150 or a MeanWell Electronics GST120A15, both of which will power up the chip without issue.
We've also connected a benchtop power supply set for 17V, and powered up the chip without issue.
Trouble comes when we connect our replacement AC-DC brick candidate, a 16V CUI Inc SDI120-16-U. Immediately upon connection, this power supply causes the TPS54620 to internally short to ground. No other components on the path are damaged when this happens (an internal circuit breaker in the brick trips, cutting off all voltage and current), and removing and replacing the TPS chip and reconnecting one of the known-good power supplies allows the entire design to power up again normally.
The only difference I can find between the known-good power supplies and the problem one is that the problematic one produces a sawtooth wave with a ~175 ms period and 200 mV switching step:
whereas the known-good supply has a much higher switching frequency and much smaller step, to the point that my oscilloscope doesn't give a very accurate picture of what's happening:
----------
I expected there to be issues switching from one power supply to another, especially dealing with increased noise from using a cheaper power supply, but I didn't expect the DC-DC converter to completely fail every time I plug in the chip. The switching step is very large, but still shouldn't be exceeding the absolute maximum rating on the TPS54620 of 20V.
Any idea what might be causing this?
