Hi all,
we have a design using a TPS54308 to supply a dual-uC system at 3.3V / 700mA.
It works very well, aside from an issue that most probably came from spikes (cable inductance) larger than input abs max ratings, destroying the converter, and everything that was connected to the 3.3V rail... None of the converter safety features seem to work here. We then included a low ohm damping resistor from power input to input caps, and it looks good up to now...
What still is not solved:
We power from a 10 to 22V secondary cell block and are using a hardware power-on logic, starting the converter for a limited period of time (25ms) when power is connected. This starts-up the 1st uC, and if everything is correct, this 1st uC then sets a hold-signal to keep the system alive (EN stays high). If something fails during start-up, the hold-signal is missing and EN falls back to low, even if the power stays connected.
Unfortunately, this hold-signal is not available during flash programming of the uC or if during debugging the uC is reset. Thus, we want to connect an external power supply to the converter output (!) to keep the 3.3V rail at nominal and the uC alive even when the hold-signal is set inactive. While this flash programming / debugging period the converter input power is present (10..22V), but EN would be low.
Are there any restrictions?
Could this external "power from the back" lead to any damage, maybe to the fet from SW to GND?
(we had one damaged converter with R_SW_GND = 8 ohms, but we are not sure if this was from the input spikes or due to externally powering the 3.3V)
I could not find any info regarding this issue in the datasheet.
Thank you in advance!
Kind regards
Gotthard
