We are investigating the use of the LM5008A 100V, 350 mA Constant On-Time Buck Switching Regulator as a pre-regulator for a 28V-based military vehicle, which must meet MIL-STD-1275 and be able to have its output shutdown with an SCR crowbar (short to ground).
The incoming vehicle battery bus (28 VDC) would be pre-regulated to about 12 VDC (at 150mA to 200mA) through the LM5008A. The most severe high energy transient in MIL-STD-1275 is a 100V surge for 50 msec then a linear decay to 40 VDC over the next 450 msec. The risetime from 28 VDC to 100 VDC is probably a step function (<5 usec) with a source impedance of < 500 milliohm.
The other requirement (not in MIL-STD-1275), we have, is the tolerance of the LM5008A to having its regulated output shorted to ground with an SCR crowbar circuit for an emergency shutdown. The vehicle system would then be reset after the emergency and operation would return to a normal operating mode, without any damage.
We have bench tested one part with a step input from 0V to 90V and the LM5008A was destroyed. We will try again but with a step from 28V to 100V on the input.
Another question is the proper configuration of the output capacitor with additional ESR to ensure some ripple in the output feedback, per the data sheet. What is the best way to daisy chain the LM5008A output to other inputs of downstream regulators (which will have bulk caps with very low ESR)? The downstream bulk caps will bypass the small resistor added per the data sheet. Should the downstream regulators have a small series resistor put in before their bulk caps?
Any feedback, recommendations, or alternates would be appreciated.
