So I've been having a PSU failure day. Was working with a synchro test set running off a little vintage Ferranti 115V 400Hz inverter. Inverter powered by a 360W variable bench supply (Farnell G30-12 in a box with meters etc) set to 28V. Suddenly the inverter stops (no 400Hz whine from test set) and input current goes to 10A
Switched off quickly, disconnected the inverter and checked the DC supply - all OK.
No info on the inverter, I don't even know it's rating, but opened it up to have a look 4 TO3 power transistors a couple of power diodes, some magnetics and a PCB wit small transistors R's C's and a mistery black epoxy potted block with about 15 wires coming out. Anyway quick check of TO3's (3 x 2N3772 and one house marked) showed a probable short. Disconnecting emitter and base confirmed a C-E short on of the 2N3772's. Coudn't find one in my workshop stash and it's tipping dowon so not going out to the garage. Circuit around two of the transistors, including the shorted one, is clearly switching the ends of a centertapped transformer primary to ground. Clearly not parameter critical and as I had a couple of 2N6338's handy I replaced both the 2N3772's.
Switched the mains on the DC supply on, before connecting inverter, and there is a flash and a bang and that dies
.
The DC supply's mans switch failed "ON" a few months ago and I replaced it. It's a minature 5A toggle. After about a month the replacement failed. Getting suspicious I fitted a double pole switch with a 1R in series with each pole. I think the flash-bang was the resistors. The G-series PSUs have RIFA's built in and I think the X cap is breaking down and self healing at switch on. No smoke yet thankfully. I'll investigate that later.
So stuck the inverter on a fixed 28V DC supply and at least that is working
The 2N6338 is a reasonable upgrade from the 2N3772. More than twice the Vceo with lower saturation voltage, higher gain and IC. The 30 odd percent lower saturation voltage will reduce the dissipation quite a bit too.