On 01/26/2018 07:15 PM, Jules Richardson via cctalk wrote:
After that... well, I need to work out how to get the entire supply
assembly to start up while it's open, so that I can work on it outside the
machine - then I can at least start looking for differences between the two
supplies (boy, are those things complex, but maybe I'll get lucky and it's
a bad solder joint or cap somewhere)
OK, now I feel like an idiot :-)
I started looking into this, and found that neither PSU would do anything
without being plugged into the (currently unpopulated) backplane. My
initial assumption was that there was some interaction between the
backplane and the "other" wires between it and the PSUs (i.e. the ones that
don't carry +5V or +12V).
However, on a whim (and I really don't know why I did this) I plugged the
+5V PSU connectors up to the backplane, but left the others (carrying +12V
and the other signals) unplugged. Doing that, things behaved as before -
the top PSU started operating normally, and the lower PSU started putting
out around 2.5V.
It got me wondering though - maybe it was simply a load issue, and with my
dummy loads I was right at the (lower) limit for the PSU regulation to
work; perhaps I had just enough for the top PSU to operate normally (but
only with the backplane hooked up too) and the lower PSU (for whatever
reason) wanted just that little bit more than the top one.
Anyway, I tweaked my loads to draw a little more current - and bingo! I'm
now getting stable +5V and +12V outputs on both PSUs, a working card cage
fan, and the DC OK light comes on.
So, in summary... with no load at all, the PSUs do nothing, but with *some*
load they either work OK, or the regulation is goofy and the outputs are low.
Time to add CPU/memory in, wire up a console cable and see what happens, I
guess.
cheers
Jules
