I traced out scheamtics of the PSUs in all my
IBMs...
Great work ! Normally I trace out schematics only when I need to repaire its
and I cannot find any schematics on the net.
Oh, I like to understand how things work _while they're still working_.
Then if I need to make any measurements on th working device (even
something as trivial as determining if a house-coded transsitor is NPN or
PNP), I can do it knowing the componnent is good.
Of course this is not always possible (e.g. when I'm given a non-working
machine).
How quickly do you need this? I am wondering if
handing a copy to a
friend at an HPCC meeting in a couple of weeks time would be soon enough.
No problem for the time, it's a hobby (great but hobby) non work :)
OK, pester me a little before the meeting (12th Nov) and I'll take a copy
along.
Yes, and the control chip is on the
'output' side of the PSU, the drive
to the bases of the chopper transistors is transformer-coupled. There are
3 little daughterboards in there.
I have take a look at the circuit. It' similar to some off line converters
and other supplies, with control on the secondary side and not in the "live"
part of the switch mode.
Yes, there's nothing particularly odd about it.
One is an inrush limiter and contains a
relay that shorts out a power resistor once the smoothing capacitors have
charged.
This work, I've tested. The 4 ohm resistor is shorted correctly by the relay
OK.
Another contains the overcurrent protection
circuit, and, IIRC,
a 12V regulator. The last contains the overvoltage protection circuit
(the chips on that board are LM339 quad comparators).
I'm thinking about some protection circuit, because the switching try to
startup, for a time less then half second, and then stop. Now I can try to
There are 2 protection circuits. One is a current sensor, there's a
current transformer in series with the chopper transformer primary
winding. The other is a set of voltage sensors.
Whatever you do, don't disable any of the trips. If there really is
something wrong, the results would be spectacular!.
check il any protection become active during startup.
The SG3524 il correctly powered and try to fire up the switching
transistors, via the pulse transformer, but only for a short time.
Do you get any voltage appearing on any of the outputs? Even for that
half-second?
-tony