• Please review our updated Terms and Rules here

PS/1 2011 DC/DC converter board mystery...

whaka

Experienced Member
Joined
Feb 26, 2021
Messages
90
Location
France
Hello,

i see something strange with the motherboard DC/DC converter.
while checking voltages, i have somewhat low +/- 12V rails. i get 11.80 and -11.65V. thoses are usually around 11.90.

so, the mystery is here : if i bend enough the converter pcb in a way, all voltages raise (5, +/-12 they all raise ) and i can get a really strong 12V. if i bend it in reverse way, voltages fall...
it's not a connetor issue, i'm 100% sure. i tried to look, but i see nothing wrong. if someone have some ideas on what could happen ?
i thinked about bad ground vias, but i checked them, and they're good.

here is a little video of what happen : https://youtu.be/a5ZuQN6s7O0
if i can use my 2 hands for bending, i can get above 12V. i really don't understand what happen here, if it's an impedance issue, contact or whatever...

thanks
 
i double checked for the connector, on both board. the issue was clearly not here. voltages were low at the converter output, before the connector.
and btw, the mystery probably never be solved... the converter has just died between my last message and now...

well, i guess i'm good for build a replacement, or maybe find another if i'm lucky enough.
but as the mosfet are glued, i really can't fix it :(
 
Having died permanently will actually make it easier to fix. You're no longer looking for something that changes resistance depending on board deflection, now you're looking for a plain broken connection.

You're probably looking for a broken trace or bad solder in the feedback path from the output to the switching regulator.
 
i'll refresh joints on the converter pcb connector, but i doubt.
for now, the 2 feedback traces continuity are ok at least from the connector till the glued part. but of course, i can't check under the glued circuitery.
feedback come from the motherboard, i guess. and now you speak of that, i remember the 12V feedback was 11.23V
 
nothing better.
i tried to power the motherboard with an atx psu providing the 5 and +/-12, and it doesn't boot. the fan spin, i just get a blink on the 3 leds keyboard, and that's all.

there's a mysterious pin i don't know if it's an input or output. i supposed it was the 12V feedback (the one i got 11.23V when it worked), but it might not be a feedback as i got nothing on it when using the atx psu. it's not connected directly to the 12V rail for sure.
so, i don't know if i dare pluging a 12V here or not... if it's a feedback and got nothing, maybe that's why the original converter doesn't start.

and the pin i was considering as a 5V feedback is probably a power good. as i effectively got 5V on it when powered with the atx psu.
it's a pain without schematics :D
 
ok, it's a bit more clear now. the mysterious pin is in fact connected to pin 42 of the acu connector.
it's the reserved IRQ12 named FAULT by ibm. the fact is i still don't know if it's an input or output.
i guess an output from the DC/DC converter, and that's maybe why it doesn't boot with the atx psu. but that's weird, 11.23V on an irq line... what do you think ?
i don't know how an isa card can react if IRQ12 is wired on it... i'm lost on that point :D

edit : maybe they simply not wired it on the isa port of their acu... and make me seriously think i should certainly remove it from my acu project as i wired it.
the documentation is not very clear on this particular point. they say it's reserved, they maybe better saying it's not used on the isa port himself.
and considering that, the motherboard should certainly boot without it as it used for the acu. so, my motherboard is probably faulty...
 
Last edited:
and yes... the technical reference answered all my questions, shame i didn't look in it before :D
now it boot with the atx psu, he need the power good. so now, i'm really sure it's only the DC converter.
 
Back
Top