wrote:
> Paul,
>
> Thanks for the response. I have taken the board out of the case and
> plugged in the PSU (a different one from the one that was in the case)
> and the fans still run as before. The MB is currently sitting on a
> non-conductive surface, so I don't see how anything could be shorting
> out underneath. Should I just toss the board or is there anything I
> can do to make it run correctly?
>
If the power supply is sitting on its own, not plugged into any hardware,
and you plug it in and turn it on, does it stay off ? That proves that
the PS_ON# signal is properly pulled high by the power supply. Without
a motherboard to pull that signal to ground, the power supply fan should
not start to run, on its own.
The following is included to give some history, and an example of how
it used to work. In fact, there is probably little you can do to trace
or debug it down further, due to the level of integration on the motherboard,
and lack of components like the older motherboards used.
********* PS_ON# and friends ... *********
To see how the driver scheme used to work for PS_ON#, there is a reference
schematic here. On PDF page 18, a signal called SUSC# on the 82371EB Southbridge,
drives an inverter and a 74F07 (upper right corner of the page). The output
of the 74F07 is "B_SUSC", english equivalent being "buffered SUSC". That signal
goes to PDF page 32 of the schematic, and connects to pin 14 of the main power
connector. Pin 14 is the PS_ON# signal, meaning the output of the 74F07 open
collector driver, is what drives the power supply PS_ON#.
http://www.intel.com/design/chipsets...x/BXDPDG10.PDF
To verify what the motherboard is doing, you would take a multimeter
and probe on either side of the 74F07 (marked as pin 1 and pin 2 on the
schematic). When there is a 0 volt level on the input of the 74F07,
there would be 0 volts on the output, which would turn on the supply.
In particular, when the computer "shuts down" or is supposed to, you'd
want to examine the voltage level on the input to the 74F07.
(A 74F07 datasheet, doesn't explain much..., just so you can see one)
http://hep.physics.lsa.umich.edu/alpha/images/74f07.pdf
The reason for looking on either side of the 74F07, is to see if the
motherboard is "trying" to shut off the PSU or not. If the input to the
74F07 is high, and yet the output is low, it means something is shorting
the output, or the 74F07 is dead. The 74F07, being open collector,
actually allows more than one 74F07 output to be connected together,
and that is referred to as wired-OR logic. But in this case, the open
collector has the advantage, that a person testing the motherboard,
can actually force the motherboard on (by grounding pin 14), without
damage to the 74F07. If a non open collector driver was used, it could
be damaged by grounding. But the open collector logic does not allow
forcing the output high (which is what you'd want right now, to test
whether the supply can be forced off or not). Another benefit of the
74F07, is if the PS_ON# signal is damaged, the 74F07 is cheap to replace,
which might be an advantage in a warranty repair situation. When they
were used, they used to cost about $0.25.
Modern designs could well be more economical, in that they probably
no longer use external logic or buffers at all. It could be that the
Southbridge drives PS_ON# directly, and the lack of a buffer like the
74F07, makes it harder to "see" what the motherboard is trying to do.
The signal to be probed to see if the motherboard is "trying" to switch
off, could be inside the Southbridge.
At the simplest level, the Southbridge contains a small circuit, that
converts the momentary pulse from the front panel "ON" button, into a
steady level for PS_ON#. But in terms of other part of power management,
both the Southbridge and the SuperI/O chip, contain agents that have
"wake" capability, such as wake on LAN, wake on ring, etc. So both
the Southbridge and the SuperI/O have to be connected together in some
way, to "wake" the computer and influence the state of PS_ON#. But
even if the logic was exposed so a person could check the logic levels,
there is no guarantee you could see it screwing up.
**********
At this point, you're probably looking at a replacement board. I wouldn't
throw it out, until you're certain that the motherboard is at fault.
In other words, if a replacement works properly, and for a few weeks,
then I might toss the other one. If the replacement board misbehaves,
it could turn out that something else is doing it, but I don't know
what. Something shorting the PS_ON# is about all I can think of, if
the PSU behaves when it is by itself and not connected to anything.
Paul
Similar Threads
Linear Mode
