MOBO power light on but comp not working? cpu?

  • Thread starter Thread starter Joel
  • Start date Start date
J

Joel

http://tinyurl.com/dczbv


I just got that above mobo from newegg to replace an old one that stopped
working. I had narrowed the problem down to the mobo or cpu. The green power
light on both mobo's comes on when I hit the power switch, but nothing
works . . . even the chassy fan. So I have to assume I got a bad cpu, right?
The psu works in another comp. So does the nickel battery (cmos).

So I guess I will order a cpu. Is this cpu the right one for the MOBO I
linked above? It looks right.

http://tinyurl.com/2cd49q

Thanks!!!
 
Joel said:
http://tinyurl.com/dczbv


I just got that above mobo from newegg to replace an old one that stopped
working. I had narrowed the problem down to the mobo or cpu. The green power
light on both mobo's comes on when I hit the power switch, but nothing
works . . . even the chassy fan. So I have to assume I got a bad cpu, right?
The psu works in another comp. So does the nickel battery (cmos).

So I guess I will order a cpu. Is this cpu the right one for the MOBO I
linked above? It looks right.

http://tinyurl.com/2cd49q

Thanks!!!

No, if the chassis fan doesn't work, you have no 12V output. If you
have no 12V, there is no way a CPU can run. (The CPU connector is
12V.)

I would start with the basics. You say the green LED on the surface
of the motherboard works. That means you have +5VSB. The +5VSB is needed
to power the chip that listens to the power switch on the front of
the chassis. Now, the motherboatd sends a signal called PS_ON# to
the power supply. The signal is inverted - with a multimeter, PS_ON#
will read 5 volts when the computer is off. When you push the
button on the front, what is supposed to happen, is PS_ON# drops to
pretty close to 0 volts. A good value is about 0.4V, and the spec says
the signal has to be lower than 0.8V to turn the computer on.

So you have to determine what is going on with +12V first. It could
be you have a bad power supply. By checking the PS_ON# signal, you can
determine if it is the motherboard's fault (PS_ON# not going to 0 volts).
If PS_ON# is zero volts, and the fans don't spin, then it is the PSU's
fault.

Once the fan is running, and if the thing still won't POST, then
you can think about buying another CPU. But at the moment, I'm not
sure your power supply is working right. Prove there is 12V available
first, before you panic.

Here are three power supply specs. The first two are for 20 pin power
supplies. The first one is for really old power supplies that still
have -5V on pin 18. The third spec is for newer supplies that have a
24 pin main connector. To orient yourself, the three black wires next
to one another, are the three COM pins. PS_ON# is right next to them.

http://web.archive.org/web/20030424...org/developer/specs/atx/ATX_ATX12V_PS_1_1.pdf
http://www.formfactors.org/developer/specs/atx/ATX12V_1_3dg.pdf

http://www.formfactors.org/developer/specs/ATX12V_PSDG_2_2_public_br2.pdf

If you don't own a multimeter, the next thing to swap is the PSU.
And don't go any further until you get some fans running.

Paul
 
http://tinyurl.com/dczbv


I just got that above mobo from newegg to replace an old one that stopped
working. I had narrowed the problem down to the mobo or cpu. The green power
light on both mobo's comes on when I hit the power switch, but nothing
works . . . even the chassy fan. So I have to assume I got a bad cpu, right?

Unless the heatsink fan had stopped working or the heatsink
was off (fell off, since prior system build had worked then
stopped working, right?) it is not likely to be the CPU.

Perhaps now would be a good time to back up and briefly
describe all major components in the system, including PSU
make/model/wattage and rated 12V current.

The psu works in another comp. So does the nickel battery (cmos).

Having the PSU work in another system is not necessarily a
sign it is "good" enough or working properly for this
system. What was this other system like?
CPU/motherboard/video type? It is especially true if a
generic PSU or with limited 12V current available.
Similarly so with 5VSB current, if the other system didn't
have PS2 or USB ports jumpered to 5VSB (versus 5V), and
present system does and has significant PS2 or USB
peripherals plugged in, it could interfere with system
startup.

I suggest you strip the system down to bare essentials.
Only CPU, heatsink/fan, 1 memory module, and the least power
hungry video card you have. An old PCI video card might be
ideal, nevermind if it's too ancient to use the system as
intended, at this point in troubleshooting. Leave
keyboard/mouse/printer/drives/etc all unplugged from power
and data cables unplugged from motherboard.

Clear CMOS and retry system. Also check behind the battery
for a plastic insulator to preserve it's charge. Also
recheck clear CMOS jumper position in case they had left it
in the clear position. The motherboard manual will detail
this or there should be a silkscreen on the board with Pin 1
designated (but might be too hard to see with the board in
the case so the manual may be most helpful).

If the board has a jumper to set bus speed, jumper it to the
slowest speed temporarily.


As Paul had mentioned, until you get the fans working it
looks like the PSU is iffy. You might unplug the PSU from
the board and jumper the PS_On line to ground with a
paperclip to see if the PSU then comes on. You might be
unlucky and have received a bad new motherboard but it seems
more likely the PSU is the problem. If there are any
questions as to whether there might be any standoffs in the
case shorting out the new board, take it out of the case and
test on a non-conductive surface (and not on ESD discharge
protective packaging or foam). A couple of magazines on a
desktop to elevate the board enough that the video card
bracket can overhang the edge of the magazines, works
nicely.

You might also take the CPU out and see if you get a
different response. It's not impossible to have a CPU fail,
just very unlikely. However with the CPU out you also
greatly decrease 12V current required so it might be that
the PSU had been shutting down (so fast it never even
appeared to start) when it couldn't keep rail voltages
within tolerance but by removing most of the 12V load it
might be able to do it. Unfortunately if this is how it
responds, it does not narrow down whether the CPU or PSU is
at fault, but along with other evidence collecting could be
useful to know.
 
Joel said:
http://tinyurl.com/dczbv


I just got that above mobo from newegg to replace an old one that stopped
working. I had narrowed the problem down to the mobo or cpu. The green power
light on both mobo's comes on when I hit the power switch, but nothing
works . . . even the chassy fan. So I have to assume I got a bad cpu, right?
The psu works in another comp. So does the nickel battery (cmos).

So I guess I will order a cpu. Is this cpu the right one for the MOBO I
linked above? It looks right.

http://tinyurl.com/2cd49q

Thanks!!!

Ever thought of the PSU? Is it Pentium-4 capable, esp. if the prior
motherboard was not for a Pentium 4 CPU?
 
Hey Joel,

The same thing happened to me last month. I tried everything to figure out
what was wrong and in the end had to put it in for repair. It turned out
that the motherboard had died. I know you have had this happen with two
boards which makes this seem unlikely but you may well be that unlucky.

Cheers.

Si
 
Back
Top