So... is this Athlon XP fried or not?

[J.Clarke]

On 2 Oct 2003 12:45:47 -0700

Hi,
I posted a question about problems I was having with an Athlon-XP
based system yesterday. Now I'm getting round to thinking the CPU is
fried.
Background: Athlon XP 1800+ system w/ Gigabyte mobo.

(1) When switched on, the power fan starts momentarily, and the
"RAM_LED" LED flickers. Then the power cuts out. (However, sometimes
nothing happens at all on power-up.)
(2) If I connect the (Athlon-friendly) power supply from my
Pentium-4 machine instead, everything starts up *without* the front
power switch being pushed, RAM_LED flickers continuously, and there's
a slightly irregular noise like the POST RAM-test (or maybe not).

This is strange, unless the CPU *and* the original power-supply
both have different faults. Case (1) sounds suspiciously similar to a
few other 'fried-Athlon' stories I've googled.

Importantly, the heatsink is attached (correctly, I assume, the
system has worked for a year), but the cpu fan DOES NOT WORK! Could
the power be cutting out when the fan fails? If so, why the different
behaviour with my P4 power supply?

At first I thought I might have damaged the fan when poking
components to make sure they were in place. Basically, it looked
squint, and had slight friction at one point when pushing it round.
But the machine isn't mine, and the main fault existed before I looked
at it. My feeling is that the fan failed and the CPU overheated.

What do you think? Any feedback appreciated, thanks.

- Michael Strorm
(e-mail address removed)

Are you completely certain that there is nothing shorting out the
motherboard? Could be a screw lying on the board, or could be an extra
standoff in the wrong place if this is a newly constructed system. What
you're describing sounds an awful lot a piece of metal lying on the
board that got moved when you changed the power supply.

 

[Michael Strorm]

Hi,
I posted a question about problems I was having with an Athlon-XP
based system yesterday. Now I'm getting round to thinking the CPU is
fried.
Background: Athlon XP 1800+ system w/ Gigabyte mobo.

(1) When switched on, the power fan starts momentarily, and the
"RAM_LED" LED flickers. Then the power cuts out. (However, sometimes
nothing happens at all on power-up.)
(2) If I connect the (Athlon-friendly) power supply from my
Pentium-4 machine instead, everything starts up *without* the front
power switch being pushed, RAM_LED flickers continuously, and there's
a slightly irregular noise like the POST RAM-test (or maybe not).

This is strange, unless the CPU *and* the original power-supply
both have different faults. Case (1) sounds suspiciously similar to a
few other 'fried-Athlon' stories I've googled.

Importantly, the heatsink is attached (correctly, I assume, the
system has worked for a year), but the cpu fan DOES NOT WORK! Could
the power be cutting out when the fan fails? If so, why the different
behaviour with my P4 power supply?

At first I thought I might have damaged the fan when poking
components to make sure they were in place. Basically, it looked
squint, and had slight friction at one point when pushing it round.
But the machine isn't mine, and the main fault existed before I looked
at it. My feeling is that the fan failed and the CPU overheated.

What do you think? Any feedback appreciated, thanks.

- Michael Strorm
(e-mail address removed)

 

[rstlne]

Hi,
I posted a question about problems I was having with an Athlon-XP
based system yesterday. Now I'm getting round to thinking the CPU is
fried.
Background: Athlon XP 1800+ system w/ Gigabyte mobo.

(1) When switched on, the power fan starts momentarily, and the
"RAM_LED" LED flickers. Then the power cuts out. (However, sometimes
nothing happens at all on power-up.)
(2) If I connect the (Athlon-friendly) power supply from my
Pentium-4 machine instead, everything starts up *without* the front
power switch being pushed, RAM_LED flickers continuously, and there's
a slightly irregular noise like the POST RAM-test (or maybe not).

This is strange, unless the CPU *and* the original power-supply
both have different faults. Case (1) sounds suspiciously similar to a
few other 'fried-Athlon' stories I've googled.

Importantly, the heatsink is attached (correctly, I assume, the
system has worked for a year), but the cpu fan DOES NOT WORK! Could
the power be cutting out when the fan fails? If so, why the different
behaviour with my P4 power supply?

At first I thought I might have damaged the fan when poking
components to make sure they were in place. Basically, it looked
squint, and had slight friction at one point when pushing it round.
But the machine isn't mine, and the main fault existed before I looked
at it. My feeling is that the fan failed and the CPU overheated.

What do you think? Any feedback appreciated, thanks.

- Michael Strorm
(e-mail address removed)

I guess before anything I would say to reset the bios (read the book) then
see if it turns on..

If your connecting 2 different power supplys and one is "Starting" up and
the other isnt
then that right there says to me that one of the supplys is bad..

the CPU Fan should turn as soon as everything tries to cut on..
Did you try spinning the fan by hand to see if it's verry loose? It's
possible that the board is shutting down due to the fan RPM being too low..

I would remove EVERYTHING from the board, Ram, Cpu, all cards, all IO
connections and all fans and then use the power supplys and see if they both
do the same thing, if they dont then I would first say that the psu is bad
... Asto why it doenst "instantly" cut on.. I dont know but I could make a
blind assumption that the "Restart after power loss = true" and the +5v VSB
isnt feeding to the motherboard, so the good p4 psu cuts it on instantly (5v
feeds the mobo, mobo see's the power being avail, throws the switch and
request power).. This could also explain why both systems give different
faults when started..

Okay, so if you do what I said and see that it is indeed different then that
will tell you that one supply is "probably" faulty.. If you have a DVM you
can check the +5v VSB line (say i fyou want to do this and someone/i'll tell
you what pins).. So assuming that the one power supply is bad, the p4 one
will give you an irregular noise..
Noises are important, are they beeps?.. What are they..
I say put back the processor and ram (if more than one stick then just use 1
stick) with the p4 suppy and try it..
First things first.. Does the cpu fan spin.. If no then that can cause
problems..(so keep it in mind) but the system still should boot so you
should get at least 20 or 30 seconds before thermal protection kicks in (if
the giga board has protection, if not then I would probably sort the fan
first, give it a whirl with ur finger and see if it starts).. Look up the
error codes and see what they say.. if ram is bad then try another stick..

a mobo shouldnt give you an error code other than "processor faulty" if the
"processor is faulty" because that's the first thing that's tested and if it
test bad then it's all halted (but not power'd off)..

I have 2 bad athlon Xp's (Pal 1700+ and Pal 1800+) and they would both
actually get hot!.. I tried them in 2 different mobo's, one of them just
shows bad processor for both chips, the other one will shut down with one
chip, and just show bad cpu with the other.. so bad chips can show anything

 

[Ben Pope]

Hi,
I posted a question about problems I was having with an Athlon-XP
based system yesterday. Now I'm getting round to thinking the CPU is
fried.
Background: Athlon XP 1800+ system w/ Gigabyte mobo.

(1) When switched on, the power fan starts momentarily, and the
"RAM_LED" LED flickers. Then the power cuts out. (However, sometimes
nothing happens at all on power-up.)

How long does it take to cut out?

(2) If I connect the (Athlon-friendly) power supply from my
Pentium-4 machine instead, everything starts up *without* the front
power switch being pushed, RAM_LED flickers continuously, and there's
a slightly irregular noise like the POST RAM-test (or maybe not).

I'm suspecting a sticky power switch or a short in it.

That would explain the first PSU starting then stopping... like holding the
power button, yeah?

It would also explain the reason for the second PSU starting up without the
button being pressed.

Why one stops and the other doesn't could be down to different ways in which
they deal with forced powering down.

At this stage that doesn't really explain the flickery RAM light or
irregular noise.

Try swapping the reset button and power switch on the header and using the
reset button on the front to power up, that'll eliminate the sticky
button/shorted lead guess.

It would also be beneficial to play around with the ram - remove all but one
stick and see if you still get the flickery LED, if so try a different slot,
preferably in a different bank. If you still get the flickerey LED, try
another stick.

Ben

 

[Kent_Diego]

I just had an old computer die and the CPU fan did not spin. Turned out to
be shorted hard drive. The 12 Volt is used for CPU fan. Unplug and remove
everything but video card and see if works.

-Kent

 

[Michael Strorm]

I guess before anything I would say to reset the bios (read the book) then
see if it turns on..

I don't have the BIOS book (the computer isn't mine)- I could get it,
but the question is how much time (and risk) am I willing to take on
someone else's computer...

If your connecting 2 different power supplys and one is "Starting" up and
the other isnt
then that right there says to me that one of the supplys is bad..

And something strange is happening with the other one since it
switches on when it *shouldn't*!

the CPU Fan should turn as soon as everything tries to cut on..
Did you try spinning the fan by hand to see if it's verry loose? It's
possible that the board is shutting down due to the fan RPM being too low..

The fan is *not* loose... it's slightly sticky, and now it's even more
f*****, as I tried to fix it (note: it didn't work at all before!)

I will nick the one off my Pentium-233 to see if that works (even if
it's insufficient for an Athlon, the heatsink doesn't get hot within
the few seconds I've had the computer on so far, implying that so long
as the heatsink is on, the Athlon won't fry without the fan- for a few
seconds anyway- and if it did, then the CPU is already fried from a
faulty fan).

I would remove EVERYTHING from the board, Ram, Cpu, all cards, all IO
connections and all fans and then use the power supplys and see if they both
do the same thing, if they dont then I would first say that the psu is bad
.. Asto why it doenst "instantly" cut on.. I dont know but I could make a
blind assumption that the "Restart after power loss = true" and the +5v VSB
isnt feeding to the motherboard, so the good p4 psu cuts it on instantly (5v
feeds the mobo, mobo see's the power being avail, throws the switch and
request power).. This could also explain why both systems give different
faults when started..

It might. I'm not a hardware expert (I built the P4 system myself with
no major problems, but that's my widest experience really...). I'll
definitely consider it though.

Okay, so if you do what I said and see that it is indeed different then that
will tell you that one supply is "probably" faulty.. If you have a DVM you
can check the +5v VSB line (say i fyou want to do this and someone/i'll tell
you what pins).. So assuming that the one power supply is bad, the p4 one
will give you an irregular noise..

I tried a multimeter, and only got a reading from the blue line
(0.28V), but AFAIK, PSUs don't kick into action fully if some mobo
components are missing (IIRC).

Noises are important, are they beeps?.. What are they..

No beeps, just rapid clicks, like POST RAM-check, but slightly more
irregular and they don't stop after a while.

I say put back the processor and ram (if more than one stick then just use 1
stick) with the p4 suppy and try it..
First things first.. Does the cpu fan spin.. If no then that can cause
problems..(so keep it in mind) but the system still should boot so you
should get at least 20 or 30 seconds before thermal protection kicks in (if
the giga board has protection, if not then I would probably sort the fan
first, give it a whirl with ur finger and see if it starts).. Look up the
error codes and see what they say.. if ram is bad then try another stick..

a mobo shouldnt give you an error code other than "processor faulty" if the
"processor is faulty" because that's the first thing that's tested and if it
test bad then it's all halted (but not power'd off)..

I have 2 bad athlon Xp's (Pal 1700+ and Pal 1800+) and they would both
actually get hot!.. I tried them in 2 different mobo's, one of them just
shows bad processor for both chips, the other one will shut down with one
chip, and just show bad cpu with the other.. so bad chips can show anything

Thanks for this info. To tell the truth, I'm going to try some of the
obvious things (replacing the fan is easy and should work so long as
the mobo isn't expecting fan feedback which my old P233 one might not
supply), but... I think I'll have to draw the line somewhere.

Fortunately, I got the important data of the machine's hard drive, so
the mailing lists etc should be okay even if everything else is 'cked.

- Michael S

 

[Michael Strorm]

How long does it take to cut out?

No more than half a second.

I'm suspecting a sticky power switch or a short in it.
That would explain the first PSU starting then stopping... like holding the
power button, yeah?

I thought that too... but I disconnected the power switch altogether
and it makes no difference.

Why one stops and the other doesn't could be down to different ways in which
they deal with forced powering down.

At this stage that doesn't really explain the flickery RAM light or
irregular noise.

Try swapping the reset button and power switch on the header and using the
reset button on the front to power up, that'll eliminate the sticky
button/shorted lead guess.

I might still try that, though I doubt it's the problem now.

It would also be beneficial to play around with the ram - remove all but one
stick and see if you still get the flickery LED, if so try a different slot,
preferably in a different bank. If you still get the flickerey LED, try
another stick.

Worth a check....
Thanks!

- Michael S

 

[Michael Strorm]

Are you completely certain that there is nothing shorting out the
motherboard? Could be a screw lying on the board, or could be an extra
standoff in the wrong place if this is a newly constructed system. What
you're describing sounds an awful lot a piece of metal lying on the
board that got moved when you changed the power supply.

The standoff theory sounded plausible, except that the system was in
use for over a year before it failed.

My gut reaction was 'possible short-circuit' too, so I took the mobo
out, checked it, and replaced it, with no noticable effect.

- Michael S

 

[Michael Strorm]

I just had an old computer die and the CPU fan did not spin. Turned out to
be shorted hard drive. The 12 Volt is used for CPU fan. Unplug and remove
everything but video card and see if works.

The HDD is okay- I copied data off it. IIRC I had the drives
disconnected when I tested the machine at one stage and it made no
difference (though I'll recheck that).

- Michael S

 

[rstlne]

I don't have the BIOS book (the computer isn't mine)- I could get it,

but the question is how much time (and risk) am I willing to take on
someone else's computer...

There is little/no risk to resetting the bios, turn power off, remove
battery, move the jumper.. should work on any motherbaord..

And something strange is happening with the other one since it
switches on when it *shouldn't*!

Some mobo's will switch themselves back on after a power loss.. So "if" the
switch hasnt been hit and the power has gone off for another reason then it
can try to turn on.. if THIS is the case then only the good power supply
will cause this..

The fan is *not* loose... it's slightly sticky, and now it's even more
f*****, as I tried to fix it (note: it didn't work at all before!)

yea they are hard to fix

I will nick the one off my Pentium-233 to see if that works (even if
it's insufficient for an Athlon, the heatsink doesn't get hot within
the few seconds I've had the computer on so far, implying that so long
as the heatsink is on, the Athlon won't fry without the fan- for a few
seconds anyway- and if it did, then the CPU is already fried from a
faulty fan).

Yea any fan will work.. as long as the heatsink is on properly then you'll
probably be okay anyhow..

I tried a multimeter, and only got a reading from the blue line
(0.28V), but AFAIK, PSUs don't kick into action fully if some mobo
components are missing (IIRC).

Blue is the -12v line and it wont come on until power is applied ..
Purple is the +5v VSB line and it should allways give 5v regardless of the
psu being on or off..

No beeps, just rapid clicks, like POST RAM-check, but slightly more
irregular and they don't stop after a while.

Does it sound like "click click click click click click click"
electronic componets that go bad will sometimes give this type of noise..

 

[Michael Strorm]

Does it sound like "click click click click click click click"
electronic componets that go bad will sometimes give this type of noise..

Nah, it doesn't sound like "when good components go bad", the period
between clicks is slightly irregular, but the clicks themselves sound
fine.

- Michael S

 

[Spiffy]

The purpose of the RAM_LED on Gigabyte mobos is to indicate the status of
the memory. If the LED stays continuously lit, the ram is operating as it
should. flashing or flickering RAM_LED indicates a potential problem with
your memory. if this lite is flashing/flickering i would swap out the ram
firstly. hope this helps.