bust processor ?

Hi all,

I was given a "KM18G Pro" motherboard with an
AthlonXP 2500 socketA (or socket462) processor
the other day, being told that the mobo was shot and
the processor OK.

On strength of that (and having no means to test either
independantly) I purchased a "KM400A + 8237" board
and intended to upgrade my son's PII machine.

420 bucks later I powered it up with the following
known-to-work items:

512 DDR333 memory module,
300W PSU
6gig HDD

.....and nada. The machine turns on and then off again
after a few seconds when jumpered to 166 FSB and
turns on only when jumpered to 133 FSB. In both cases
I get no signal to the monitor and (thus?) no post. The
HDD spins up and CPU fan turns on. Appropos CPU
fan - I'm not entirely sure how critical this is to post. It's
a new fan but of PII or III calibre - if you know what I
mean. Also, the manual-stated pinout for the CPU fan
seems to be incorrect: with +12v in the middle, but the
ground and signal wires are reversed. I doubt that bears
any significance though.

So, my question is the obvious - can I be sure the processor
is indeed buggered ? What are the typical symptoms of a
shot processor ?

TIA
Jona.

On Sat, 17 Sep 2005 10:03:08 +0200, "jona"
wrote:

Hi all,

I was given a "KM18G Pro" motherboard with an
AthlonXP 2500 socketA (or socket462) processor
the other day, being told that the mobo was shot and
the processor OK.

On strength of that (and having no means to test either
independantly) I purchased a "KM400A + 8237" board
and intended to upgrade my son's PII machine.

420 bucks later I powered it up with the following
known-to-work items:

How did you spend 420 if you already had the CPU, bought a
cheap mobo, PSU, tiny drive and only 512MB of memory?
Unless you bought a $300 video card, something is amiss.



512 DDR333 memory module,
300W PSU
6gig HDD

Is the PSU name-brand with at least 180, preferribly 200W of
combined 3V/5V power rating? If not, it is suspect.



....and nada. The machine turns on and then off again
after a few seconds when jumpered to 166 FSB and
turns on only when jumpered to 133 FSB.
In both cases
I get no signal to the monitor and (thus?) no post. The
HDD spins up and CPU fan turns on. Appropos CPU
fan - I'm not entirely sure how critical this is to post. It's
a new fan but of PII or III calibre - if you know what I
mean. Also, the manual-stated pinout for the CPU fan
seems to be incorrect: with +12v in the middle, but the
ground and signal wires are reversed. I doubt that bears
any significance though.


If the fan spins, you've hooked it up correctly.

If this heatsink is smaller, as many from the P2 or P3 era
were, you may need a bigger/better heatsink. That's not the
current problem though, so long as you have it installed
well (such that it's making good contact with the CPU,
double-check on that) the CPU certainly won't have
overheated in the first few seconds of turning the system on
from being cold. Some systems may turn off if the bios is
set to expect a fan RPM signal, but this is generally after
it has POSTed, and you'd have the option to disable such a
setting.

So, my question is the obvious - can I be sure the processor
is indeed buggered ? What are the typical symptoms of a
shot processor ?

No, it is not likely that the processor is bad in general,
not from your described symptoms. It may be possible
considering the state of the other motherboard, though, or
if you have had the CPU powered on without the heatsink on
it good.

I'd first suspect an inadequate power supply, and suggest
taking voltage readings with a multimeter and/or swapping in
another known good/adequate PSU. You might also recheck all
cards, cables, etc. Then you could strip system down to
bare essentials leaving only CPU, heatsink/fan, video, 1
memory module... disconnecting everything else including
keyboard, mouse, cable wiring. Turn on the system by
shorting the two power-on motherboard pins with a
screwdriver tip.

Since it seems your PSU may be marginal, I suggest leaving
the board set to the slowest FSB setting until it works at
that speed, THEN trying it higher. Also substitute a
less-power-hungry video card if the present one is power
hungry... any old PCI or low-end AGP card should suffice.

Also try clearing CMOS (or pull battery for 10 minutes)
while AC power is disconnected, and check the battery
voltage.

"kony" wrote in message
...
On Sat, 17 Sep 2005 10:03:08 +0200, "jona"
wrote:

Hi all,

I was given a "KM18G Pro" motherboard with an
AthlonXP 2500 socketA (or socket462) processor
the other day, being told that the mobo was shot and
the processor OK.

On strength of that (and having no means to test either
independantly) I purchased a "KM400A + 8237" board
and intended to upgrade my son's PII machine.

420 bucks later I powered it up with the following
known-to-work items:

How did you spend 420 if you already had the CPU, bought a
cheap mobo, PSU, tiny drive and only 512MB of memory?
Unless you bought a $300 video card, something is amiss.

LOL.
No, that would be 420 of *my* bucks (Namibian $'s) for
the board only. That's approx 70 US$. All the other mentioned
parts are those I have for testing purposes and have been
used in various other setups.


512 DDR333 memory module,
300W PSU
6gig HDD

Is the PSU name-brand with at least 180, preferribly 200W of
combined 3V/5V power rating? If not, it is suspect.

It's "brand name" for around here. In fact that very PSU is
from my own machine when I replaced it with a 350W unit
after adding a video capture card. It's good.

....and nada. The machine turns on and then off again
after a few seconds when jumpered to 166 FSB and
turns on only when jumpered to 133 FSB.
In both cases
I get no signal to the monitor and (thus?) no post. The
HDD spins up and CPU fan turns on. Appropos CPU
fan - I'm not entirely sure how critical this is to post. It's
a new fan but of PII or III calibre - if you know what I
mean. Also, the manual-stated pinout for the CPU fan
seems to be incorrect: with +12v in the middle, but the
ground and signal wires are reversed. I doubt that bears
any significance though.

If the fan spins, you've hooked it up correctly.

Ja, this would not be the first manual I've seen with inacurate
diagrams.

If this heatsink is smaller, as many from the P2 or P3 era
were, you may need a bigger/better heatsink. That's not the
current problem though, so long as you have it installed
well (such that it's making good contact with the CPU,
double-check on that) the CPU certainly won't have
overheated in the first few seconds of turning the system on
from being cold. Some systems may turn off if the bios is
set to expect a fan RPM signal, but this is generally after
it has POSTed, and you'd have the option to disable such a
setting.

That answers my CPU fan doubt, thanks. I figured that
perhaps the board detected a too low rpm and thus shut off.

So, my question is the obvious - can I be sure the processor
is indeed buggered ? What are the typical symptoms of a
shot processor ?

No, it is not likely that the processor is bad in general,
not from your described symptoms. It may be possible
considering the state of the other motherboard, though, or
if you have had the CPU powered on without the heatsink on
it good.

I haven't, but maybe the previous "owner" has had the CPU
powered on without h/sink & fan.

I'd first suspect an inadequate power supply, and suggest
taking voltage readings with a multimeter and/or swapping in
another known good/adequate PSU.

Unless the fact that AMD processors require a significant
amount of power (they operate on 3.3v if I'm not mistaken),
then the PSU is good. It has hauled my own P4 with many
devices/peripherals for about a year. By that I mean one ATi
Radeon 9800 agp card, two additional pci network cards, pci
modem, pci IDE card and 5 IDE devices. When I added the
video capture card I decided not to tempt failure and upgraded
to a same-brand 350W PSU.

You might also recheck all
cards, cables, etc. Then you could strip system down to
bare essentials leaving only CPU, heatsink/fan, video, 1
memory module... disconnecting everything else including
keyboard, mouse, cable wiring. Turn on the system by
shorting the two power-on motherboard pins with a
screwdriver tip.

I have no cards or anything else - it's using (or attempting to)
the onboard vga and the 512 RAM is a single unit. In the meantime
I disconnected the keyboard, HDD and mouse as well - no go.
I also tried with a low-end pci vga card.


Since it seems your PSU may be marginal, I suggest leaving
the board set to the slowest FSB setting until it works at
that speed, THEN trying it higher. Also substitute a
less-power-hungry video card if the present one is power
hungry... any old PCI or low-end AGP card should suffice.

Also try clearing CMOS (or pull battery for 10 minutes)
while AC power is disconnected, and check the battery
voltage.

I've tried that and now I think I'll resort to ripping my sister's
AMD processor out of her machine just to get some sort of
clarity on the issue. It might just turn out that I now have two
perfectly good AMD socketA boards and no processors.

Thanks for your input.

On Sat, 17 Sep 2005 16:27:53 +0200, "jona"
wrote:


Is the PSU name-brand with at least 180, preferribly 200W of
combined 3V/5V power rating? If not, it is suspect.

It's "brand name" for around here. In fact that very PSU is
from my own machine when I replaced it with a 350W unit
after adding a video capture card. It's good.

I can presume that means it "works" in general, but we don't
know what your own machine was, whether that qualifies it
for this different system. Even similar system (total load)
can vary these days since some may draw a more balanced
amperage from both 5V and 12V rails, while older era Athlon
XP platforms do not. KM400 based boards were among the last
that, in general (most of them), didn't use 12V for CPU
power, rather 5V rail still.


That answers my CPU fan doubt, thanks. I figured that
perhaps the board detected a too low rpm and thus shut off.

It's possible still, though it should POST immedately when
you turn it on and then seconds later shut off. It can't do
that at first as the fans can't be presumed to be at max RPM
yet by the time the system (would have) already started
POSTing and displaying the video card then motherboard bios.


So, my question is the obvious - can I be sure the processor
is indeed buggered ? What are the typical symptoms of a
shot processor ?

No, it is not likely that the processor is bad in general,
not from your described symptoms. It may be possible
considering the state of the other motherboard, though, or
if you have had the CPU powered on without the heatsink on
it good.

I haven't, but maybe the previous "owner" has had the CPU
powered on without h/sink & fan.

Possible... When I come across a system dead from mysterious
causes, it's usually easiest to have an entire 2nd system so
each part can be swapped between the two. That is, unless
it's something obvious.



I'd first suspect an inadequate power supply, and suggest
taking voltage readings with a multimeter and/or swapping in
another known good/adequate PSU.

Unless the fact that AMD processors require a significant
amount of power (they operate on 3.3v if I'm not mistaken),
then the PSU is good.

Well yes they do require a significant amount, as do Intels.
The motherboard uses a step-down switching voltage regulator
powered by 5V from the PSU, on every KM400 board I've seen
and most likely yours as well.

It has hauled my own P4 with many
devices/peripherals for about a year. By that I mean one ATi
Radeon 9800 agp card, two additional pci network cards, pci
modem, pci IDE card and 5 IDE devices. When I added the
video capture card I decided not to tempt failure and upgraded
to a same-brand 350W PSU.

Another thought is that this PSU was a bit marginal
originally and has been stressed by running the described
parts, was about ready to fail either way. It is not
necessary the case but yet another reason I would take
voltage readings and try a higher capacity name-brand PSU.

I have no cards or anything else - it's using (or attempting to)
the onboard vga and the 512 RAM is a single unit. In the meantime
I disconnected the keyboard, HDD and mouse as well - no go.
I also tried with a low-end pci vga card.

Have you tried moving the memory to another slot, or using
different memory?



Since it seems your PSU may be marginal, I suggest leaving
the board set to the slowest FSB setting until it works at
that speed, THEN trying it higher. Also substitute a
less-power-hungry video card if the present one is power
hungry... any old PCI or low-end AGP card should suffice.

Also try clearing CMOS (or pull battery for 10 minutes)
while AC power is disconnected, and check the battery
voltage.

I've tried that and now I think I'll resort to ripping my sister's
AMD processor out of her machine just to get some sort of
clarity on the issue. It might just turn out that I now have two
perfectly good AMD socketA boards and no processors.


Perhaps, or it could be some bios quirk where IF you got it
running with a different CPU, updating the bios could
possibly help with the other CPU or memory compatibility.

I would tend to suspect even this new motherboard as much as
the CPU though, unless the prior owner had indeed done
something reckless with it.

"kony" wrote in message
...
On Sat, 17 Sep 2005 16:27:53 +0200, "jona"
wrote:
That answers my CPU fan doubt, thanks. I figured that
perhaps the board detected a too low rpm and thus shut off.

It's possible still, though it should POST immedately when
you turn it on and then seconds later shut off. It can't do
that at first as the fans can't be presumed to be at max RPM
yet by the time the system (would have) already started
POSTing and displaying the video card then motherboard bios.

[......]

I haven't, but maybe the previous "owner" has had the CPU
powered on without h/sink & fan.

Possible... When I come across a system dead from mysterious
causes, it's usually easiest to have an entire 2nd system so
each part can be swapped between the two.

Yes, that would make my current situation easier. Unfortunately
I don't have a second AMD setup and they seem to be scarce
around here.

[.....]

It has hauled my own P4 with many
devices/peripherals for about a year. By that I mean one ATi
Radeon 9800 agp card, two additional pci network cards, pci
modem, pci IDE card and 5 IDE devices. When I added the
video capture card I decided not to tempt failure and upgraded
to a same-brand 350W PSU.

Another thought is that this PSU was a bit marginal
originally and has been stressed by running the described
parts, was about ready to fail either way. It is not
necessary the case but yet another reason I would take
voltage readings and try a higher capacity name-brand PSU.

Have you tried moving the memory to another slot, or using
different memory?

Moved to a different slot and I temporarily used the PSU
and a RAM chip from my own machine. No go.

[.......]

I've tried that and now I think I'll resort to ripping my sister's
AMD processor out of her machine just to get some sort of
clarity on the issue. It might just turn out that I now have two
perfectly good AMD socketA boards and no processors.

Perhaps, or it could be some bios quirk where IF you got it
running with a different CPU, updating the bios could
possibly help with the other CPU or memory compatibility.

I tried my sister's Athlon-PECM, 807Mhz processor and again,
it won't even post. I'm not sure whether it should/would though.
I mean, that processor is "ancient" technology and certainly not
DDR compatible, but ..... ??

I would tend to suspect even this new motherboard as much as
the CPU though, unless the prior owner had indeed done
something reckless with it.

The only new development is that when testing now, or
with any other PSU (i tried two), the machine powers up
directly when connecting to the wall socket - whether I
have the switch connected to it's pinout or not. Still no
post though. I also tried the CPU in question on my sister's
board (Asus A7V) where it doesn't work either - but that
doesn't surprise me. BUT, under correction, her Athlon
should have worked on the new board, right ?

Further, the "new" board came all boxed with it's various
goodies, including IDE cable, manual, driver CD, etc -
but *without* the I/O backplate/panel. That almost suggests
that it has before been installed somewhere and returned
to the vendor .....

sigh

On Mon, 19 Sep 2005 16:27:51 +0200, "jona"
wrote:


Perhaps, or it could be some bios quirk where IF you got it
running with a different CPU, updating the bios could
possibly help with the other CPU or memory compatibility.

I tried my sister's Athlon-PECM, 807Mhz processor and again,
it won't even post. I'm not sure whether it should/would though.
I mean, that processor is "ancient" technology and certainly not
DDR compatible, but ..... ??

It should've worked, the memory controller is in the
northbridge so the CPU cares not what it is, and the FSB was
always DDR on a socket A athlon. It could be necessary to
clear CMOS before the old CPU would POST the system, but
beyond that it should've worked.


The only new development is that when testing now, or
with any other PSU (i tried two), the machine powers up
directly when connecting to the wall socket - whether I
have the switch connected to it's pinout or not. Still no
post though. I also tried the CPU in question on my sister's
board (Asus A7V) where it doesn't work either - but that
doesn't surprise me. BUT, under correction, her Athlon
should have worked on the new board, right ?

I would've expected it to. If this board is still in a box,
I'd try it out of the box on a non-conductive (not
anti-static material) surface and without the case wiring
connected. Sometimes it's the simple things like a stuck
case switch or a case standoff shorting the back of the
board that'll be the hardest things to find if you're not
looking for them.


Further, the "new" board came all boxed with it's various
goodies, including IDE cable, manual, driver CD, etc -
but *without* the I/O backplate/panel. That almost suggests
that it has before been installed somewhere and returned
to the vendor .....


If it uses the old standard ATX style of rear port
locations, like the following picture, it's fairly common
for the panel to not be included on mid to low end boards
due to it being expected that any standard case would
accomodate it.

http://www.uvem.com/images/case_comp...2_rear_big.jpg

"kony" wrote in message
news
On Mon, 19 Sep 2005 16:27:51 +0200, "jona"
wrote:


Perhaps, or it could be some bios quirk where IF you got it
running with a different CPU, updating the bios could
possibly help with the other CPU or memory compatibility.

I tried my sister's Athlon-PECM, 807Mhz processor and again,
it won't even post. I'm not sure whether it should/would though.
I mean, that processor is "ancient" technology and certainly not
DDR compatible, but ..... ??

It should've worked, the memory controller is in the
northbridge so the CPU cares not what it is, and the FSB was
always DDR on a socket A athlon. It could be necessary to
clear CMOS before the old CPU would POST the system, but
beyond that it should've worked.

That's what I thought.

The only new development is that when testing now, or
with any other PSU (i tried two), the machine powers up
directly when connecting to the wall socket - whether I
have the switch connected to it's pinout or not. Still no
post though. I also tried the CPU in question on my sister's
board (Asus A7V) where it doesn't work either - but that
doesn't surprise me. BUT, under correction, her Athlon
should have worked on the new board, right ?

I would've expected it to. If this board is still in a box,
I'd try it out of the box on a non-conductive (not
anti-static material) surface and without the case wiring
connected. Sometimes it's the simple things like a stuck
case switch or a case standoff shorting the back of the
board that'll be the hardest things to find if you're not
looking for them.

I can always give it one more try just before I repack it
in it's box to send back to the supplier.

Further, the "new" board came all boxed with it's various
goodies, including IDE cable, manual, driver CD, etc -
but *without* the I/O backplate/panel. That almost suggests
that it has before been installed somewhere and returned
to the vendor .....

If it uses the old standard ATX style of rear port
locations, like the following picture, it's fairly common
for the panel to not be included on mid to low end boards
due to it being expected that any standard case would
accomodate it.

http://www.uvem.com/images/case_comp...2_rear_big.jpg

No it doesn't. The audio sockets are vertically aligned (same
direction as keyb & mouse sockets).

Thanks for your time and suggestions.

"kony" wrote in message
news
On Mon, 19 Sep 2005 16:27:51 +0200, "jona"
wrote:


Perhaps, or it could be some bios quirk where IF you got it
running with a different CPU, updating the bios could
possibly help with the other CPU or memory compatibility.

I tried my sister's Athlon-PECM, 807Mhz processor and again,
it won't even post. I'm not sure whether it should/would though.
I mean, that processor is "ancient" technology and certainly not
DDR compatible, but ..... ??

It should've worked, the memory controller is in the
northbridge so the CPU cares not what it is, and the FSB was
always DDR on a socket A athlon. It could be necessary to
clear CMOS before the old CPU would POST the system, but
beyond that it should've worked.

That's what I thought.

The only new development is that when testing now, or
with any other PSU (i tried two), the machine powers up
directly when connecting to the wall socket - whether I
have the switch connected to it's pinout or not. Still no
post though. I also tried the CPU in question on my sister's
board (Asus A7V) where it doesn't work either - but that
doesn't surprise me. BUT, under correction, her Athlon
should have worked on the new board, right ?

I would've expected it to. If this board is still in a box,
I'd try it out of the box on a non-conductive (not
anti-static material) surface and without the case wiring
connected. Sometimes it's the simple things like a stuck
case switch or a case standoff shorting the back of the
board that'll be the hardest things to find if you're not
looking for them.

I can always give it one more try just before I repack it
in it's box to send back to the supplier.

Further, the "new" board came all boxed with it's various
goodies, including IDE cable, manual, driver CD, etc -
but *without* the I/O backplate/panel. That almost suggests
that it has before been installed somewhere and returned
to the vendor .....

If it uses the old standard ATX style of rear port
locations, like the following picture, it's fairly common
for the panel to not be included on mid to low end boards
due to it being expected that any standard case would
accomodate it.

http://www.uvem.com/images/case_comp...2_rear_big.jpg

No it doesn't. The audio sockets are vertically aligned (same
direction as keyb & mouse sockets).

Thanks for your time and suggestions.