K7S5A Pro Won't POST

A year or so ago I bought three K7S5A Pro boards to replace older PC's on an
as-needed basis. Got three identical cases with Austin 300W P4 capable PSs,
three Athalon 2400XP CPUs and 3 strips of Centon 512DDR. The first two
machines went together without a hitch (other than my forgetting to bump the
bus speed up to 133) using 512MB DDR and 2400 Athalon XP processors.
They've been running fine for months.

Now I've gotten to the third one and it refuses to POST. I can even pull
the RAM (and have swapped in and out known good SD and DDR modules) but
nada, zip, zilch, bupkis. Speaker is connected but never beeps. The CPU fan
spins and it's obvious the 'power on' switch works, but no signal is sent to
the CRT and no beeps are heard, even if I pull the video card or the memory
modules. Before I go fishing into two perfectly good machines to start
swapping CPUs and PS's, can anyone suggest something else to try? I've
already changed out the battery and cleared CMOS via the J1 jumper and tried
all sorts of PCI and AGP video cards. Only signs of life are the
motherboard-powered case and CPU fans.

I'm think it's a bad CPU. How often does that occur?

TIA,

Eric

Eric Wolfe wrote:

> A year or so ago I bought three K7S5A Pro boards to replace older PC's on an
> as-needed basis. Got three identical cases with Austin 300W P4 capable PSs,
> three Athalon 2400XP CPUs and 3 strips of Centon 512DDR. The first two
> machines went together without a hitch (other than my forgetting to bump the
> bus speed up to 133) using 512MB DDR and 2400 Athalon XP processors.
> They've been running fine for months.
>
> Now I've gotten to the third one and it refuses to POST. I can even pull
> the RAM (and have swapped in and out known good SD and DDR modules) but
> nada, zip, zilch, bupkis. Speaker is connected but never beeps. The CPU fan
> spins and it's obvious the 'power on' switch works, but no signal is sent to
> the CRT and no beeps are heard, even if I pull the video card or the memory
> modules. Before I go fishing into two perfectly good machines to start
> swapping CPUs and PS's, can anyone suggest something else to try? I've
> already changed out the battery and cleared CMOS via the J1 jumper and tried
> all sorts of PCI and AGP video cards. Only signs of life are the
> motherboard-powered case and CPU fans.
>
> I'm think it's a bad CPU. How often does that occur?
>
> TIA,
>
> Eric
>
>
>
>

While you are doing this test, have you disconnected your cables to the
IDE bus and the floppy disk controller??? It is possible for a reversed
IDE cable to ground out the address bus and prevent a boot up. Also,
since it is easy to short out the underside of a motherboard with a
standoff in the wrong position, I would suggest you remove the MB from
the case until you get it running if you are not already doing that.

It is possible that you have a bad processor, but I would not conclude
that until everything (including the PS) had been eliminated. This
means clearing CMOS with the jumper for that purpose as well.

Thanks for the reply, Ken!

> While you are doing this test, have you disconnected your cables to the
> IDE bus and the floppy disk controller??? It is possible for a reversed
> IDE cable to ground out the address bus and prevent a boot up. Also,
> since it is easy to short out the underside of a motherboard with a
> standoff in the wrong position, I would suggest you remove the MB from
> the case until you get it running if you are not already doing that.

Did all of those and even removed all front panel cables except power on
switch.

> It is possible that you have a bad processor, but I would not conclude
> that until everything (including the PS) had been eliminated. This
> means clearing CMOS with the jumper for that purpose as well.

Well, I had a spare 1Ghz Duron lying around that I had forgotten about and
it boots just fine so I suppose I have eliminated the PS and everything else
but the damn CPU.

Is there anything that could be wrong with the default settings that would
not allow a 2400XP to boot? This was a retail special I got from Newegg a
while back so it's still under AMD "alleged" warranty. I'd just like not to
send back a perfectly good chip but I am reluctant to bust up a well-tested
sister machine just to swap CPUs. I know with some other boards you have to
made sure BIOS settings and jumpers are correct (in fact I bought the Duron
just to be able to boot a Gigabyte board that would not see chips faster
than 2000 unless you made a BIOS change - you see the Catch 22 there! It's
possibly the crappy stock AMD fan isn't spinning fast enough but since it
never posts, I assume it's a bad CPU. Since the boards, the memory and the
CPUs were all bought at the same time and had very close serial numbers, I
assume that it's a bad CPU. I am going to run Memtest + for a while to see
if I can spot any other anomalies.

Anyone know how hard/easy it is to get AMD to make good on their guarantee?
Everything I've read seems to indicate they're pretty fussy on what they
will take back for repair or replacement.

TIA,

Eric

Eric Wolfe wrote:
> Thanks for the reply, Ken!
>
>
>>While you are doing this test, have you disconnected your cables to the
>>IDE bus and the floppy disk controller??? It is possible for a reversed
>>IDE cable to ground out the address bus and prevent a boot up. Also,
>>since it is easy to short out the underside of a motherboard with a
>>standoff in the wrong position, I would suggest you remove the MB from
>>the case until you get it running if you are not already doing that.
>
>
> Did all of those and even removed all front panel cables except power on
> switch.
>
>
>>It is possible that you have a bad processor, but I would not conclude
>>that until everything (including the PS) had been eliminated. This
>>means clearing CMOS with the jumper for that purpose as well.
>
>
> Well, I had a spare 1Ghz Duron lying around that I had forgotten about and
> it boots just fine so I suppose I have eliminated the PS and everything else
> but the damn CPU.
>
> Is there anything that could be wrong with the default settings that would
> not allow a 2400XP to boot? This was a retail special I got from Newegg a
> while back so it's still under AMD "alleged" warranty. I'd just like not to
> send back a perfectly good chip but I am reluctant to bust up a well-tested
> sister machine just to swap CPUs. I know with some other boards you have to
> made sure BIOS settings and jumpers are correct (in fact I bought the Duron
> just to be able to boot a Gigabyte board that would not see chips faster
> than 2000 unless you made a BIOS change - you see the Catch 22 there! It's
> possibly the crappy stock AMD fan isn't spinning fast enough but since it
> never posts, I assume it's a bad CPU. Since the boards, the memory and the
> CPUs were all bought at the same time and had very close serial numbers, I
> assume that it's a bad CPU. I am going to run Memtest + for a while to see
> if I can spot any other anomalies.
>
> Anyone know how hard/easy it is to get AMD to make good on their guarantee?
> Everything I've read seems to indicate they're pretty fussy on what they
> will take back for repair or replacement.
>
> TIA,
>
> Eric
>
>

As far as settings on the MB, I do not recall any as I think the MP set
the voltage itself. CHECK YOUR MANUAL HOWEVER, AS I AM TRUSTING ON MY
MEMORY. Clearing the CMOS with the jumper for that purpose would set
you into a state that the MP should be accepted if it were good. It
would be set to a lower FSB and clock setting, but it should still start
up since the settings are less than the rated specs of the MP. I had a
bad MP on my purchase of a MB and MP combo, so it does sometimes happen.
It has run flawlessly since however.

> As far as settings on the MB, I do not recall any as I think the MP set
> the voltage itself. CHECK YOUR MANUAL HOWEVER, AS I AM TRUSTING ON MY
> MEMORY. Clearing the CMOS with the jumper for that purpose would set
> you into a state that the MP should be accepted if it were good. It
> would be set to a lower FSB and clock setting, but it should still start
> up since the settings are less than the rated specs of the MP. I had a
> bad MP on my purchase of a MB and MP combo, so it does sometimes happen.
> It has run flawlessly since however.

I've checked the manual and there is nothing to configure. I've contacted
AMD to arrange a replacement but they've already warned me that they may not
have any 2400+ XP 266FSB's to replace it with. I'm not sure what will
happen if they don't because there aren't many equivalent or faster CPUs
that I am aware of that will run on the K7S5A Pro. I guess this means when
batch buying I've got to buy 4 CPUS for every 3 motherboards to really be
safe.

One good thing I've learned is that when fans spin but there's no video and
no beeps even without a video card or RAM installed it's very likely to be
the CPU. I remember chasing my tail for a long, long time when first I
encountered a bad CPU. Of course, AMD could inspect the sucker and find out
that it's working A-OK. I suppose that's easier than me breaking apart a
fully cabled and squirreled away machine buried very deep behind the desk
just to test it. That's another thing I've learned hacking on PCs. Don't
bust up good machines to test bad ones if you can avoid it.

The irony is I bought my first ECS board because it was so cheap that I was
going to keep it around as a spare component tester. I liked its ease of
use and dual RAM capabilities so I bought three more to upgrade older
machines as time permitted. The only problem I've had was with the onboard
sound connectors. I didn't realize the CD1 and CD2 were wired differently.
Now if only I had a little robot I could feed into the PC case through a
hatch that would switch the cable from one header to the other.

Thanks again for your input, Ken!

Eric

On Thu, 21 Oct 2004 00:34:28 -0400, "Eric Wolfe"
<> wrote:

>> As far as settings on the MB, I do not recall any as I think the MP set
>> the voltage itself. CHECK YOUR MANUAL HOWEVER, AS I AM TRUSTING ON MY
>> MEMORY. Clearing the CMOS with the jumper for that purpose would set
>> you into a state that the MP should be accepted if it were good. It
>> would be set to a lower FSB and clock setting, but it should still start
>> up since the settings are less than the rated specs of the MP. I had a
>> bad MP on my purchase of a MB and MP combo, so it does sometimes happen.
>> It has run flawlessly since however.
>
>I've checked the manual and there is nothing to configure. I've contacted
>AMD to arrange a replacement but they've already warned me that they may not
>have any 2400+ XP 266FSB's to replace it with. I'm not sure what will
>happen if they don't because there aren't many equivalent or faster CPUs
>that I am aware of that will run on the K7S5A Pro. I guess this means when
>batch buying I've got to buy 4 CPUS for every 3 motherboards to really be
>safe.
>
>One good thing I've learned is that when fans spin but there's no video and
>no beeps even without a video card or RAM installed it's very likely to be
>the CPU. I remember chasing my tail for a long, long time when first I
>encountered a bad CPU. Of course, AMD could inspect the sucker and find out
>that it's working A-OK. I suppose that's easier than me breaking apart a
>fully cabled and squirreled away machine buried very deep behind the desk
>just to test it. That's another thing I've learned hacking on PCs. Don't
>bust up good machines to test bad ones if you can avoid it.
>
>The irony is I bought my first ECS board because it was so cheap that I was
>going to keep it around as a spare component tester. I liked its ease of
>use and dual RAM capabilities so I bought three more to upgrade older
>machines as time permitted. The only problem I've had was with the onboard
>sound connectors. I didn't realize the CD1 and CD2 were wired differently.
>Now if only I had a little robot I could feed into the PC case through a
>hatch that would switch the cable from one header to the other.
>
>Thanks again for your input, Ken!
>
>Eric

You are being extremely paranoid about using a working machine to
test a suspect machine.

In the time you have spent typing these messages, you could have
easily eliminated all suspect parts from the suspect machine. It is
the fastest and easiest troubleshooting tool you have. Use it.

The only caveat is that you use the good machine as your platform for
testing the suspect parts.

Pj

"PJX" <> wrote in message

>You are being extremely paranoid about using a working machine to
> test a suspect machine.

Ubetcha! Personal quirk along the line of "if it ain't broke, don't fix
it."

We'll see if it was the right course of action. If it was, AMD will send me
a new processor. If the Duron ran and the Athlon didn't, that's good enough
for me. I can't afford to have the working machine down right now and every
time you pop a heatsink and swap a CPU you're inviting chipped cores,
breaking the tabs off the ZIF socket, static damage and a host of other
maladies that I would rather avoid. The solution is to get a cheap MB to
use solely as a test board but that's why I bought the dirt cheap ECS K7S5A.
Somehow, when they survive all the abuse a test bench machine takes, I can't
help but turning them into workhorse machines.

My track record, up until now, has been good with ECS boards. Far better
than Giga*bite*, Abit, Tyan, Soyo and Asus - and all of those board cost a
lot more. It's going to cost me six bucks to send the CPU back to AMD. I
can live with than gamble easier than downing a working machine because I'm
an avowed klutz.

Besides, there's that old joke about swapping parts that goes like this. An
IBM SW guy, HW guy and marketing guy are driving along and get a flat tire.
The SW guy says "Let's drive it a while more and see if the problem fixes
itself." The HW guy says "Let's swap tires and see if that fixes the
problem" and the marketing guy says "We need a new car!"

Eric

Forgive the top posting but I wanted to follow up on my original post. I
just got a notice from AMD that they are sending out a new processor. The
one I returned apparently was DOA so I did the right thing in not tearing
down a working machine to try to test the CPU further when it failed to POST
in the K7S5A. I figured if the old Duron booted with no problems, it was
pretty likely that the Athlon was dead.

Eric

"PJX" <> wrote in message

> >I've checked the manual and there is nothing to configure. I've
contacted
> >AMD to arrange a replacement but they've already warned me that they may
not
> >have any 2400+ XP 266FSB's to replace it with. I'm not sure what will
> >happen if they don't because there aren't many equivalent or faster CPUs
> >that I am aware of that will run on the K7S5A Pro. I guess this means
when
> >batch buying I've got to buy 4 CPUS for every 3 motherboards to really be
> >safe.
> >
> >One good thing I've learned is that when fans spin but there's no video
and
> >no beeps even without a video card or RAM installed it's very likely to
be
> >the CPU. I remember chasing my tail for a long, long time when first I
> >encountered a bad CPU. Of course, AMD could inspect the sucker and find
out
> >that it's working A-OK. I suppose that's easier than me breaking apart a
> >fully cabled and squirreled away machine buried very deep behind the desk
> >just to test it. That's another thing I've learned hacking on PCs.
Don't
> >bust up good machines to test bad ones if you can avoid it.
> >
> >The irony is I bought my first ECS board because it was so cheap that I
was
> >going to keep it around as a spare component tester. I liked its ease of
> >use and dual RAM capabilities so I bought three more to upgrade older
> >machines as time permitted. The only problem I've had was with the
onboard
> >sound connectors. I didn't realize the CD1 and CD2 were wired
differently.
> >Now if only I had a little robot I could feed into the PC case through a
> >hatch that would switch the cable from one header to the other.
> >
> >Thanks again for your input, Ken!
> >
> >Eric
>
> You are being extremely paranoid about using a working machine to
> test a suspect machine.
>
> In the time you have spent typing these messages, you could have
> easily eliminated all suspect parts from the suspect machine. It is
> the fastest and easiest troubleshooting tool you have. Use it.
>
> The only caveat is that you use the good machine as your platform for
> testing the suspect parts.
>
> Pj
>
>
>
>
>
>

Good job.

"Eric Wolfe" <> wrote in message
news:IeCdnVt95fZ1K-PcRVn-...
> Forgive the top posting but I wanted to follow up on my original post. I
> just got a notice from AMD that they are sending out a new processor. The
> one I returned apparently was DOA so I did the right thing in not tearing
> down a working machine to try to test the CPU further when it failed to POST
> in the K7S5A. I figured if the old Duron booted with no problems, it was
> pretty likely that the Athlon was dead.
>
> Eric
>
> "PJX" <> wrote in message
>
> > >I've checked the manual and there is nothing to configure. I've
> contacted
> > >AMD to arrange a replacement but they've already warned me that they may
> not
> > >have any 2400+ XP 266FSB's to replace it with. I'm not sure what will
> > >happen if they don't because there aren't many equivalent or faster CPUs
> > >that I am aware of that will run on the K7S5A Pro. I guess this means
> when
> > >batch buying I've got to buy 4 CPUS for every 3 motherboards to really be
> > >safe.
> > >
> > >One good thing I've learned is that when fans spin but there's no video
> and
> > >no beeps even without a video card or RAM installed it's very likely to
> be
> > >the CPU. I remember chasing my tail for a long, long time when first I
> > >encountered a bad CPU. Of course, AMD could inspect the sucker and find
> out
> > >that it's working A-OK. I suppose that's easier than me breaking apart a
> > >fully cabled and squirreled away machine buried very deep behind the desk
> > >just to test it. That's another thing I've learned hacking on PCs.
> Don't
> > >bust up good machines to test bad ones if you can avoid it.
> > >
> > >The irony is I bought my first ECS board because it was so cheap that I
> was
> > >going to keep it around as a spare component tester. I liked its ease of
> > >use and dual RAM capabilities so I bought three more to upgrade older
> > >machines as time permitted. The only problem I've had was with the
> onboard
> > >sound connectors. I didn't realize the CD1 and CD2 were wired
> differently.
> > >Now if only I had a little robot I could feed into the PC case through a
> > >hatch that would switch the cable from one header to the other.
> > >
> > >Thanks again for your input, Ken!
> > >
> > >Eric
> >
> > You are being extremely paranoid about using a working machine to
> > test a suspect machine.
> >
> > In the time you have spent typing these messages, you could have
> > easily eliminated all suspect parts from the suspect machine. It is
> > the fastest and easiest troubleshooting tool you have. Use it.
> >
> > The only caveat is that you use the good machine as your platform for
> > testing the suspect parts.
> >
> > Pj
> >
> >
> >
> >
> >
> >
>
>

On Tue, 26 Oct 2004 16:40:34 -0400, "Eric Wolfe"
<> wrote:

>Forgive the top posting but I wanted to follow up on my original post. I
>just got a notice from AMD that they are sending out a new processor. The
>one I returned apparently was DOA so I did the right thing in not tearing
>down a working machine to try to test the CPU further when it failed to POST
>in the K7S5A. I figured if the old Duron booted with no problems, it was
>pretty likely that the Athlon was dead.
>
>Eric

I disagree.

The quickest and easiest way would have been to stick the bad
processer in the good board and watched it not boot.

5 minutes max troubleshooting time and your good system is back up
and running and you send off the bad processor.

Time is money.

Pj








>
>"PJX" <> wrote in message
>
>> >I've checked the manual and there is nothing to configure. I've
>contacted
>> >AMD to arrange a replacement but they've already warned me that they may
>not
>> >have any 2400+ XP 266FSB's to replace it with. I'm not sure what will
>> >happen if they don't because there aren't many equivalent or faster CPUs
>> >that I am aware of that will run on the K7S5A Pro. I guess this means
>when
>> >batch buying I've got to buy 4 CPUS for every 3 motherboards to really be
>> >safe.
>> >
>> >One good thing I've learned is that when fans spin but there's no video
>and
>> >no beeps even without a video card or RAM installed it's very likely to
>be
>> >the CPU. I remember chasing my tail for a long, long time when first I
>> >encountered a bad CPU. Of course, AMD could inspect the sucker and find
>out
>> >that it's working A-OK. I suppose that's easier than me breaking apart a
>> >fully cabled and squirreled away machine buried very deep behind the desk
>> >just to test it. That's another thing I've learned hacking on PCs.
>Don't
>> >bust up good machines to test bad ones if you can avoid it.
>> >
>> >The irony is I bought my first ECS board because it was so cheap that I
>was
>> >going to keep it around as a spare component tester. I liked its ease of
>> >use and dual RAM capabilities so I bought three more to upgrade older
>> >machines as time permitted. The only problem I've had was with the
>onboard
>> >sound connectors. I didn't realize the CD1 and CD2 were wired
>differently.
>> >Now if only I had a little robot I could feed into the PC case through a
>> >hatch that would switch the cable from one header to the other.
>> >
>> >Thanks again for your input, Ken!
>> >
>> >Eric
>>
>> You are being extremely paranoid about using a working machine to
>> test a suspect machine.
>>
>> In the time you have spent typing these messages, you could have
>> easily eliminated all suspect parts from the suspect machine. It is
>> the fastest and easiest troubleshooting tool you have. Use it.
>>
>> The only caveat is that you use the good machine as your platform for
>> testing the suspect parts.
>>
>> Pj
>>
>>
>>
>>
>>
>>
>