crash and burn

I've been having endless problems with my computer, and hope someone
here can suggest something. It abruptly crashes and restarts several
times a day.

This occurs under both Windows XP and Linux, so I don't believe it's a
software issue. I took the following steps to diagnose it, although
nothing has helped:

1. I tried using the "failsafe" settings in the BIOS. No change.
2. I updated to my latest BIOS and latest motherboard manufacturer's
set of drivers. Nothing. The motherboard is a Tyan S5101, by the way.
3. I monitored the temperature and found that the crashes still happen
when the CPU is only at 38C. The video card is also fairly cool.
4. I swapped out the power supply for a 400 watt new one. An Antec. No
change.
5. I replaced the video card (an ATI 9800 Pro) with a known good
Nvidia 440-MX. No change.
6. I extensively tested the RAM, using a diagnostic program. No errors
there. I then tried running the system with the RAM underclocked in
the BIOS. No help there. I then removed the RAM and tried in different
slots. I tried them one at a time, hoping one of them was bad. Nothing
made a difference.
7. I pulled the hard drive and put in a new one. I installed Windows
XP fresh, with no applications. The crash and restart still happened.
8. I checked the processor and heatsink and CPU fan. All in order. I
did use thermal compound, and the heatsink was snug against the
processor.
9. I removed ALL PCI cards from the system, and disconnected ALL
devices except the hard drive. No CD-ROM, no modem, nothing. The
crashes still happened.
10. I turned off all integrated peripherals, in the BIOS (networking,
for example), except the hard drive controller. Still, the crashes
happened.
11. I inspected the motherboard for loose components and blackened
areas. Nothing seemed amiss.

While I did mention this happened in Linux as well as Windows, it's
worth mentioning that I've thoroughly scanned my system for spyware
(Adware, SpyBot) and viruses (AVG). I've checked the logs (on Linux)
and the Event Viewer (on Windows XP).

I have now replaced every piece of my system, with the exception of:
a) The motherboard itself, and
b) The CPU, and
c) the RAM, although as I said I ran RAM tests, underclocked it, tried
it in different slots, and tried them one at a time.

Frustrating. I'm considering setting the whole mess on fire,
extinguishing the smoldering ashes in a manner I won't describe here,
and buying a slow, overpriced Macintosh.

My questions:
Is there something that stands out, that I am not considering?
My next step is a new CPU. Is this reasonable?
Or should I proceed with a new motherboard?
Or should I buy name-brand memory, even in light of the above testing?
Should I be suspecting an issue with the power from the wall and/or
surge protector?

Any advice would be greatly appeciated.

I went through almost as much trouble once only to find it was my "power
on" button - had to replace it. Other than that, I'd suspect motherboard.

If you are not getting blue screens with warnings on them, it may be
the power supply. Did you check the event and error logs?


Jerry G.
========


(Marc Missire) wrote in message om...
I've been having endless problems with my computer, and hope someone
here can suggest something. It abruptly crashes and restarts several
times a day.

This occurs under both Windows XP and Linux, so I don't believe it's a
software issue. I took the following steps to diagnose it, although
nothing has helped:

1. I tried using the "failsafe" settings in the BIOS. No change.
2. I updated to my latest BIOS and latest motherboard manufacturer's
set of drivers. Nothing. The motherboard is a Tyan S5101, by the way.
3. I monitored the temperature and found that the crashes still happen
when the CPU is only at 38C. The video card is also fairly cool.
4. I swapped out the power supply for a 400 watt new one. An Antec. No
change.
5. I replaced the video card (an ATI 9800 Pro) with a known good
Nvidia 440-MX. No change.
6. I extensively tested the RAM, using a diagnostic program. No errors
there. I then tried running the system with the RAM underclocked in
the BIOS. No help there. I then removed the RAM and tried in different
slots. I tried them one at a time, hoping one of them was bad. Nothing
made a difference.
7. I pulled the hard drive and put in a new one. I installed Windows
XP fresh, with no applications. The crash and restart still happened.
8. I checked the processor and heatsink and CPU fan. All in order. I
did use thermal compound, and the heatsink was snug against the
processor.
9. I removed ALL PCI cards from the system, and disconnected ALL
devices except the hard drive. No CD-ROM, no modem, nothing. The
crashes still happened.
10. I turned off all integrated peripherals, in the BIOS (networking,
for example), except the hard drive controller. Still, the crashes
happened.
11. I inspected the motherboard for loose components and blackened
areas. Nothing seemed amiss.

While I did mention this happened in Linux as well as Windows, it's
worth mentioning that I've thoroughly scanned my system for spyware
(Adware, SpyBot) and viruses (AVG). I've checked the logs (on Linux)
and the Event Viewer (on Windows XP).

I have now replaced every piece of my system, with the exception of:
a) The motherboard itself, and
b) The CPU, and
c) the RAM, although as I said I ran RAM tests, underclocked it, tried
it in different slots, and tried them one at a time.

Frustrating. I'm considering setting the whole mess on fire,
extinguishing the smoldering ashes in a manner I won't describe here,
and buying a slow, overpriced Macintosh.

My questions:
Is there something that stands out, that I am not considering?
My next step is a new CPU. Is this reasonable?
Or should I proceed with a new motherboard?
Or should I buy name-brand memory, even in light of the above testing?
Should I be suspecting an issue with the power from the wall and/or
surge protector?

Any advice would be greatly appeciated.

You are spending way too much times speculating and wildly replacing
things rather than first find reasons for problems. For example, if
that power supply was a problem, then the simple procedure with a 3.5
digit multimeter would have identified the problem immediately -
whether it be a supply too small or a supply slowly failing. Having
confirmed the power supply was OK, then you could have moved on to
other potential reasons for failure. Instead you are now spending
money on useless solutions.

Operating systems come with historical information. What did the
system (event) logs reports? Was there any problem indicted in Device
Manager? Notice the trend. First get the facts before replacing
things.

Comprehensive diagnostics. When running the memory diagnostic, then
intermittent memory will pass. That is until memory is at the normal
higher temperature such as room temperature of 100 degrees F, or by
heating memory with a hair dryer on high (memory hot enough so that it
is uncomfortable to touch but does not leave skin). Same with
diagnostics for the rest of that machine. Those comprehensive
diagnostics are provided by all good manufacturers on their web site
for free. Otherwise you must download each hardware diagnostic from
every vendor. That's right. A decent motherboard manufacturer should
provide comprehensive diagnostics. Many do not.

Those hardware tests are best executed in a 100 degree F room.
Don't believe it? Then look up the specs for those computer parts.
Detailed numerical specs that many 'discount' manufacturers can't be
bothered to provide. All fully functional computers must execute
every comprehensive diagnostic when in a 100 degree F room. This was
only part of what we used to do when burn-in testing. In fact, the
standard PC must execute normally in a 100 degree room. Look up their
specs.

A computer running hot tends to expose its intermittent problems
rather repeatedly. Too many instead, fix the symptoms rather than the
problem. They install more fans. A single 80 mm fan is usually
sufficient for any computer even running in a 100 degree F room. That
is if the hardware is not intermittent or designed defectively.

Just a first (and quick) step to solving intermittent problems. IOW
first find the problem. Only then try to fix it. Best way to make an
intermittent into an 'easy to find' failure? Heat.

(Marc Missire) wrote in message om...
I've been having endless problems with my computer, and hope someone
here can suggest something. It abruptly crashes and restarts several
times a day.

This occurs under both Windows XP and Linux, so I don't believe it's a
software issue. I took the following steps to diagnose it, although
nothing has helped:

1. I tried using the "failsafe" settings in the BIOS. No change.
2. I updated to my latest BIOS and latest motherboard manufacturer's
set of drivers. Nothing. The motherboard is a Tyan S5101, by the way.
3. I monitored the temperature and found that the crashes still happen
when the CPU is only at 38C. The video card is also fairly cool.
4. I swapped out the power supply for a 400 watt new one. An Antec. No
change.
5. I replaced the video card (an ATI 9800 Pro) with a known good
Nvidia 440-MX. No change.
6. I extensively tested the RAM, using a diagnostic program. No errors
there. I then tried running the system with the RAM underclocked in
the BIOS. No help there. I then removed the RAM and tried in different
slots. I tried them one at a time, hoping one of them was bad. Nothing
made a difference.
7. I pulled the hard drive and put in a new one. I installed Windows
XP fresh, with no applications. The crash and restart still happened.
8. I checked the processor and heatsink and CPU fan. All in order. I
did use thermal compound, and the heatsink was snug against the
processor.
9. I removed ALL PCI cards from the system, and disconnected ALL
devices except the hard drive. No CD-ROM, no modem, nothing. The
crashes still happened.
10. I turned off all integrated peripherals, in the BIOS (networking,
for example), except the hard drive controller. Still, the crashes
happened.
11. I inspected the motherboard for loose components and blackened
areas. Nothing seemed amiss.

While I did mention this happened in Linux as well as Windows, it's
worth mentioning that I've thoroughly scanned my system for spyware
(Adware, SpyBot) and viruses (AVG). I've checked the logs (on Linux)
and the Event Viewer (on Windows XP).

I have now replaced every piece of my system, with the exception of:
a) The motherboard itself, and
b) The CPU, and
c) the RAM, although as I said I ran RAM tests, underclocked it, tried
it in different slots, and tried them one at a time.

Frustrating. I'm considering setting the whole mess on fire,
extinguishing the smoldering ashes in a manner I won't describe here,
and buying a slow, overpriced Macintosh.

My questions:
Is there something that stands out, that I am not considering?
My next step is a new CPU. Is this reasonable?
Or should I proceed with a new motherboard?
Or should I buy name-brand memory, even in light of the above testing?
Should I be suspecting an issue with the power from the wall and/or
surge protector?

Any advice would be greatly appeciated.

I eventually replaced the Tyan S5101 motherboard with and Intel D865PERL.

The crashes are gone. Thanks to all who posted advice.

(Marc Missire) wrote in message . com...
I eventually replaced the Tyan S5101 motherboard with and Intel D865PERL.

The crashes are gone. Thanks to all who posted advice.

I've the same problem on my system but i've alredy changed the
motherboard!
In addition there's a problem with the BIOS checksum "BIOS Rom
Checksum Error" and the MB reload the default BIOS settings.
The UPS is ok and all the tensions are good even during the system up
and with load.
My processor is a PURE AMD Athlon XP 2400+ with a liquid cooled
heatsink...
In the errorlog of Win XP I've found SYSTEM FAILURE (1003) error 102
and when the system rise up after the reboot the error was "Product_ID
256_1". Someone could help me?