Help - random crashes in XP but not 98

  • Thread starter Thread starter Kadiir
  • Start date Start date
K

Kadiir

Hello:

Sorry, this is a bit long...

My system has recently started randomly crashing (no 2 BSOD's the same
in a row and I've seen at least 5 different ones). So far (I'll do
more extensive testing tonight) it only happens in Windows XP Pro SP1
(all the latest WindowsUpdate patches) and not Windows 98SE (fully
patched as well).

At first it appeared to be a heat issue. According to BIOS case temp
(the only measure I have), when it hit 49C it would CTD (games) or
reboot. So, I changed cases (from a case w/ poor airflow to an Antec
Sonata w/ the optional front fan). I've got round cables, so now
nothing is blocking airflow (I had a flat cable that was blocking it
in the old case - replacing it gave me 3C more before crashing).

I also replaced the stock Intel HSF with a Zalman 3100+ (includes the
FM123 + 1 92mm fan) and on my GF I put a Zalman ZM-80A (just this
weekend I added another 92mm fan over that heatsink, but it still
crashes).

Doing the above bought me some time - it seemed to stay at a higher
temp longer for a short while. Then, it was crashing many times per
day.

At this point, let me cover the hardware:

Abit BX6-R2 (QR BIOS)
Intel P3-850
512MB RAM (oh, I ran memtest for 5+ hours last night - no errors).
Gainward GF3-Ti500 (product info is a Ti-550 Golden Sample)
Turtle Beach Santa Cruz w/ Roland daughterboard
3 Hard drives (15GB IBM, 75GB IBM, & 120GB WD)
Adaptec 2930CU SCSI card
Plextor UltraMax 40 CDROM (SCSI)
Plextor PlexWriter 8/2/20 (SCSI)

The mobo is a slot one design and I've got a Socket 370 CPU w/ a
slocket. In order to fit the ZM-80a, I removed the support on the
side of the slot closest to the vidcard, however the slocket still has
a slight contact with the heatsink.

A mech. engineer friend of mine assisted me with performing a little
surgery on the board - I have a rev 2.01 board which has (had) the CPU
temp monitoring capability (very inaccurate and had a bug in the last
bios release). Because of the bug (couldn't run ACPI in Win2K or XP),
I finally got around to doing the fix (remove resistor R266 which is
part of the temp monitoring line). After that, all of my crashes
stopped.

In retrospect, I think the crashes stopped because I had to take
everything out of the case in order to remove the resistor. I suspect
that one or both heatsinks it too heavy and is either warping the mobo
sufficiently to cause a crash or pulling one or both devices (cpu
and/or vidcard) out of its socket. However, since Win98 doesn't
crash, it really has me perplexed.

Any ideas?

TIA!
 
Left a couple things out...

[snip]
Gainward GF3-Ti500 (product info is a Ti-550 Golden Sample)
Click to expand...

Runnin 30.82 drivers.

[snilp]
A mech. engineer friend of mine assisted me with performing a little
surgery on the board - I have a rev 2.01 board which has (had) the CPU
temp monitoring capability (very inaccurate and had a bug in the last
bios release). Because of the bug (couldn't run ACPI in Win2K or XP),
I finally got around to doing the fix (remove resistor R266 which is
part of the temp monitoring line). After that, all of my crashes
stopped.
Click to expand...

I forgot to mention that I reinstalled Windows (both versions - also
re-partitioned & re-formatted the drive) after doing the above. After
about5-6 days, the crashes started again.

kad
 
Does the Abit BX6-R2 officially support a P3 850?

I tried running a Celeron 900 in this board, while in Windows 98, it
worked fine, XP would freeze on bootup. I eventually came to the
conclusion that while a Celeron 900 would work in this board for the
most part, the 900 was not fully compatible. I tried running a Celeron
300 with XP and there was no problem, I ended up changing the MB to a
slightly newer Asus socketed model and the problem disappeared. I
wonder if the P3 850 has the same problems.

If you disable the L2 cache, and your problem disappears, you will
know there is a CPU/MB compatibility problem.
 
Does the Abit BX6-R2 officially support a P3 850?
Click to expand...

With the QR BIOS, yes (all will support the P3 900 and some will
support the 1GHz in slot 1 format).

Also, it has worked fine for over a year and only recently started
having problems (when I started doing some changes - changed from a
super tower to mid tower to fit it in a desk, changed things because
it has always been too loud, & changed things (heatsinks and
eventually the case) because it was running too hot in the desk).
I tried running a Celeron 900 in this board, while in Windows 98, it
worked fine, XP would freeze on bootup. I eventually came to the
conclusion that while a Celeron 900 would work in this board for the
most part, the 900 was not fully compatible. I tried running a Celeron
300 with XP and there was no problem, I ended up changing the MB to a
slightly newer Asus socketed model and the problem disappeared. I
wonder if the P3 850 has the same problems.

If you disable the L2 cache, and your problem disappears, you will
know there is a CPU/MB compatibility problem.
Click to expand...

I'll try it, but like I said above - it's worked for over a year (I've
had the processor since 2000 or 2001 & XP since about 6 months after
release).

kad
 
Kadiir said:
That's what I thought, too, until I changed the case to the Sonata -
it comes with an Antec TruePower 380.

Although, I suppose it could be a bad one.

kad
Click to expand...

I just RMA'd an Antec TruePower 480. Made the system unstable and
smelled like it was burning up. It also got really hot and made my
system temps skyrocket. I've heard great things about Antec and that's
why I tried one. I went back to using my old Sparkle and system is
stable and cool.

Bob
 
I just RMA'd an Antec TruePower 480. Made the system unstable and
smelled like it was burning up. It also got really hot and made my
system temps skyrocket. I've heard great things about Antec and that's
why I tried one. I went back to using my old Sparkle and system is
stable and cool.

Bob
Click to expand...

Heh, my old one is a sparkle, too - but the fan is really loud :-(

kad
 
Check the motherboard for any caps that are domed on top, or are leaking
any brown or black substance.

You can go to this site http://www.motherboardrepair.com/ for pictures
of bad caps, and information on repairing the board.

Doug
Click to expand...

I didn't see leaks of any kind from any of the caps on the mobo nor on
the peripheral cards - they all looked clean (aside from a little
dust).

Since writing my first post, I added another side blowing fan to the
Zalman FM123 bracket next to the video card. Since then, it hasn't
crashed randomly. I decided to run a 'heat test' (multiplayer
Battlefield 1942 w/ desk door closed over the cpu compartment). I
CTD'd a couple times when the temp (reading from the BIOS using MBM 5)
when the case temp hit 44C. Before moving everything into this case,
I was able to get to 49-52C before it would crash. I did have one
infinite loop on the video driver (nvidia reference driver v30.82).

Just when it looks like one problem, the pendulum swings the other way
and it looks like something else :-(

kad
 
I just RMA'd an Antec TruePower 480. Made the system unstable and
smelled like it was burning up. It also got really hot and made my
system temps skyrocket. I've heard great things about Antec and that's
why I tried one. I went back to using my old Sparkle and system is
stable and cool.

Bob
Click to expand...

Well, the air coming out of it is much warmer than the air coming out
of the fan right below it (just feeling it, it seemed like a 20F
difference). no burnt smell, but it does smell like a very warm
monitor. This weekend (when I have more time) I'll hook up my old
Sparkle to it and see what happens.

kad
 
[snip]
Since writing my first post, I added another side blowing fan to the
Zalman FM123 bracket next to the video card. Since then, it hasn't
crashed randomly. I decided to run a 'heat test' (multiplayer
Battlefield 1942 w/ desk door closed over the cpu compartment). I
CTD'd a couple times when the temp (reading from the BIOS using MBM 5)
when the case temp hit 44C. Before moving everything into this case,
I was able to get to 49-52C before it would crash. I did have one
infinite loop on the video driver (nvidia reference driver v30.82).

Just when it looks like one problem, the pendulum swings the other way
and it looks like something else :-(
Click to expand...

So much for that - it just BSOD'd at 37C (I was reading email). :-(
 
[snip]
Since writing my first post, I added another side blowing fan to the
Zalman FM123 bracket next to the video card. Since then, it hasn't
crashed randomly. I decided to run a 'heat test' (multiplayer
Battlefield 1942 w/ desk door closed over the cpu compartment). I
CTD'd a couple times when the temp (reading from the BIOS using MBM 5)
when the case temp hit 44C. Before moving everything into this case,
I was able to get to 49-52C before it would crash. I did have one
infinite loop on the video driver (nvidia reference driver v30.82).

Just when it looks like one problem, the pendulum swings the other way
and it looks like something else :-(
Click to expand...

So much for that - it just BSOD'd at 37C (I was reading email). :-(
Click to expand...

It BSOD'd a few more times this evening (different each time). Also,
now MBM is having wild fluctuations in its readings - the temp ranged
from 0-79C and all voltages ranged from 0 to about double the named
voltage (i.e., the 3.3 was showing from 0-6.6 or so).

This is really weird - it was good for 3 days (only crashing during my
"stress" tests) and now it's crashing a lot. The pc has been on for
about 6 hours now. In the first 2.5-3 hours it crashed 3 times
(probably would have crashed more except I was out & it crashed while
I was gone). Since the last one, though, it hasn't crashed.

kad
 
It BSOD'd a few more times this evening (different each time). Also,
now MBM is having wild fluctuations in its readings - the temp ranged
from 0-79C and all voltages ranged from 0 to about double the named
voltage (i.e., the 3.3 was showing from 0-6.6 or so).

This is really weird - it was good for 3 days (only crashing during my
"stress" tests) and now it's crashing a lot. The pc has been on for
about 6 hours now. In the first 2.5-3 hours it crashed 3 times
(probably would have crashed more except I was out & it crashed while
I was gone). Since the last one, though, it hasn't crashed.
Click to expand...

Why don't you send an e-mail to '(e-mail address removed)' He is the guy with this
web page http://www.motherboardrepair.com/ and see what he says about
these problems. The board you have is known to have the defective caps
that flooded the market a few years ago. Just because you don't see
anything leaking from them does not guaranty they are good.

Doug
 
I just RMA'd an Antec TruePower 480. Made the system unstable and
smelled like it was burning up. It also got really hot and made my
system temps skyrocket. I've heard great things about Antec and that's
why I tried one. I went back to using my old Sparkle and system is
stable and cool.

Bob
Click to expand...

OK, definitely not the power supply - I swapped out the Antec for my
Sparkle and no change.

Interestingly, I noticed that if I run it outside of the desk with the
cover off, it doesn't crash and the temp didn't go higher than 38C (it
crashes at around 42-43C).

Just for fun, I took out the Zalman fan bracket (with a fan over the
cpu & a fan over the vidcard - both 92mm) and the system crashed very
quickly and at a relatively low temp (34C IIRC).

kad
 
Why don't you send an e-mail to '(e-mail address removed)' He is the guy with this
web page http://www.motherboardrepair.com/ and see what he says about
these problems. The board you have is known to have the defective caps
that flooded the market a few years ago. Just because you don't see
anything leaking from them does not guaranty they are good.

Doug
Click to expand...

I'll do that - thanks for the suggestion.

kad
 
Back
Top