PC crashes when CPU is at spec speeds

  • Thread starter Thread starter Christopher Rawlison
  • Start date Start date
C

Christopher Rawlison

Hey, odd question here.

I have a system with the following hardware:

Motherboard: DFI NF-II Ultra-AL
Processor: AMD Athlon XP 2100+ (Palomino core - 133MHz, 13x multiplier)
RAM: Kinsgston ValueRam 512mb, DDR333

Now here's where it gets odd.

The computer boots fine with the FSB at 100mhz, 13x multiplier, 66mhz
AGP bus, memory at 166mhz. This essentially gives me an Athlon XP
1500+. Motherboard reports a CPU temp of 49-50 Celsius.

When I change the FSB to 133mhz however, the motherboard reports a CPU
temp of about 55 Celsius. Windows crashes mid boot. I'll get through
the POST stuff fine, but Windows takes a dump.

Now I know this CPU works at 133FSB as in my old motherboard it was
fine. Unfortunately I broke off a cap swapping a PCI card so I had this
motherboard sitting around and did a fresh WinXP install. I don't
*need* it to run as a 2100, but its going to be the house NAT/file/print
server so I would like to have the extra speed if possible.

Any ideas why Windows it taking a dump at the higher FSB?
 
Christopher said:
When I change the FSB to 133mhz however, the motherboard reports a CPU
temp of about 55 Celsius. Windows crashes mid boot. I'll get through
the POST stuff fine, but Windows takes a dump.
Click to expand...

What exactly is it doing. Restarting? Blue screening? Freezing?
Did you properly reseat the heatsink and reapply thermal compound?
What's the temperature of the CPU /after/ a 'crash'? Have you tried
booting into safe mode?

Ari

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/
 
spodosaurus said:
What exactly is it doing. Restarting? Blue screening? Freezing?
Did you properly reseat the heatsink and reapply thermal compound?
What's the temperature of the CPU /after/ a 'crash'? Have you tried
booting into safe mode?
Click to expand...

First off, yes there is a good joint between heatsink and CPU, using
Arctic Silver. I forgot to mention I'm running SP2. Regarding
temperatures, it never seems to break mid 50s C for a core temp,
although the only time I can get a read is during the boot process. As
far as the behavior of the system, it breaks down like this:

I change from 13x100Mhz to 13x133Mhz and when the computer boots, it
will go to the WinXP logo screen (with the scrolling blue progress bar)
and then it will reboot. Upon rebooting regardless of if I attempt safe
mode or normal boot sequence it will hang.

If I boot into safe mode, it will hang on the driver for the Iomega disk
manager (iomdisk.sys) for my zip drive. If I tell it to boot normally,
it will hang and return the message: "Windows could not start because
the following file is missing or corrupt: <Windows
Root>\system32\ntoskrnl.exe"

One thing that has me a bit confused is that whenever I switch from
13x100Mhz to the 13x133Mhz, the raid array (mirrored) on my Silicon
Image card has to get rebuilt. No data loss, but the array needs to be
rebuilt. This is making me think that there's some type of issue with
the PCI bus running out of spec, but the nForce 2 chipset is supposed to
have a lock on the PCI bus speed.
 
First off, yes there is a good joint between heatsink and CPU, using
Arctic Silver. I forgot to mention I'm running SP2. Regarding
temperatures, it never seems to break mid 50s C for a core temp,
although the only time I can get a read is during the boot process.
Click to expand...
Sounds way too hot to me. Unless your case temp is rediculous, your
heatsink isn't on good.
 
Chris said:
Sounds way too hot to me. Unless your case temp is rediculous, your
heatsink isn't on good.
Click to expand...

No, that seems normal for these CPUs even with good case ventilation. I
have my 2000+ in an ASUS board and it's running at 60C despite me having
three case fans (including one blowing are through a tunnel right onto
the CPU's fan). My 2400+ and 2800+ don't run as hot with poorer airflow.

Ari

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/
 
No, that seems normal for these CPUs even with good case ventilation. I
have my 2000+ in an ASUS board and it's running at 60C despite me having
three case fans (including one blowing are through a tunnel right onto
the CPU's fan). My 2400+ and 2800+ don't run as hot with poorer airflow.
Click to expand...


It has been a while, but I don't think any of the builds I did in that
generation got that hot. I'd have done something about it if they
did. I never had to resort to more than one case fan.
 
Chris said:
It has been a while, but I don't think any of the builds I did in that
generation got that hot. I'd have done something about it if they
did. I never had to resort to more than one case fan.
Click to expand...

I didn't resort to it, the case had room and I had extra vantec stealth
fans laying around, so I used them.

Ari

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/
 
Chris said:
Sounds way too hot to me. Unless your case temp is rediculous, your
heatsink isn't on good.
Click to expand...

I have a 120mm case intake and a 120mm case exhaust fan, Volcano 9
heatsink with a 80mm fan ducted to draw air directly from outside the
case. The Palomino core was actually one of the hottest chips that AMD
ever put out.

Anyway, I still think its something relating to the PCI bus somehow
running above 33mhz, but I have no way to verify this.
 
Christopher Rawlison said:
Hey, odd question here.

I have a system with the following hardware:

Motherboard: DFI NF-II Ultra-AL
Processor: AMD Athlon XP 2100+ (Palomino core - 133MHz, 13x multiplier)
RAM: Kinsgston ValueRam 512mb, DDR333

Now here's where it gets odd.

The computer boots fine with the FSB at 100mhz, 13x multiplier, 66mhz AGP
bus, memory at 166mhz. This essentially gives me an Athlon XP 1500+.
Motherboard reports a CPU temp of 49-50 Celsius.

When I change the FSB to 133mhz however, the motherboard reports a CPU
temp of about 55 Celsius. Windows crashes mid boot. I'll get through the
POST stuff fine, but Windows takes a dump.
Click to expand...

Chris, seems I recall some of those Palo's running better at 142 fsb
than at 133. Dunno if your DFI supplies it, but if so you may find this
works for you. Also, I think they ran better using synchronous memory
settings, i.e. 133 vs your 166 mhz. HTH, s
 
sdlomi2 said:
Chris, seems I recall some of those Palo's running better at 142 fsb
than at 133. Dunno if your DFI supplies it, but if so you may find this
works for you. Also, I think they ran better using synchronous memory
settings, i.e. 133 vs your 166 mhz. HTH, s
Click to expand...
Run a memory test. I once had som PC133 memory which worked when new,
but later started crashing. It worked fine at 100, but failed at 133.

Mike.
 
Christopher said:
I have a 120mm case intake and a 120mm case exhaust fan, Volcano 9
heatsink with a 80mm fan ducted to draw air directly from outside the
case. The Palomino core was actually one of the hottest chips that AMD
ever put out.
Click to expand...

That's very similar to what I'm running except I have a fan at the
opening of the duct right to the fan on the AMD HSF unit. Still 60C most
of the time.
Anyway, I still think its something relating to the PCI bus somehow
running above 33mhz, but I have no way to verify this.
Click to expand...

Do you know any ways the OP might be able to verify it?

Ari



--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/
 
spodosaurus said:
That's very similar to what I'm running except I have a fan at the
opening of the duct right to the fan on the AMD HSF unit. Still 60C most
of the time.

Do you know any ways the OP might be able to verify it?
Click to expand...


Well its about the best you can do without going liquid cooling, but I'm
the OP. ;) I remember they used to make pci cards that would display
what frequency they were running at, but I don't even know where to get one.
 
Back
Top