New Build Keeps Rebooting

[Bill Woessner]

I put this build together 2 days ago. It starts up and will run for a
little while. But it reboots seemingly randomly. Initially, I
thought it was overheating. But when I check the CPU temp after
reboot, it's always in the 40-50C range.

Here's the build:

Cooler Master Centurion 5 case
Fortron AX450-PN power supply
MSI K9N4 Ultra-F motherboard
AMD Athlon 64 X2 4000+ (Brisbane) CPU
WinTec AMPX 1GB DDR2-800 DIMM
Western Digital 160 GB SATA hard drive
Sony/NEC 18x SATA DVD burner
Sapphire Radeon x1950 Pro (PCI-E)

I took everything apart and put it back together. Didn't help. I
upgraded the BIOS. Didn't help. I adjusted the CPU and RAM voltage
settings. Didn't help. At this ponit, I'm totally stumped. Maybe
the power supply might not be beefy enough to handle the graphics
card?

I know the Brisbane cores have to use odd clock multipliers. Is it
possible that this motherboard has trouble with the Brisbane core?
One NewEgg reviewer suggested the BIOS doesn't know how to set the
voltage for the Brisbane core, which is why I set it manually.

Any advice? I appreciate any help you can give me. Also, if anyone
has experience with NewEgg's return policy, I'd like some advice on
that. I suspect I might be returning the whole mess.

Thanks in advance,
Bill

 

[meerkat]

I put this build together 2 days ago. It starts up and will run for a
little while. But it reboots seemingly randomly. Initially, I
thought it was overheating. But when I check the CPU temp after
reboot, it's always in the 40-50C range.

Here's the build:

Cooler Master Centurion 5 case
Fortron AX450-PN power supply
MSI K9N4 Ultra-F motherboard
AMD Athlon 64 X2 4000+ (Brisbane) CPU
WinTec AMPX 1GB DDR2-800 DIMM
Western Digital 160 GB SATA hard drive
Sony/NEC 18x SATA DVD burner
Sapphire Radeon x1950 Pro (PCI-E)

I took everything apart and put it back together. Didn't help. I
upgraded the BIOS. Didn't help. I adjusted the CPU and RAM voltage
settings. Didn't help. At this ponit, I'm totally stumped. Maybe
the power supply might not be beefy enough to handle the graphics
card?

I know the Brisbane cores have to use odd clock multipliers. Is it
possible that this motherboard has trouble with the Brisbane core?
One NewEgg reviewer suggested the BIOS doesn't know how to set the
voltage for the Brisbane core, which is why I set it manually.

Any advice? I appreciate any help you can give me. Also, if anyone
has experience with NewEgg's return policy, I'd like some advice on
that. I suspect I might be returning the whole mess.

Check youe Event Viewer, to eliminate any Apps or
System failures.

 

[Michael Hawes]

I put this build together 2 days ago. It starts up and will run for a
little while. But it reboots seemingly randomly. Initially, I
thought it was overheating. But when I check the CPU temp after
reboot, it's always in the 40-50C range.

Here's the build:

Cooler Master Centurion 5 case
Fortron AX450-PN power supply
MSI K9N4 Ultra-F motherboard
AMD Athlon 64 X2 4000+ (Brisbane) CPU
WinTec AMPX 1GB DDR2-800 DIMM
Western Digital 160 GB SATA hard drive
Sony/NEC 18x SATA DVD burner
Sapphire Radeon x1950 Pro (PCI-E)

I took everything apart and put it back together. Didn't help. I
upgraded the BIOS. Didn't help. I adjusted the CPU and RAM voltage
settings. Didn't help. At this ponit, I'm totally stumped. Maybe
the power supply might not be beefy enough to handle the graphics
card?

I know the Brisbane cores have to use odd clock multipliers. Is it
possible that this motherboard has trouble with the Brisbane core?
One NewEgg reviewer suggested the BIOS doesn't know how to set the
voltage for the Brisbane core, which is why I set it manually.

Any advice? I appreciate any help you can give me. Also, if anyone
has experience with NewEgg's return policy, I'd like some advice on
that. I suspect I might be returning the whole mess.

Thanks in advance,
Bill

Run memtest86 to test the memory. Check all PSU cables are fuly engaged,
if you tug them, they should stay in place. Have you connected ALL required
power cables? Check Event Viewer for clues. PSU is good enough, but could be
faulty. Stop windows Restarting on error, you may then get a BSOD.

Mike.

 

[SirWillae]

Run memtest86 to test the memory. Check all PSU cables are fuly engaged,
if you tug them, they should stay in place. Have you connected ALL required
power cables? Check Event Viewer for clues. PSU is good enough, but
could be faulty. Stop windows Restarting on error, you may then get a
BSOD.

Tried all that. memtest86 reported no errors. There's nothing in the
Event Viewer, either. I tried 3 different power supplies and got the
same results with all 3. Also, I noticed that, after SOME reboots
(but not all), I get a POST message saying something like "Hyper
transport sync flood on previous boot". Not sure what that's all
about. I Googled it and got ambiguous results.

So I swapped out the video card. I had an old ATI X300 card (PCI-E
but doesn't require its own power) sitting around. So far, I haven't
had any random reboots. That sort of indicates a PSU problem, but,
like I said above, I've tried 3 different PSUs, one of which powers a
similar system with no problems.

Now I'm totally stumped. The mysterious hyper transport error
combined with the video card issues are leading me to believe the
problem is with the motherboard. At tihs point, the system is up and
runnign and stable, but keeping the X300 card in this system is not
acceptable.

Any suggestions?

Thanks in advance,
Bill

 

[Michael Hawes]

Tried all that. memtest86 reported no errors. There's nothing in the
Event Viewer, either. I tried 3 different power supplies and got the
same results with all 3. Also, I noticed that, after SOME reboots
(but not all), I get a POST message saying something like "Hyper
transport sync flood on previous boot". Not sure what that's all
about. I Googled it and got ambiguous results.

So I swapped out the video card. I had an old ATI X300 card (PCI-E
but doesn't require its own power) sitting around. So far, I haven't
had any random reboots. That sort of indicates a PSU problem, but,
like I said above, I've tried 3 different PSUs, one of which powers a
similar system with no problems.

Now I'm totally stumped. The mysterious hyper transport error
combined with the video card issues are leading me to believe the
problem is with the motherboard. At tihs point, the system is up and
runnign and stable, but keeping the X300 card in this system is not
acceptable.

Any suggestions?

Thanks in advance,
Bill

1950 is a pretty demanding board. All 3 PSUs may not be upto the job.
Need 450W, good brand to be sure.

Mike.

 

[Bill Woessner]

1950 is a pretty demanding board. All 3 PSUs may not be upto the job.
Need 450W, good brand to be sure.

Except that one of the PSUs I tried can run the exact same X1950. I
took them both from another computer I have. Admittedly, that other
computer is a 3.0 GHz P4 HT. Not sure how much power that requires
compared to my new build. I thought the older HT P4s were total power
hogs, though.

--Bill

 

[Paul]

Except that one of the PSUs I tried can run the exact same X1950. I
took them both from another computer I have. Admittedly, that other
computer is a 3.0 GHz P4 HT. Not sure how much power that requires
compared to my new build. I thought the older HT P4s were total power
hogs, though.

--Bill

Did you examine the CPU support table for the product, and see
what BIOS revision is the minimum to support your processor ?

Is Cool N' Quiet disabled ? By disabling that, you'd stop any
dynamic VID changes, and the processor should operate at the
voltage you set in the BIOS. The voltage will still vary,
as processors are designed to operate on a load line. A
typical figure for the amount of droop due to loading,
might be 0.15V less voltage, when the processor is running
a 100% computing load.

I would try to verify, by using whatever hardware monitor
is available, as to what the Vcore is when in Windows
(idle or busy). If the idle voltage is supposed to be 1.35V,
maybe under load it will drop to 1.25V or a bit less.

Does the computer "fall over" when you run Prime95 or Orthos ?
Does it reboot or shut off, just as this test starts ?
Give this a try.

http://sp2004.fre3.com/beta/beta2.htm

Orthos is also a good test for the memory, and seems to detect
memory problems better than memtest86+. The main value of
memtest86+, is finding stuck bits. Orthos and Prime95 cannot
test the entire memory, and only cover the areas not occupied
by the OS. In that sense, the two test methods are complimentary,
and both kinds of testing have value (memtest and one of the others).

Paul

 

[Michael Hawes]

Except that one of the PSUs I tried can run the exact same X1950. I
took them both from another computer I have. Admittedly, that other
computer is a 3.0 GHz P4 HT. Not sure how much power that requires
compared to my new build. I thought the older HT P4s were total power
hogs, though.

--Bill

May be time to test out of the case. Place motherboared on piece of
cardboard or other non-conducting surface and test. Check that you don't
have standoff pillar in incorrect position, casusing a short.

Mike.

 

[Bill Woessner]

May be time to test out of the case. Place motherboared on piece of
cardboard or other non-conducting surface and test. Check that you don't
have standoff pillar in incorrect position, casusing a short.

Heh. Great minds think alike. I tried that a long time ago.
Besides which, that wouldn't explain why it works with the X300 but
not the X1950.

I finally sent the motherboard back and got a different one. Between
the "hyper transport flood sync", the mysterious floppy drive showing
up in Windows and the onboard audio not working (not to mention the
random reboots) I figured there was enough wrong with the motherboard
to switch it out. I'm replacing it with the GigaByte GA-M61P-S3.
Probably should have gone that way to begin with.

Thanks to everyone for all the suggestions. Hopefully the new
motherboard will fix all the problems. If not... well, I hate to
think I have a defective CPU. Wish me luck!

--Bill

 

[Mac Cool]

:

I tried 3 different power supplies and got the
same results with all 3.

So I swapped out the video card. I had an old ATI X300 card (PCI-E
but doesn't require its own power) sitting around. So far, I haven't
had any random reboots. That sort of indicates a PSU problem

More likely it indicates a faulty video card unless you are testing with
weak PSU. RMA the 1950.

 

[ProfGene]

I put this build together 2 days ago. It starts up and will run for a
little while. But it reboots seemingly randomly. Initially, I
thought it was overheating. But when I check the CPU temp after
reboot, it's always in the 40-50C range.

Here's the build:

Cooler Master Centurion 5 case
Fortron AX450-PN power supply
MSI K9N4 Ultra-F motherboard
AMD Athlon 64 X2 4000+ (Brisbane) CPU
WinTec AMPX 1GB DDR2-800 DIMM
Western Digital 160 GB SATA hard drive
Sony/NEC 18x SATA DVD burner
Sapphire Radeon x1950 Pro (PCI-E)

I took everything apart and put it back together. Didn't help. I
upgraded the BIOS. Didn't help. I adjusted the CPU and RAM voltage
settings. Didn't help. At this ponit, I'm totally stumped. Maybe
the power supply might not be beefy enough to handle the graphics
card?

I know the Brisbane cores have to use odd clock multipliers. Is it
possible that this motherboard has trouble with the Brisbane core?
One NewEgg reviewer suggested the BIOS doesn't know how to set the
voltage for the Brisbane core, which is why I set it manually.

Any advice? I appreciate any help you can give me. Also, if anyone
has experience with NewEgg's return policy, I'd like some advice on
that. I suspect I might be returning the whole mess.

Thanks in advance,
Bill

I had an Athlon 64 system that rebooted all the time and I happened to
have some RAM that I had bought without knowing it was overclocking RAM
and it didn't work on a dual core system I had built but when I put it
in the Athlon 64 machine it started working like a charm and has never
given me any trouble since. When I called up Kingston they told me that
the RAM I had been sold was mainly used for builders who liked to
overclock. I have never overclocked and only tried the RAM on a hunch.