Hardware that goes boom!

[Talas]

I've got a little issue and before I strike again I wanna make sure I understand what happend... Any input is appreciated!

I'm in the process of making a minor upgrade. I had ordered a new GeForce 6800 and XP 2800+ (Barton core) to upgrade my Ti4600 and XP1700+. Since my PS was only a Sparkle 300W, I also purchased a new Antec TRU430 PS as recommended by the 6800. So we're setup to make the upgrades.... now.

I swap everything out, corrupt the CMOS, boot, change the BIOS config for the new processor (no OC'ing or OV'ing) and reboot. The system start to come up with Windows and it shuts down. Odd I'm thinking... the 430W supply should be plenty, so I boot it again. Same thing. I swap the video cards out and try again with the same result (auto shut down while booting Windows). I go into the BIOS and note that the processor is running a little hot (85C), so I exit and the system reboots... ONLY THIS TIME, I hear something that sounds like a capacitor exploding !!!!

After a few explatives, I inspect the MB and don't see any busted caps. Long story shorter.... I've got a dead processor with a chunk missing from the die and even with the old processor the system won't boot so I'm guessing a dead MB as well.

What caused this ?

I'm thinking that it was either
(a) power supply was bad
(b) using the CPU heatsink and fan from XP1700 wasn't adequate for the XP2800

My guess is (b) but not seeing any blown traces, I'm not sure why it would have taken out the MB as well. If the PS was bad, I could see a way out of tolerance voltage doing something like this... but I dunno for sure.

Any thoughts ? Anyone ?

 

[pelikan]

The heatsink should have been fine to boot up with. Maybe it wouldn't have been powerful enough under load.

 

[imported_rod]

That PSU has way more than enough power - no way it would have had trouble, unless it was devective (possible, but unlikely).

The HSF should have been enough, maybe not for long-term use, but enough for a boot up. Was the fan plugged in? Did you apply thermal paste?

Maybe take everything out of the case, disassemble and reasemble it, and try with the other processor. I dunno...

RoD

 

[Talas]

Yeah, that was my initial thought. Although, 85C while doing nothing more the sitting the in BIOS was a bit suprising. I'm now thinking that the auto shutdowns were probably done by the CPU Overheat Protection on my A7N8X.... but I don't have an explaination on why it was running above avg temps. XP2800 (2.083 GHz spec'd) was clocked @ 12.5x 166MHz so it wasn't a clock rate issue. The only other things I can imagine are voltages well out of spec or defective processor (if it wasn't before, it is now =/)

 

[Talas]

Originally posted by: rod
That PSU has way more than enough power - no way it would have had trouble, unless it was devective (possible, but unlikely).

The HSF should have been enough, maybe not for long-term use, but enough for a boot up. Was the fan plugged in? Did you apply thermal paste?

Maybe take everything out of the case, disassemble and reasemble it, and try with the other processor. I dunno...

RoD

The HSF was applied with Artic Silver 5... I did manage to check and verify that the HSF was running when I noticed the temps in the BIOS. My inital thought was that I had forgotten to plug the fan in.

Even with the old PS + old proc + MB only (no video, HDs, etc), the system doesn't power up (i.e. PS doesn't get the 'on' signal from the MB).

I'm at least looking at a new MB + CPU... but not sure if the new PS is the culprit or if it's safe to use again.

 

[pcgeek11]

Originally posted by: Talas

Originally posted by: rod
That PSU has way more than enough power - no way it would have had trouble, unless it was devective (possible, but unlikely).

The HSF should have been enough, maybe not for long-term use, but enough for a boot up. Was the fan plugged in? Did you apply thermal paste?

Maybe take everything out of the case, disassemble and reasemble it, and try with the other processor. I dunno...

RoD

The HSF was applied with Artic Silver 5... I did manage to check and verify that the HSF was running when I noticed the temps in the BIOS. My inital thought was that I had forgotten to plug the fan in.

Even with the old PS + old proc + MB only (no video, HDs, etc), the system doesn't power up (i.e. PS doesn't get the 'on' signal from the MB).

I'm at least looking at a new MB + CPU... but not sure if the new PS is the culprit or if it's safe to use again.

Not to insult you, But are you sure the heat sink was not on backwards? or maybe the CPU die was chipped when installing the HSF. These are the two most common causes of new CPU failures other than not plugging in the fan. The cutout in the bottom of the HSF goes towards the hinge of the locking lever. The corners of the CPU die are very fragile.

pcgeek11

 

[Texun]

As PCGEEK said..... not to insult you, but the HS you used on the 1700 should have been enough to keep the 2800 well below 85C if it was installed correctly. I'm not saying it was or wasn't enough of a HS for long term use, but even if it were an OEM HS it should have kept it below 85.

A real easy mistake to make during install is to allow the HS to ride up on the front of the cpu socket, near the ridge, just enough to prevent even contact. That will skyrocket the temps in a matter of seconds.

I doubt that it was the PSU. Did you notice anything odd with the voltages when you were in the BIOS checking the temp?

 

[flexy]

did you apply AS5 right (according to specs on arcticsilver website)...did you mount the HSF right ?

either your VCore voltages etc. where WAY out of specs or you messed something up installing the HSF....not seated right, or too much AS5 (more than a tiny grain)...or you got AS5 where you were not supposed to.....

Or CPU pins were bent/cut short/missing.....or anything the like...

 

[Budman]

I have to agree with others here, looks like you mounted the heatsink improperly.

for future reference this is the correct way to install http://www.omnicast.net/~tmcfadden/guides/Heatsink/index.html