Telltale signs of a dead chip?

[HeaterCore]

I'm pretty sure I blew out my old 2800 Barton, which had been running at 2500MHz @ 1.95v for months. My system just died in the middle of a game, and now hitting the power switch does nothing but start a two-tone alarm, which continues until I cut the power entirely.

Anyway, the PSU checks out, I've tried two different sticks of RAM, and I've even unplugged the video card. That leaves the chip and the motherboard. Any way to tell which bit the bullet, short of dropping in a new chip (which I don't have)?

There are no scorch marks on the little ah heck, but I'm guessing that's far from conclusive?

-HC-

 

[CheesePoofs]

Whats your motherboard, and what does the alarm mean? (there should be a key somewhere in your manuel for all of the different alarms)

 

[HeaterCore]

Should've specified earlier:

NF7-S Rev.2 (Manta Ray's XT D26)
2800 Barton (200x12.5 @ 1.95v)
9800 Pro @ 465/380
Danger Den TDX/Maze4 GPU
HyperX DDR333 @ 400 (2.5-3-3-7)
36.7GB Raptor
2x80GB WD PATA

The siren means one of two things, overheating or out-of-spec voltages. It's not overheating, or it would've done so much earlier than this. My cooling rig keeps the chip at ~42C idle. Obviously the voltages were way out of whack, but I've cleared CMOS multiple times and by multiple methods and so all settings should be back at default.

My best guess is that the chip fused a few circuits and now overheats instantly, no matter what current is run through it. But wouldn't that be at least minimally visible?

-HC-

 

[MDE]

I'd say it was too much voltage. Pony up the cash for a Mobile XP and you'll probably be running 2.5GHz in no time.

 

[imported_whatever]

Originally posted by: MDE
I'd say it was too much voltage. Pony up the cash for a Mobile XP and you'll probably be running 2.5GHz in no time.

agreed. at 1.95v on a Maze4 I could run almost 2.6GHz.

 

[superkdogg]

I take it you didn't smell anything, huh? Usually burning up a CPU is a sensory experience. Do you have a friend with an AXP rig he'd let you 'borrow' for a 1/2 hr. to test with? I'm thinking it could be possible that mobo circuitry or caps let go from providing all that voltage. Probably a CPU, but could be either-you need parts to troubleshoot with.

 

[RichUK]

Originally posted by: superkdogg
I take it you didn't smell anything, huh? Usually burning up a CPU is a sensory experience. Do you have a friend with an AXP rig he'd let you 'borrow' for a 1/2 hr. to test with? I'm thinking it could be possible that mobo circuitry or caps let go from providing all that voltage. Probably a CPU, but could be either-you need parts to troubleshoot with.

defo .. if you dont want to by another CPU

 

[sangyup81]

You probably did this but since you didn't mention it, I'll ask. Did you reset the BIOS? And did you use both the jumper and the battery method?

Given my experiences, I would think the MB is the one to go out first. The MB is the one with the voltage regulator afterall.

 

[HeaterCore]

Yeah, I switched the jumper, pulled the battery, flipped the power switch, you name it. Always a good idea to cover all those bases. It's amazing most boards don't ship with a foolproof way to clear CMOS, flip a switch and you're done.

Unfortunately, I don't have a spare CPU or mobo lying around, and my buddies have upgraded long since (while I dropped the cash on a watercooling rig). And I don't want to pick up a new CPU without knowing that mine is actually dead.

Since the CPU looks fine, and -- apparently -- it's usually pretty easy to tell when they've blown, it's time to pull out the mobo and give it a close-up scan with a penlight. What should I look for? Burnt ICs, busted caps, scorch marks?

And if I can't figure out the problem, I'm thinking it doesn't make sense to pony up a hundred bucks for a new CPU without any assurance that I'm just buying a second functional chip for a busted board. Clearly the only sensible thing to do would be to drop the extra cash for a new A64 chip and mobo. And no point in sticking with AGP either, right? And the only reasonable PCI-E card for my setup at the moment is an X800 XL, so I'd have no choice but to add that to the mix. Hmmm. What a disaster that would be.

The wife will be appalled. 😀

-HC-

 

[MDE]

There's nothing wrong with AGP. I'm running an nForce3 Ultra board with a 3000+ @ 2.5GHz with a 6800GT and I'd bet it feels no different from a comparable nForce4 setup. If your 9800 Pro is still running AGP is still a very viable option.

 

[HeaterCore]

There's nothing wrong with AGP performance-wise, definitely. But I'm a compulsive video-card upgrader, a new one every year or so, and I think we've seen the end of the line for AGP. Maybe R520 and nVidia's next-gen will come out for AGP, but I wouldn't bet on it, and even if they do they won't show up on the shelves until several months after the initial launch. I just figure that if I'm going to upgrade platforms, might as well go full-bore.

-HC-