A tale of PC death

[jo blo]

Any theories or experience with the following would be helpful:

 

I just put together my 2nd build ever (first was a DFI Ultra D that this forum helped me make work), an asus A8N-VM CSM based HTPC. For the processor, I used a two year old Athlon 64 3500+ that was first in my main rig for a year, then replaced and stuck in a box (full specs for this new system at the end of this article). I got it all wired up and bolted down by Tuesday night. By Wednesday night she was up and singing. By Thursday night, dead dead dead.

 

Here's what happened: When I powered it on after work on Thurs, there was a continuous high pitched tone from around the CPU fan, and when I put my finger on the fan and stopped it, the tone disappeared. Nothing came up on the screen, and she didn't boot. Powered down, powered back on. The tone came back, then went away, and the system began to boot, and made it to the Windows green progress bar screen. Here the whine came back sporadically. When it did, the progress bar would freeze. When the tone went away, the progress bar would resume. Then blam, system is off. Nothing on screen, no fans moving, but the LED on the board is on (there's only one LED on this board, showing that the PSU is delivering power). I power it up again. Same behavior during the windows progress screen, but this time it makes it all the way into the OS. There is nothing out of the ordinary in the system log. I start messing around, and after about 10 minutes, WHAMMO, system down. No warning, no tone this time, just system off. It has not even booted to bios since. When I power it on, all systems get juice, the fans spin up, the drives chatter for a sec, but nothing ever comes up on the monitor (the monitor remains in suspend).

 

I cleared the bios short, I cleared the bios long (24 hrs) with the battery removed, I took out a stick of ram, and I plugged in a different power supply, still no joy. I unplugged everything from the board but the keyboard and monitor, no improvement. Different monitor, no improvement. I actually ended up with two boards for this build, one refurbished and one retail. This was with the refurbished board, and that was the first suspect. I replaced it with the brand new board, still no luck. I pulled a stick of known good RAM out of my main system, that I'm now typing on, and tried it in each of the four slots, with a bios clear in between each, she just stayed dead. So, trying it with a different PSU, different MOBO, different RAM, two different monitors (one DVI, one VGA) does not help. The only thing that remains the same is the CPU, so I'm hoping that's it, except the way if failed has me spooked. That tone suggests to me a failing power component, or something being driven in an unnatural way, like all of a sudden there was a lot of interference or harmonics on the voltage rails or something. PSU #1 was the one that came with the Ahanix case. They make some good PSU's, so I didn't feel bad about using this one, but maybe it got rabies and killed board #1, then killed board #2 as soon as I plugged it in (I tried board #2 with both PSUs). Is that possible? Or is this really just a clear case of a bad CPU? I never overclocked it, never dropped it, and in storage it was in it's original AMD package. I did bend a couple pins when I removed it from the main machine, but I bent them back nicely so that it just slipped in the socket. It's just that in spite of the evidence, I trust the CPU more than I trust some of the other parts. And what the hell was making the fan make that noise? The board does not come with a buzzer or diagnostic LEDs, and I only plugged a speaker in after the thing died, and it never beeped on any subsequent power-up.

 

Here's the original setup:

ASUS A8N-VM CSM nVidia 6150 mATX motherboard, refurbished

AMD 3500+ (single core), used 1 year, stored 1 year

Zalman CNPS8000 CPU heatsink

Crucial value ram 2x512, also used 1 year, stored 1 year

Seagate 7200.8 300GB parallel IDE HD

Pioneer 112D DVD burner

Ahanix MCE301 case with Ahanix 350 PSU

Ahanix/iMon VFD front panel display, powered with a floppy connector an plugged into a USB header on the MOBO

 

Here are the replacements:

New A8N-VM CSM motherboard

2 year old, 1 year stored Seasonic S12-430 PSU

Geil 1GB stick of RAM from main system

Stock Heatsink/fan

[stepbackjack]

I did bend a couple pins when I removed it from the main machine, but I bent them back nicely so that it just slipped in the socket

 

This and the supply choices you have made.

 

:sad:

[red930]

This and the supply choices you have made.

 

:sad:

 

What's wrong with that? I've bent pins on all three of my 939-based processors and bent them back with a pencil... to date, all three still work as they are being used by me, my father, and my sister. Bent pins don't impede a chip's function.

 

His supply choices seem fine to me. With the exception of the PSU (which to me should have been fine anyway), everything seems in order.

 

@ jo blo: This may not be much of a recommendation, but if you're pretty sure it's the CPU you could try RMA'ing it back to AMD.

[dr_bowtie]

agreed...

 

test the PSU and the CPU...could be either...the giveaway was swaping the board did no good...

[AnUnknownSource]

About ten years ago I sent three different motherboards packing before I realised the PSU I was using was the culrit... It took board 2 and 3 out as soon as I powered on... (luckily nothing else went with them!) replaced the PSU and tested all three boards and found them dead, but the 4th board with the new PSU and all the rest of the components were fine... I'd check that PSU out on a PSU tester if I were you...

[Clay]

Have you considered. possible board shorting? Perhaps the boards are grounding somewhere? (All stand-offs and paper washers in their proper place ect,ect)

Long shot but maybe....

[bldegle2]

i have had fans that have fubared so badly before it tanked the OS, once replaced, everything went back to normal.....

 

here is the clue that lead me to this conclusion...

 

"and when I put my finger on the fan and stopped it, the tone disappeared."

 

this is a classic example of interference created by the bad CPU fan, causing the system to flake........

 

your PS choice is fine power wise.

 

had case fans do this to my system and i have had several customer's rigs that have had this happen over the years, simple fix, just annoying as h3ll.....

 

laterz

[wevsspot]

Hopefully a malfunctioning cpu heatsink fan did not cause your CPU to overheat one too many times and cook it. Hopefully a malfunctioning cpu fan didn't fry your cpu fan header on the motherboard.

 

I've bent way too many AMD cpu pins to even admit it. In every case (except when I broke one straightening it), the cpu's have worked exactly as they should have before I bent the pins. So if you didn't break any of the pins, and they are straight enough to drop freely into the 939 socket I wouldn't be worrying about that.

 

However, an overheating condition (i.e. "When I powered it on after work on Thurs, there was a continuous high pitched tone from around the CPU fan, and when I put my finger on the fan and stopped it, the tone disappeared." could have fried the CPU. So with all of the other trouble shooting that you've done, the first suspect on my list would be a bum cpu, caused by a malfunctioning cpu fan.

[radodrill]

I'd have to say that it seems like a dead CPU.