I think i just killed my processor
 

Hopefully this won't sound like a totally stupid question, but i was experimenting w/ a fan adaptor the other day and i think i killed my proc. my computer has overheat shutdown on the board so no smoke or burnt smell, but when i went to start it up w/ the original setup it id'd the processor, did the memory check and then locked. i couldn't even get it into bios. so my question is this, what qualifies as some basic indicators that i've toasted my chip & how would you recommend troubleshooting this? thanks for helping out.:rolleyes:

What CPU is it?

AMD's normally are fubar after overheat

whereas P4's are normally saveable after that due to the smarts on the chip to shut it down at higher temps.

Your CPU works or it wouldn't post at all. Make absolutly sure your getting full contact with the HS. I did this same damn thing this last weekend with a block I made. The bolts went to far past the bottom of the base and when I installed it the bolts hit the CPU body and prevented the Die of the CPU from making contact. What happened was exactly as you state. It would post, do the memory check, overheat (101C by the thermo coupler on the bottom of the CPU) and then lockup. I took the block off, ground the bolts down flush with the base, reseated the block, and it is working just fine now. In fact it is overclocked 146FSB at 1540mhz stock Vcore. It is an original XP1600 from when they first came out.

Like I said though if the board posts the CPU isn't completely dead as it is needed in order to make the post. Double check the contact on the HS.

Ok, first off, its an amd 2000xp, and as far as the HSF making contact, i took off the setup that was causing problems and replaced that w/ the original HSF. this stopped the automatic shutdown, so the board aparently isn't detecting extreme temps, but it still locks up. and yes, the fact that the cpu clock speed and multiplier were recognized (or show up on the screen at least) made me kinda curious.

oh, on a side note, i think what caused the problems wasn't the contact being made, but the 80mm fan i put on in place of the original 60mm didn't have the pressure required to force enough air through the HS (its a rather thin-finned type, closest match i can make is to one of the dynatron HSF)

Incorrect. you can damage a CPU to the point it doesnt work right but works enough to get POST action started. Or its extremely temp sensative now ( where it works when its cold but on the ramp up of heat at boot it just gets to a point where the damaged parts of the core just stop working)

Every AMD chip I have seen killed had a distinctive smell :)

Read the whole post eh? ;)

I never said it wasn't damaged. :p

I was pretty impressed to see my XP1600+ at 101C with no HS attached and not get the order of death (been there done that) and then actually have it work like new after installing the block right. I thought for sure it was damaged. The one I fried before was a TBird 1400 and she went faster than I could blink. Bumped the power switch and it was an instant death. Always unplug the damn PS when removing water blocks/HS's!!! :D

Could it be a CPU fan RPM sensor fault? Try wiring it directly to the PSU (aka through a Molex, instead of the mobo).

well, i initially thought of the fan also, but among other things my board has auto temp shutdown & RPM detection for the cpu. but as i initially mentioned, i switched back to the stock setup after my little disaster so see if it would work that way. No go. :cry: after reading through here, i'd be inclined to think that i managed not to kill it (refer to auto-shutdown), but fubar'ed it up pretty good.
oh yeah, something i'm pissed about, the auto-shutdown gives you 5 (FIVE!!!) warning beeps before killing the machine once it detects a temp over the limit. i mean crap, i'd rather have it just kill power & let me figure it out later than waste time w/ that, especially if one has the sound off so as not to hear the beeps. Oh well, i suppose i'll just use this as an excuse to upgrade a bit. i'm thinking a new chip (i'll stick w/ the 2000+), another 256 memory stick, & a new power supply (i've only got a 300W right now & its causing all kinda problems w/ some of my games). btw, any suggestions as to an 80mm fan w/ some decent pressure? i was experimenting w/ a panoflo.

probably the easiest way to test if to borrow a cpu from a friend and see what happens.

Recently I was installing a cpu in an albatron board. We normally only use Asus, but unfortunately just about everyone was out of stock. I had the power plugged in for a reason I can't remember.

Anyway, as soon as I pushed down the socket lever, the computer turned on. I was so surprised that by the time I had reacted the magic smoke had already escaped.

You ought to try the CPU on another mobo, to confirm if it's dead. Right now, I wouldn't know which one died: the CPU or the mobo, all I know is that you don't have the parts to tell the difference.

well, i just realized i do have a chip i can test the mobo with. although i do have a related question. i was looking at processor pricing and i saw an xp2000 advertised as either a palomino or a thorobred, what exactly is the difference, and will it matter?

Thoroughbred is the latest core. You ought to opt for it, if you can.

"you ought to opt for it, if you can"! more like you ought to avoid the palomino like the plague. The palomino is a larger core (.18 insead of .13 process, I beleive) which means it makes more heat, performs slower, and overclocks less. I can't imagine a situation in which a t-bred is not far superior to a pally.

well, it was the cpu, i plugged in my bro's 1700+ & the computer worked perfecly (well, at least like its old self anyway). so repair parts are on order. also, as soon as i get the roll of film developed & scanned i'll have pictures of my little experiment gone wrong, at least the fan adapter part anyway.