MY CPU fried, but no idea why...

[dreIU]

Here is my story:

I took the board out of me case to install NB-1 cooler on the NB, and move the NB synk to the SB. I did this, put it back together... and when I connected the power connector to the mobo, the CPU fan started spinning for about a fraction of a second, THEN it stopped, making some odd clicking noise.

I then power my computer back up, and it does not do anything, just sits with a black screen. So I just figure it is a cold boot issue, so I just reboot, and nothing still. I shut down, make sure everything is connected right, and all is.

I power back on, then start to smell somehting nasty, and the computer shutts off after about 3 seconds. I take the HSF off, and take the processor out to find burn marks on it.

Could it be because of static electricity? Surely not, because I was extra careful with it, and I have taken my PC apart many times before. It is not because of the OC either, nor the voltage. It was at 1.8volts, and had more than adequate cooling on it.

Also, while the computer was on, the fans never stopped spinning, and the HS was seated correctly at all times (it was screwed onto the mobo).

So why one earth would this happen? I am baffled... I am also hoping that the rest of the system is not fried. I ordered an 2600+mobile from excaliber, should be in wed. should there be any possibilities of that one getting fried too? This is the first time it has happened to me... anybody got any clues as to what cause this mishap?

 

[itim100]

All I can say is don't put that mobile in your board yet. I had two A7N8Xs die on me, and when the first one died, It killed my thunderbird. Then I put in another thunderbird and it fried that too. I would advise putting in like a Duron 500 or something cheapo or getting the board RAMd.

It might not kill it, but my experiences show that there is a chance of it killing things. All it would take is excessive volts getting to the vcore (+5v to vcore kills, for example) from a mobo fault. I guess wait and see what other owners of your mobo say about their experiences before you put in another good cpu.

 

[germanjulian]

mhhh....
sounds really really dodgy. I would RMA the board! if the heatsink is scewed in and the CPU fried there most be something wrong with the mobo for sure.

 

[dreIU]

Well here is another thing that might have done it... I had the power cord connected to the PSU (which was conected to the wall at the other end), AND the power button on the PSU was in the ON position. maybe that is what did it.

That is an excellent idea about putting a old CPU in, I just hope the local computer store has an older one.

I am skeptical about RMAing it... it was a monster OCer with the original bios. I didnt have to go through the bios death crap with it. But, if it is the mobo's fault, gone it is. anybody else got any input? thanks guys.

 

[hitechjb1]

Sorry to hear that.

I think both the motherboard and CPU may have been damaged.

1. The smell means there were some short current.

2. Fan was always ON, and HS in good contact, but CPU got burn mark for few second during boot. This points to high possibility that some how the motherboard got damaged, probably the Vcore regulator portion, and in turn passed high voltage >> max_Vcore_limit and killed the CPU.

3. Even the power switch is OFF but the PSU power cord is connected and PSU main switch (usually at the back) are still ON, the PSU still supply voltages to the motherboard (as told by the LED's lights on motherboard). So you moved or pressed the power connector to the motherboard, I suspect so how something got short circuited and motherboard got damaged, since the motherboard is flexible and some traces or pins or ... got shorted to some other things, ...

When working on any internal of a system, it is a good idea to completely power down the whole system, such as disconnecting the power cord, or at least turn OFF the back PSU switch.

 

[Yodums]

I think that the motherboard or PSU may have fried the CPU.

I'm assuming when you took off the heatsink, you saw a big fat CPU imprint of it on the bottom of the heatsink right? Like it pushed all the excessive thermal paste to make an outline? Just gotta make sure you had 100% contact.

I would definitely test the PSU. I'm not exactly sure on how to do this but you might wanna give a shout to the PSU forum and they should help you. Faulty PSUs sometimes do a lot of damage.

The motherboard could definitely be an issue and if it's possible, try to put in an old chip or if you have another system using AMD, it'd be nice.

About the PSU and the switch and cord.. I don't think it would really matter. When I work with stuff, I keep the cord in, but the switch turned off. It keeps everything grounded, or at least from what I've read anyways.

Good luck.

 

[hitechjb1]

I
...
About the PSU and the switch and cord.. I don't think it would really matter. When I work with stuff, I keep the cord in, but the switch turned off. It keeps everything grounded, or at least from what I've read anyways.

Good luck.

If I understand you correctly, the power cord and back PSU switch matter when working on internal parts.

Have at least the back PSU switch OFF. Even better have the power cord disconnected from the PSU while working on open case and inside components, putting in HS, fans, drives, wirings, ....

When power cord is connected and back PSU switch is ON, the motherboard is still live. It may damage components if some wires, traces, ... got short-circuited, motherboard is elastic and may bend to certain extent physically leading to unexpected contacts and short circuits, ....

IMO, actually electro-static is relatively less a concern, as chips have electro-static protection diodes at the I/O these days. Of course, safe precaustion is always a good idea.

 

[OC Detective]

It is possible that when you removed the mobo from the case the HSF may have become dislodged from its secure seating - thus causing the cpu to overheat and scorch.

 

[dreIU]

It is possible that when you removed the mobo from the case the HSF may have become dislodged from its secure seating - thus causing the cpu to overheat and scorch.

No, it was SCREWED onto the motherboard, there is NO way it came loos or came d isplaced.

I did have 100% contact between the HS and the CPU.

Bad PSU? Antec true480, 2 months old. I doubt it would go bad just like that.

As far as the flexibility think goes... It might have flexed a bit, so I am leaning towards that being the problem. so does this mean that it will damage my next chip as well? I have no other AMD chip to test, and the cheapes I can find on the new is about 50bucks for a 1.3 gig Socket A chip. I cant throw that out the window just like that...

Should I play it safe and RMA the board? I do not know what to do... I was told to take it out and inspect it along with the back side of it.. but if I miss something, there goes another CPU.

I am getting ready to see if my video card and memory got fried as well.

 

[itshondo]

you might want to separate each component and do a sniff test to find out what smoked- I found a bad psu this way- but I had to seperate the components first in order to identify it. Sorry for you loss.......

 

[dreIU]

you might want to separate each component and do a sniff test to find out what smoked- I found a bad psu this way- but I had to seperate the components first in order to identify it. Sorry for you loss.......

Good, idea, that's how I figured out what general area the smell was coming from.


I just confirmed that my Radeon still works, but my 350 dollar memory is completely shot. Gone. I put it in my parent's Dell, and it just started beeping left and right.

So, I am going to RMA the mobo ( ) and the memory.

Thanks for the help guys, I guess I will be more careful next time, although I did nothing this time than any other time (well the power cord is the only difference). 250fsb was too god to be true...

 

[hitechjb1]

...
1. The smell means there were some short current.

2. Fan was always ON, and HS in good contact, but CPU got burn mark for few second during boot. This points to high possibility that some how the motherboard got damaged, probably the Vcore regulator portion, and in turn passed high voltage >> max_Vcore_limit and killed the CPU.
...

As reasoning above, very high chance the motherboard was damaged. I would not putting more parts onto it risking more parts get damaged.

If you feel handy, there are ways just to test the PSU alone without connecting to any component.

I would just replace the motherboard and start from there.

 

[dreIU]

As reasoning above, very high chance the motherboard was damaged. I would not putting more parts onto it risking more parts get damaged.

If you feel handy, there are ways just to test the PSU alone without connecting to any component.

I would just replace the motherboard and start from there.

I am getting ready to go to Fedex and send the mobo and memory back to newegg.

tell me more about testing the PSU please, would be much appreciated.

EDIT: if my memory has gone bad too, could there (knock on wood) be a chance that anything else is damaged? aside from the PSU, which I will test as soon as I am told how to. Like the harddrive... and dvd & cdr drives? thanks.

 

[hitechjb1]

Testing PSU alone without connecting to any part

The PSU is NOT connecting to anything, except its power cord.

1. Turn OFF the back PSU switch, and unplug the power cord from the AC supply.

2. Since PSU is not connected to motherboard and the normal power switch, the POWER ON (pin 14 of the ATX connector, normally GREEN wire) has to be connected to GROUND (e.g. pin 13 of the ATX connector, there are more than one GROUND pins, normally BLACK wire) in order to turn ON the PSU.

ATX connector is that 20 pin (rectangular shape) connector that connects to the motherboard normally.

The pin numbers, from 1 to 20, are usually labelled (very small) on the ATX connector. I checked my Antec True Power ATX connector and some other PSU's, the labels are there.

Double check the connections. Be very careful to have the pin numbers correct, use the wire color to confirm. Wrong connection may damage PSU.

3. Then connect the PSU power cord to the AC supply.

4. Turn ON the back PSU switch. The PSU fan should spin.

5. Use a voltmeter to measure the following voltages:
- 3.3 V is orange wire
- 5 V is red wire
- 12 V is yellow wire
- -12 V is blue wire (pin 11 of ATX connector)

The voltages should be roughly (say within 5% for each voltage) the same as the expected voltage value. If not, it may indicate PSU problem.

If it reads correctly, the PSU is probably OK, but not sufficient since it is not a full load test.

6. Turn OFF the PSU.

You may connect a 12 V fan to the PSU for testing if the above test passed.

 

[dreIU]

That is an interesting test... BUT I do not have a voltmeter, so maybe I can get away without doing this test. I have a backup PSU I can use in case I have to RMA this one. I doubt it would go bad just like that, nor do I think that this little incident (a short) has affected the PSU, but correct me if I am wrong.

I just sent in my RMA fot newegg, so I probabily will be out of a PC for about 2 weeks, my new porcessor shall arrive sometime this week... in the meantime, I may see if my neighbor has a voltmeter in order to test the PSU. thsnks for all the help guys.