Motherboard Problems -- Computer Won't Start

[btsdev]

Just put a new PC together with an EPoX nForce 2 / AMD 2500+ 2.0GHz.

Haven't got the PC started yet. Hit power on -- Heard CPU fan going for about 1.5 - 2 seconds, then everything stops. Then we only hear a very faint alternating high/low pitch sound that seems to come from the battery area. This continues until I switch from 1 to 0 on back of PC.

I think this is a CPU problem. Haven't tried reseating it yet. Not sure if I should -- Any suggestions?

Thanks.

 

[mechBgon]

It sounds like your heatsink's on backwards. The end of the base with the step cut into it (photo) must be oriented to match the raised, solid-plastic end of the CPU socket, which is the top end on an 8RDA+. If the heatsink isn't oriented this way, your heatsink cannot land squarely on the CPU core, resulting in instant overheat and the resulting emergency shutdown you're seeing.

Any other situation that causes a lack of thermal transfer from core to heatsink will also bring this about (forgetting to take the protective slip off a stock thermal patch, reversing the clip in the heatsink, forgetting thermal compound altogether, using a shim that keeps the CPU and heatsink from touching eachother, etc).

 

[btsdev]

Well we're looking at the heatsink now, and it looks like we had it on correctly, but we're trying to regrease and reseat it at this moment. I'm not sure exactly what the stock thermal patch is, could someone please explain?

Thanks.

 

[mechBgon]

In the photo I linked to, the stock thermal patch is that square patch in the middle of the heatsink's base. It's made of a material that "melts to fit" the gap between the heatsink and the CPU core when the CPU is first fired up. These patches are often protected with a plastic slip that should be peeled off before the heatsink's put onto the CPU socket.

Also, phase-change pads are only meant for one use.

 

[btsdev]

There's no stock thermal patch on my heatsink then :|

After re-seating the heatink/fan on the processor again, same thing happens. 2 seconds of fan and mobo activity then bam, everything stops and i hear a high/low alternating warning sound until i switch off the power completely.

1) Could the cpu really be overheating in two seconds?
2) Could it be a defective processor?
3) Should i send it back for a replacement?
4) Could it be the motherboard?

Thanks again for your help. I'd really appreciate if anyone could give any input they might have; i'm hoping to be able to fix this soon, as I have a deadline.

 

[mechBgon]

That's classic CPU overheat all the way. Yes, it can overheat in under two seconds. What heatsink/fan unit do you have (either brand/model or maybe a photo)?

Thermal grease is a suitable substitute for a stock thermal patch. It's what most enthusiasts use, although AMD doesn't recommend it for long-term use (think: several years).

 

[btsdev]

Unfornately i'm using a crap one i got from compusa for socket a/462's. The clip wasn't reaching the clip-nubs so i had to bend it a little.

I'm using a compusa one because i'm an idiot and i forgot to get a heatsink/fan when i originally ordered the parts for my computer. When we were trying to build the pc and noticed we didn't have a heatsink/fan we had only one choice: to go to comp-usa. It advertises "up to 2.4gHz".

Don't laugh .... "Copper Pro Gamer Fan".

Now say if the product was crappy quality... still wouldn't the cpu go for more than 2 seconds? Wouldn't it atleast do some cooling and let it last a tad longer?

 

[mechBgon]

Bending the clip is probably part or all of the problem. They're tight for a reason: they have to generate about 35 pounds of downward force just to get the heatsink to actually touch the CPU core at all. If the clip doesn't generate enough force to compress the cornerpads on the CPU, then the contact between the CPU core and the heatsink is zero. Bad! :Q


In addition, if the clip can slide back & forth in the heatsink a bit, then follow the instructions I've put up here and slide the heatsink as far away from the solid-plastic end of the CPU socket as you can (on an 8RDA+, as I mentioned, that's the top edge, same as the board shown in my photo).

What you want is for the thick part of the base to avoid riding up on the solid-plastic end of the CPU socket; it should look like this, where you can see that the thick section of the base is away from the solid-plastic end of the socket and not touching it whatsoever.

 

[btsdev]

Thank you very much for being so helpful with diagrams and speedy responses, etc. My friend and I really appreciate it. Next time I'll be able to get a new heatsink is in two days, so I'll respond then.

Thanks again. You've really helped us a lot.

 

[mechBgon]

Sure thing, you can see from my post count that I don't tire of helping too easily, LOL I hope it works out for you guys, good luck.

 

[btsdev]

Thanks a lot for all the tips and advice. It was a heatsink problem after all. Now we can get into BIOS, but there are some errors:

1. Floppy disk(s) fail (40) ... but the IDE cable is in firmly
2. Primary IDE channel no 80 conductor cable installed
3. Processor is reading as 1667MHz and XP 2000+ but we're running a 2500+ at 2.0GHz
-- a. The FSB is at 133MHz and the multiplier is at 12.5
-- b. Should we increment the FSB or the multiplier to raise it to the proper speed?
-- c. When I put the FSB up to 166MHz, the CPU temperature raised to 41ºC -- Is this too high?
-- d. When I put it at 166MHz, BIOS recognized the processor as a 2600+

Thanks again.

 

[mechBgon]

Originally posted by: btsdev
Thanks a lot for all the tips and advice. It was a heatsink problem after all. Now we can get into BIOS, but there are some errors:

1. Floppy disk(s) fail (40) ... but the IDE cable is in firmly
2. Primary IDE channel no 80 conductor cable installed
3. Processor is reading as 1667MHz and XP 2000+ but we're running a 2500+ at 2.0GHz
-- a. The FSB is at 133MHz and the multiplier is at 12.5
-- b. Should we increment the FSB or the multiplier to raise it to the proper speed?
-- c. When I put the FSB up to 166MHz, the CPU temperature raised to 41ºC -- Is this too high?
-- d. When I put it at 166MHz, BIOS recognized the processor as a 2600+

Thanks again.

1. This usually means that either the floppy cable needs to be flipped over at one end (doesn't matter which), or that the floppy drive has failed
2. You need an 80-wire IDE cable, these are the ones with the smaller, finer wires. Put the hard drive at the end connector so there's no vacant cable beyond it.
3. The 2500+ should be run at 11 x 166 if you want its stock speed. 41C is a very normal temperature reading on EPoX nForce2 boards.

 

[btsdev]

Thanks alot for all of your help, don't know where we'd be without you.

We just got everything working and the PC is flying.

Thanks again.

 

[mechBgon]

Originally posted by: btsdev
Thanks alot for all of your help, don't know where we'd be without you.

We just got everything working and the PC is flying.

Thanks again.

Sweet