Installing Socket 478 P4 with retail heatsink/fan...too loose? 1

[leahzero]

Hello all. First-time computer builder here. I apologize ahead of time if my questions seem clueless, but I'm trying to be very careful with my new hardware and need some advice on problems I've had installing the CPU.

Okay, first off, here's the list of parts I just bought to set up my first self-built system:

P4 2.8GHz (533 FSB)
ABIT IC7 motherboard
2x512 Corsair XMS 3200LL RAM
ATI A-I-W Radeon 9700 Pro
Western Digital 120G Special Edition Caviar (7200 RPM, 8MB cache)
Antec 1080 AMG Plus case with 430 watt TruePower PSU

(Using a floppy and CD burner from my old system, both work fine in said system.)

The P4 is boxed retail, I'm using the stock heatsink/fan. One pin on the chip arrived bent, I straightened that and installed the chip. When installing the Intel HSF, the retention locks seemed to hook on just fine, but the two side clips do not lock securely (they can be easily nudged out of locked position with a finger), and the entire HSF unit is loose atop the processor--it _is_ anchored, but when the case is standing upright, the HSF slides out a bit and loses full contact with the processor.

Unfortunately I was not aware that this is VERY BAD, and found that out when my system continually shut itself down because it was overheating. I laid the system on its side so gravity would force the heatsink to maintain full contact with the processor and checked out the temp in BIOS--it was up around 200 F, and presumably had already cooled a bit. I've read that P4 chips will automatically shut down if the temp reaches 135C/275F, so I'm assuming I was hitting that mark because of the system beeps and shut downs I kept experiencing.

Now, the problem. The stock Intel HSF just doesn't want to lock securely atop the chip! As mentioned above, the side clips are loose and can be easily nudged out of place, and while the retention locks hold the HSF and prevent it from falling off completely, the heatsink loses full contact with the processor if the case is not laid on its side. I have used what seems to me like very forceful pressure to try to get the HSF to lock tightly atop the processor, but it will not. I'm wary of using more pressure for fear of damaging the CPU or motherboard.

So what should I do? Does it sound like my HSF might be faulty, and should I RMA the CPU? Does it sound more like the motherboard is faulty and I should replace that? Has anyone else had problems with the Intel HSF on Socket 478 P4s? Am I being too delicate with this and should I just apply more pressure to force the HSF into a tight lock? Finally, since I have overheated the CPU and caused it to shut itself down several times, is it possible I might have already fried the chip, or are P4s resilient enough to take a few mistakes like this? (The last time I booted with case on its side, I received a "CPU unworkable" error in BIOS...will try again later with heatsink in full-contact, hoping that just meant it was too hot at the time...)

Thank you in advance for any advice you can give to this careless newbie.

~Leah

 

[paparazi]

Oh dear!
I think everybody I know, including myself! worried the first time at just how much pressure those levers require to lock, but rest assured you just need to be bold, apply that bit extra pressure, and they will lock into place.
P4's are more resilliant than most so I'll keep my fingers crossed for you.
I guess that your retail heatsink has a pre applied phase changing pad? (thermal pad)if it hasn't you must apply paste for good heat transfer.
Martin

Replying helps further our knowledge, without comment leaves us wondering.

 

[leahzero]

Whew! Martin, thanks so much. I was really worried about how much force to use.

Yep, the heatsink's got a thermal pad. I was kind of wondering if maybe part of the reason the HSF is so loose is that I didn't get any thermal paste to apply...but darn it, I'd expect they build these things to work with their stock HSFs!

Again, thanks. Will try to be a little more bold this time and hopefully not break anything...

 

[leahzero]

A-ha! Well, I figured out the problem.

The holes for the retention locks on the mobo are situated just a few millimeters too high. Turns out I didn't need to apply more pressure, the entire thing WAS as securely locked as it could possibly go...it's just that the retention locks will not fit tightly in the mobo as-is, there will always be a few millimeters of slack that will allow the heatsink to slide off the processor. Argh! The least painful solution seems to be to go buy a third-party HSF...

Thanks for your advice again anyway, Martin!

 

[Dreamland]

leahzero - Heat affects the longevity and stability of a CPU. The overtemp probably shortened the life of your CPU to some degree, but it's nearly impossible to tell 'how much'. If it knocks it down from 15 to 14 years, will it really matter? On the other hand, shortening it's life to two weeks is going to be a problem.

The effect will tend to either extreme (IMHO), and serious damage (short lifespan) will usually manifest itself fairly quickly as a stability problem once everything is up and running [lock-ups, blue screens, reboots, etc.]. If it runs fine for at least a few weeks, I would *hesitently* suggest that you probably won't see problems six months or a couple of years down the road directly related to this episode.

There's lots of other factors you're probably considering as far as RMAing or keeping it. Just tossing in this observation to give you a little peace of mind if you decide to hang on to it.

I think paparazi answered your first question. Your motherboard is probably fine. If it's impossible for you to seat the heatsink tightly to the CPU, then it's either the wrong model or a defective clip on the heatsink.

 

[leahzero]

Dreamland, thanks. I'll keep a close eye on things for the next few weeks. If it turns out I fried it, I won't try to scam an RMA out of it, it was my own stupid fault, and a hard (expensive) lesson for the future.

Well, tried one last time and didn't hold back, and Martin was right, you just have to FORCE the clips down. I tried to over-analyze it and look for problems elsewhere when it just needed a pinch of brute force. Now I feel especially stupid because I tried to out-think a physical problem. Sorry for my boneheaded questions, but Martin's answer did indeed do the trick.

Again, thanks to both of you.