P4 & Gigabyte Board

[Dravin]

Hello all.

I wanted to poke around for some opinions about this one.

I have been on the AMD side of the playing field since my P2 died. But just recently I built a new system for myself, and this time I went with a P4 2.8c system from scratch. Been building my own systems for a while now. But my first P4. So I was not familar with the plastic bracket surrounding the processor. I had seen them, but never worked with one. The MB is a Gigabyte GA-8IPE1000 Pro, and stock HS/Fan for the P4 2.8c.

Not being very familar with the P4 bracket system, I was extra carefull on it, and everything is seated correctly, and the HS/Fans levers are locked in possition & seated. Problem is (and I dont know if it really is a problem), it took more force to flit those levers into possition than I was expecting. And once everything was locked in place, I noticed that the back of the MB was just slightly buldged in that spot. Not enough to cause any damage or any serious concearn. But enough that I noticed.

This type of thing sound familar to anyone with the P4 brackets?

thanx in advanced.

[DocPC]

Hello all.

I wanted to poke around for some opinions about this one.

I have been on the AMD side of the playing field since my P2 died. But just recently I built a new system for myself, and this time I went with a P4 2.8c system from scratch. Been building my own systems for a while now. But my first P4. So I was not familar with the plastic bracket surrounding the processor. I had seen them, but never worked with one. The MB is a Gigabyte GA-8IPE1000 Pro, and stock HS/Fan for the P4 2.8c.

Not being very familar with the P4 bracket system, I was extra carefull on it, and everything is seated correctly, and the HS/Fans levers are locked in possition & seated. Problem is (and I dont know if it really is a problem), it took more force to flit those levers into possition than I was expecting. And once everything was locked in place, I noticed that the back of the MB was just slightly buldged in that spot. Not enough to cause any damage or any serious concearn. But enough that I noticed.

This type of thing sound familar to anyone with the P4 brackets?

thanx in advanced.

That's the way it is with the P4's. You just need to make sure there is enough clearance between the back of the board and the case.

I have yet to run into a case that did not have enough clearance.

[Dravin]

Thank you. Now I know that is the norm on these, I wont worry about attaching a new HS/Fan.

[NooNoo]

Scary though the first time, you think you are gonna break the cpu and the mobo in one fell swoop.

[JeffO93]

I'm not looking forward to that.
I am in the market to buy a Gigabyte+P4 mobo. Like you, I want the 2.8GHz. If I get the right chipset, I'll have hyperthreading and will be able to boot off a USB stick.
Using that kind of force on a board that's gonna heat up and cool down as much as it will is not a good thing, but what can you do? CPU's these days will cook in less than one second if you don't have extremely good contact and rapid heat migration.

[Platypus]

There have been motherboards produced with an extra support framework under the CPU, Soltek I think, and I gather modified mount kits are also produced to create less distortion of the board.

When I started using P4s, my supplier noted the board distortion did result in a higher occurrence of failures from broken tracks after 4 to 6 months of use. His solution, which I have also used, was to duplicate the lower tension cams in the mod kits by filing down the face of the cam by about 1mm. You smoothly follow the curve of the cam, but file away most of the thickness of the edge that runs around the cam face.

This makes a noticeable reduction in the dishing of the board, but doesn't seem to have any adverse effect on the seating of the heatsink or on heat transfer.

I'm currently using a GA8IPE1000 + 2.8G 800FSB, it's been fine, has SATA but not RAID.

[NooNoo]

There have been motherboards produced with an extra support framework under the CPU, Soltek I think, and I gather modified mount kits are also produced to create less distortion of the board.

When I started using P4s, my supplier noted the board distortion did result in a higher occurrence of failures from broken tracks after 4 to 6 months of use. His solution, which I have also used, was to duplicate the lower tension cams in the mod kits by filing down the face of the cam by about 1mm. You smoothly follow the curve of the cam, but file away most of the thickness of the edge that runs around the cam face.

This makes a noticeable reduction in the dishing of the board, but doesn't seem to have any adverse effect on the seating of the heatsink or on heat transfer.

I'm currently using a GA8IPE1000 + 2.8G 800FSB, it's been fine, has SATA but not RAID.

Platypus, have you got pics showing exactly what you mean?

[JeffO93]

Used to be tool&die machinist and CAD engineer before switching to computers. I will definitely look into reinforcment, filing and other methods. Good suggestion, Platypus.

[Platypus]

Platypus, have you got pics showing exactly what you mean?

Not specifically. The effect we're talking about is this:




This one of the locking procedure shows the cam a bit clearer:




You can see the teardrop shape of the cam on the end of the light biege coloured lever. A ridge is visible, slightly more than a millimetre thick, running around the edge, and this I file down until it's almost gone, making the bottom of the cam not stick down so far where it presses on the black frame on top of the fins. The cam lever will pop in and out of its socket if you spread the plastic top rails it sits between, but it comes out easier downward, so it's simpler to unclip the upper section that carries the fan, then refit it afterward.

[Platypus]

Well, I organised some pics. First the filed cam:




The difference in the dishing of the board:

[DocPC]

Good idea, but any modifications or physical damage will void the 3 year warranty on the boxed CPU's.....

[JeffO93]

Excellent pics!
My copper Thermaltake heatsink uses totally different clamps. I've got steel spring-clips on each side and there's no bulge to the board. The copper heatsink with fan weighs nearly a pound, which had me concerned, but it's been great.
Not sure what I'll do on my next Gigabyte.

[Platypus]

any modifications or physical damage will void the 3 year warranty on the boxed CPU's.....

You would have to allow for that possibility if contemplating modifications. Fortunately CPU failures are rare, so I've felt it's a better risk than maybe having a motherboard give up out of warranty.

[DocPC]

You would have to allow for that possibility if contemplating modifications. Fortunately CPU failures are rare, so I've felt it's a better risk than maybe having a motherboard give up out of warranty.

True, after looking over our books I see since the inception of the Socket 478 P4 we have RMA'd 2 out of nearly 10,000 units.

Both were for bent pins caused by the plastic shroud hitting them during the packaging process. Never a hard failure.