I did something similar with my Intel 2.4B processor. Before I lapped the
stock HS, I checked it with a straight-edge, and there was a considerable
concavity (is that a word) in the center of the HS. After lapping the HS,
and replacing the TIM with Arctic Silver 3, the temps dropped by about 10
degrees. This was on a unit that had been running for several weeks prior
to my hacking around on it.
Clint
"~misfit~" <> wrote in message
news:9TUpc.6137$...
> Stacey wrote:
> > Lots of talk here recently about thermal pads and what is better etc.
> > I've seen big drops in temp's ditching the thermal pads and just
> > using plain 'ol white HS compound instead. Others claim you'll only
> > see a minor change so...
> >
> > I just bought a Barton 2500+ on a chaintech 7NIF2 board for by
> > brother. It was a retail chip with a factory HSF. I just put it on as
> > supplied and when I booted it up, the idle temp was right at 50C in
> > the bios. Seemed kinda high but MBM5 said 40C so figured it was OK.
> >
> > After I got it all set up, the system was pretty noisy from the fans
> > so wanted to try to quiet it down some. Removed the fan grills, added
> > some 10ohm 1W resistors to the case fan and PSU fan etc. Rechecked
> > and the CPU temp was the same but now the CPU fan was the loudest one.
> >
> > So I took the factory HS off and removed the pad. It wasn't like
> > intel's pad, more like a gooey piece of cheeze? Anyway removed the
> > pad and sanded the bottom of the HS smooth, it was pretty rough and I
> > know HS compound likes a smooth surface. The pad was so thick this
> > didn't matter. After I reinstalled the HS with plain white compound,
> > the idle temp was 11C less, down at 39C in the bios and 29C in MBM5.
> > Now I could add a resistor to the CPU fan, drop the RPM 1000 RPM and
> > still is cooler (43C) than it was with the pad and MUCH quieter!
> >
> > So anyone installing a retail AMD chip, my advice is to ditch the
> > pad, sand the bottom of the HS on a piece of glass and throw on some
> > white HS compound and stay cool/quiet.
>
> Not a fair test. The pad needs a few days of hard work to heat up and
> squeeze out the excess material. If you'd run SETI for a week with the
pad,
> checked temps, then tried goo it would have been valuable data, as it is,
> it's junk science.
> --
> ~misfit~
>
>
|