so i finally removed the ihs on my a64 3400+.....

[chinkgai]

so heres the deal...i just removed my ihs on my a64 that i've had for about 1 week now with my dfi lanparty and xp120

i've been going slow with upping the mhz then priming to find out how high it goes exactly at each voltage...i have not reached a ceiling yet and decided to see if i can go higher with lower voltage since i saw the thread here on removing the a64 ihs...

the results are here:

before = xp120 with ihs @ 1.475 X 104% (about 1.534v) = 2620mhz (262x10) stable - temp 39-40 load

after = xp120 withOUT ihs @ same volts = not even stable at 2600 (260x10) @ same volts!!



i think it might have to do with my poor job at sanding down the base of the bracket of the xp120...since it was late at night and the only tool i could find at my disposal was a nail file and the razor i used to pry the ihs off with...lol...so needless to say its probably uneven...

but wait...heres the good part...my temps are BETTER THAN BEFORE...albeit not by much...but this further confuses me of the fact that the temps are better but not stable during load.

i was able to prime at (260x10) 2600 at the above mentioned volts for about 20 min before it froze up my system...at (262x10), it wont even boot at the same volts

are these signs of the hsf not seating properly? if so, then why are temps better? could it be that taking off the ihs has improved temps so much that even without it seated properly that it is gettin better temps than a properly seated ihs? this is the only explanation i can think of...

i'm thinking of taking off the bracket and re-sanding it tonight since i was able to pick up some sandpaper just now at work.

does anyone know how much i am supposed to sand exactly? do i sand off about as the much as the ihs is thick in the middle?

by the way...on load it is now at about 37-38 and idles like 27-29 as opposed to idling at 30-32 before

 

[chinkgai]

so if i damaged the chip somehow...would it cause something like this?

 

[cirthix]

increased interference, the heatspreader not only protects the core and aids cooling, it reduces emi that disrupts calculations and can cause instability

 

[Budman]

should have left the cap on. it's on for your protection.

 

[chinkgai]

i had a reason to remove it...and it wasnt simply to reduce temps...but i dont wana get into that

just need to do more testing on the height of the bracket i hope...

 

[Geomagick]

I'd love to know the reason - I've seen cascade setups with the ihs still in place so am intrigued about the need to remove it.

You could put it back on.

 

[Vee]

The heatspreaders purpose is to eliminate hotspots on the chip. One thing that used to plague clumsy, inept HS mounting on socket A.

Your temp reading has nothing to do with how hot your chip is on the hottest spot. Also, you may now be cooling the temp sensors more efficiently.

To replace the heatspreader with the HS, you ideally nead a very FLAT and very SMOOTH surface and very thin application of paste.