Solution for cracking paste on delidded cpus.

[Achill3uS]

guys, in the last 2 days i had 2 sessions without a single crack!

 

TIM was Grizzly under IHS and between IHS and CPU pot of course. i did not change anything like the numerous unsuccesful sessions before but one thing: i tried the "Vivi method", sanding the die! i did not glue back the ihs just gently sanded the die until the reflection lose about its shine and starts to be scratched.

 

i used the green side of the regular washing sponge

 

 

Vivi's photo

 

 

so totally confirmed from my side. if you are facing serious TIM crack issues like i did give it a try. tested with 2 different CPU's in 2 different mobos under 2 different CPU Pots. i did not even care about slow pouring! straight go to -120 start light benches and go forward to -150-160C range. no problem at all with both setups. 1st CPU had cbb at -150 other at -160C. benches were running about -165-170C and finally CPUs starts to fly.

 

[websmile]

So you are now the new sponge-Bob? - Very interesting lecture, thx for sharing your experience

[Rauf]

guys, in the last 2 days i had 2 sessions without a single crack!

 

TIM was Grizzly under IHS and between IHS and CPU pot of course. i did not change anything like the numerous unsuccesful sessions before but one thing: i tried the "Vivi method", sanding the die! i did not glue back the ihs just gently sanded the die until the reflection lose about its shine and starts to be scratched.

 

i used the green side of the regular washing sponge

 

 

Vivi's photo

 

 

so totally confirmed from my side. if you are facing serious TIM crack issues like i did give it a try. tested with 2 different CPU's in 2 different mobos under 2 different CPU Pots. i did not even care about slow pouring! straight go to -120 start light benches and go forward to -150-160C range. no problem at all with both setups. 1st CPU had cbb at -150 other at -160C. benches were running about -165-170C and finally CPUs starts to fly.

 

Thanks for sharing! Did you try full pot? I haven't had much problems at -165-170.

 

I thought I would try this myself. I tried reseal with silicon based glue and it didn't work. Cracked around -175-180 the same as without reseal. Tried using tape to simulate reaseal also but contact was too bad. No crack with tape but was seeing positive core temps with pot around -100 so it was useless.

 

I have also had some problems with gradually degradation of kryonaut when benching at -170. Works well for 30 min, then have to lower clocks about 100mhz. Then after perhaps 1,5-2 hours it cracks completely at -170. This could be a bad batch of kryonaut though, because my first tube didn't do this.

[Achill3uS]

websmile: nono pls i hate spongebob cartoon

 

Rauf: i was running the setup -150-160 like 5-6 hours no problem, yes i did try fullpot. but i dont find it any good.

 

1st CPU/system temps about -165C at full pot but got issues when running 3DMark05 and got blackscreen/cb while it swiches from GT1 to GT2 or GT2 to GT3. when i was at only 5 to 10 degree higher it went smooth. so i did torch at the end of the GT1 and cool down again if ive seen the GT2 load screen.

 

2nd CPU/system fullpot was at -175C (different roomtemp and conditions) but when i reached it idle got cb and errorcode 00 until shutdown and heat up -160 range to avoid cbb. i was not able to run fullpot only under continuous cpu load.

 

so to be honest i found it better to control temps and keep away from fullpot. less issues with cb. but before this whole situation i was not able to bench sometimes even -130C bcoz of cracking. so this solution seems valid in my case.

Edited December 21, 2015 by Achill3uS

[Vivi]

Thanks for confirming achill3us! I ran -180 (2 degree from fullpot) and can say that it runs fine there too

[Leeghoofd]

[Schmuckley]

Yeh,that would not go well for me.

Some dies really could use it, too.