Jump to content
HWBOT Community Forums

Recommended Posts

Posted

Ok, here it is. I was looking for the best sub forum for this.

 

I was wondering what the best way is to apply thermal grease to make sure a crack doesn't happen. I had it twice this weekend with 2 different methods. At -100C or beyond I saw positive temps.

Before I say how I do it I am wondering who has a method that works most of the time (besides the new Kingpin thermal paste).

Posted

Use gelid, and big globs amount of TIM, make sure all die is in nice contacts (big globs of TIM only make sure it doesnt have room to crack). Heat up the pot to +70 but make sure you not turning it on at that temp. And then you directly go to -135 or -140 before start any benchmark.

Posted

I tried a big blob before but I never heated it up before the run.

Yesterday I tried a small layer of grease but that obviosly didn't work.

I cooled it down to -80 and then heated it up to -40, after that I went down.

Posted

Aceton works fine. Clean die, inside ihs, outside ihs and pot carefully with aceton. Then apply a big blob of kryonaut on die. Mount ihs, put some kryonaut on ihs and mount pot (I mount very tight but not sure it is actually needed). I have never had a single crack this way. No need to heat up or cool down in a special way. Just pour.

 

You can reuse the kryonaut on the die if you don't remove the ihs. I tried it and could do 5-6 sessions, but eventually it will crack. So I redo it every time just to be sure...

Posted

I keep acetone in a syringe, apply it to cotton pad/qtip and wipe once with one side and once again with other side (of q-tip etc.) that way only clean cotton touches die / ihs underside. Similar to way I clean up epoxy grout, start off with 2-3 "slurry washs" and for final wash one pass diagonal across grout, clean sponge, repeat. So when ihs real dirty high % isopropal alcohol, Artic Step 1 or even goo gone/goof for wash 1 "slurry" and acetone to finish. As for tim, I've been using new 37g T.Grizz Kryo fresh out the tube and cover entire die (Don't applicaor thing that comes wih paste) Just squeeze paste out around edges of die work your way toward center of die to fill in voids . Like a downward spiral. No need to spread, shape etc as long as there no voids air pockets and plop ihs back on top while chip is in socket. (Do not apply, wipe, spread any extra paste to underside of ihs) one blob on die to ihs best. Sure T.grizz is runny so be careful, maybe pre test once before inserting into socket because paste can get into socket even with precaution. tape, blue towel etc. paste is very difficult to fully remove but acetone will dilute/ dissolve more than enough so pins can move freely, but socket will look stained. Even with it "wet" consitancy by comparison I've been having great success with no cracking for many hours with Kaby, Something I can't say about Skylake launch / following months. I was a crack head back then.

 

TLDR, I've way overcomplicated this as is't so simple in real life takes like 45 seconds.

I'll get a pic of amount used next time

Posted (edited)

If you dont want to heat up the pot to +70 i have one more tips for, i run hwbot prime for many times until your pot touch +40 Celcius. Then just go to -135 to -140 before starting any benchmark.

Edited by speed.fastest
Posted (edited)

Thanks for all the tips. I will try them soon.

 

Edit: I have a bottle of Aceton "replacement" which is 2-Butoxyethanol. Is that enough?

Edited by Sparks.nl
Posted

lapping my pot and IHS helped cracking which seems strange if the crack is between die and IHS....either way its worth doing.

 

Also remove the glue on kaby, it hurts mhz if you dont. the glue will never be correct once the ihs is removed anyways

  • Crew
Posted

Try one thing at a time Maarten, if you lap wrong you could ruin your pot

 

First Roman's suggestion, properly clean die and IHS. Apply a huge blob of grease like in Dancops guide

 

Cool down in bios to -100, then boot In OS with modest voltage settings, cool down to -150, set full pot settings and go... If it cracks , heating up to -60 could restore the contact. I Check with Realtemp or such at 5Ghz at 1.6Vcore with eg CB R15...

 

Don't waste too much time at -100, testing stuff and co !

Posted (edited)

Thanks for the advice. I wasn't planning to combine things. That way I can't name the cause of issues.

About lapping, I lapped my first heathsink/cpu more then 10 years ago. I never made any mistakes with it.

Edited by Sparks.nl
Posted

Lapping the bottom of my pot was very important since the nickle plating builds up on the edges I found myself overightening the pot trying to fix delta t's and cracks. This only made it worse because the "thickest" edga of pot would hit cpu holdown essentially crushing pins ever so slightly. This was back in Haswell days but still applies as socket is so similar.

 

I also drop temps to at least -130 in bios before booting to OS, open OC tool (Formula Drive) and CPU-z tabs and apply load from stress cpu in benchmark tab at 1.4-1.5v core and pour until my pot stops spitting ln2 in my face and sounding like a constipated devil screaming aka fullpot, nice and quite by comparison

Posted
Do tou mean the black peaces on the pcb?

You bench without any of it?

 

I will wait with lapping it for now. I do have everything I need to lap it.

 

Yes I scratch it off for real.

Posted

I used Aceton before mounting it and used a big blob. I tried to heath it up a little before I ran it and after that I did what Leeghoofd said. I could run 6GHz CB11.5 at -170C without any issue now.

Seems to be better now.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...