Bitcoin Forum
November 15, 2024, 08:56:20 AM *
News: Check out the artwork 1Dq created to commemorate this forum's 15th anniversary
 
   Home   Help Search Login Register More  
Pages: [1] 2 »  All
  Print  
Author Topic: How Thermal Compound Spreads  (Read 2671 times)
check_status (OP)
Full Member
***
Offline Offline

Activity: 196
Merit: 100


Web Dev, Db Admin, Computer Technician


View Profile
May 11, 2012, 08:20:00 PM
Last edit: May 11, 2012, 09:04:35 PM by check_status
 #1

A great video visual aide.
http://www.youtube.com/watch?v=EyXLu1Ms-q4



For Bitcoin to be a true global currency the value of BTC needs always to rise.
If BTC became the global currency & money supply = 100 Trillion then ⊅1.00 BTC = $4,761,904.76.
P2Pool Server List | How To's and Guides Mega List |  1EndfedSryGUZK9sPrdvxHntYzv2EBexGA
Gomeler
Hero Member
*****
Offline Offline

Activity: 697
Merit: 500



View Profile
May 11, 2012, 08:28:35 PM
 #2

That is an excellent video displaying what took me something like 40-50 mounts to figure out. That should pretty much end the argument for the spread method. My only complaint is that he used a little bit less paste than I'd use. I like a little excess to squeeze out to get full coverage on the IHS.
check_status (OP)
Full Member
***
Offline Offline

Activity: 196
Merit: 100


Web Dev, Db Admin, Computer Technician


View Profile
May 11, 2012, 09:03:53 PM
 #3

Sorry, this forum does not support those tags. You have to link as usual.
Just a test, 'ehh'. With SMF they are add-on modules not standard build items.

For Bitcoin to be a true global currency the value of BTC needs always to rise.
If BTC became the global currency & money supply = 100 Trillion then ⊅1.00 BTC = $4,761,904.76.
P2Pool Server List | How To's and Guides Mega List |  1EndfedSryGUZK9sPrdvxHntYzv2EBexGA
BadBear
v2.0
Legendary
*
Offline Offline

Activity: 1652
Merit: 1128



View Profile WWW
May 11, 2012, 09:13:39 PM
 #4

Yeah I've tried the spread method before, had 80+ temps idle with a core2. Was terrible, really hard to avoid air bubbles that way. I went with the line after that and with a new cooler it brought it to 44-46 under full load.

1Kz25jm6pjNTaz8bFezEYUeBYfEtpjuKRG | PGP: B5797C4F

Tired of annoying signature ads? Ad block for signatures
Roadhog2k5
Full Member
***
Offline Offline

Activity: 131
Merit: 100



View Profile
May 12, 2012, 05:26:37 AM
 #5

Notice how none of those cover the edges of the cpu. Now pretend that CPU is a bare die. If you can't see the problem with a bare die not being covered completely you shouldn't be messing with your heatsinks. For a bare die the best method is to paste it in an X from corner to corner of the die with more than enough paste than you will need so you will know 100% of the die will be covered.

oldDirty
Full Member
***
Offline Offline

Activity: 182
Merit: 100


roundhouseminer


View Profile
May 12, 2012, 10:06:58 AM
 #6

Always spread it with fingertips on the heatspreader/GPU.

CrunchingCulture

/// FREE ASSANGE \\\
1l1l11ll1l
Legendary
*
Offline Offline

Activity: 1274
Merit: 1000


View Profile WWW
May 13, 2012, 04:27:11 AM
 #7

Always spread it with fingertips on the heatspreader/GPU.

+1

Inaba
Legendary
*
Offline Offline

Activity: 1260
Merit: 1000



View Profile WWW
May 13, 2012, 05:04:28 AM
 #8

The video does not factor in viscosity change due to heat.  Air bubbles are virtually irrelevant on hot compound. They will almost immediately dissipate once the die is at operating temperature for more than a few minutes. 

The only paste that I know of that might not suffer from this is Shin Etsu, although I suspect that even the Shin Etsu paste becomes fairly free flowing at the temperatures on top of a CPU or GPU die. 

It's like worrying about air bubbles in your cookie dough.  It's pretty rare to find a giant pocket of air in the middle of your cookie... where do you think the air in your dough goes once it's in the oven?

Don't be a fool... spread for even coverage.  If you're really worried about it, heat cycle your die several times to work out the air.

If you're searching these lines for a point, you've probably missed it.  There was never anything there in the first place.
1l1l11ll1l
Legendary
*
Offline Offline

Activity: 1274
Merit: 1000


View Profile WWW
May 14, 2012, 04:37:21 AM
 #9


Nice! Thanks for the heads up!

Tomatocage
Legendary
*
Offline Offline

Activity: 1554
Merit: 1222

brb keeping up with the Kardashians


View Profile
May 15, 2012, 04:01:06 AM
 #10

Wait... we're supposed to use thermal compound?!?

Recommended Exchanges: Binance.com | CelsiusNetwork
GPG ID: 4880D85C | 1% Escrow | 8% IPO/ICO Escrow services Temporarily Closed | Bitcointalk is the ONLY place where I use this name (No Skype/IRC/YIM/AIM/etc) | 13CsmTqGNwvFXb7tD9yFvJcEYCDTB8wQTS | Beware of these SCAM sites! | *Sponsored Link
Fuzzy
Hero Member
*****
Offline Offline

Activity: 560
Merit: 500



View Profile
May 15, 2012, 04:13:21 AM
 #11

Wait... we're supposed to use thermal compound?!?

Yeah, try to use the whole tube...   Grin
RicePicker
Full Member
***
Offline Offline

Activity: 216
Merit: 100

RicePicker


View Profile
May 15, 2012, 06:07:48 AM
 #12

I thought we were suppose to use toothpaste?

Everyone is entitled to their own opinion. It's just that yours is stupid! =D
check_status (OP)
Full Member
***
Offline Offline

Activity: 196
Merit: 100


Web Dev, Db Admin, Computer Technician


View Profile
May 15, 2012, 06:24:58 AM
 #13

I made this post because people were debating the best method for spreading thermal compound in other threads.
While it didn't quell all of the debate, it is the best proof offered so far on the effects of what each method does.
Inaba can prove his theory by making a video, I hear transparent aluminum is available. Cheesy Wink

For Bitcoin to be a true global currency the value of BTC needs always to rise.
If BTC became the global currency & money supply = 100 Trillion then ⊅1.00 BTC = $4,761,904.76.
P2Pool Server List | How To's and Guides Mega List |  1EndfedSryGUZK9sPrdvxHntYzv2EBexGA
oldDirty
Full Member
***
Offline Offline

Activity: 182
Merit: 100


roundhouseminer


View Profile
May 15, 2012, 10:34:35 AM
Last edit: May 21, 2012, 04:59:06 PM by oldDirty
 #14

I thought we were suppose to use toothpaste?
One Day i was out of ThermalPaste, on Saturday evening. xD
So i look arround what i can use alternatively? And i found MoS Universalgrease.
Before i spray some silikon Aerosol to protect the gpu die.
High-GPU temp was 80°C and that saves my weekend. This works for nearly 2 weeks, but now i use Spire silvergrease pro, best compound i ever have.

CrunchingCulture

/// FREE ASSANGE \\\
Inaba
Legendary
*
Offline Offline

Activity: 1260
Merit: 1000



View Profile WWW
May 15, 2012, 12:36:09 PM
 #15

I made this post because people were debating the best method for spreading thermal compound in other threads.
While it didn't quell all of the debate, it is the best proof offered so far on the effects of what each method does.
Inaba can prove his theory by making a video, I hear transparent aluminum is available. Cheesy Wink

It's fairly easy to prove yourself.  Just put the type of compound you want to test on your die, either a pea, whatever and put the heatsink on then take it off immediately.  Look at the pattern it makes. 

Now clean everything off, put the same amount/type on your die, put the heat sink on, heat cycle your die for a few hours (up to 48 or so, with regular on/off cycles), then take your heatsink off and look at the pattern.  It will be very different for most pastes (Shin Etsu and maybe a couple of the other super thick compounds being the exceptions), as it will have spread out (and in the process of spreading out, the air will have dissipated). 


If you're searching these lines for a point, you've probably missed it.  There was never anything there in the first place.
johnyj
Legendary
*
Offline Offline

Activity: 1988
Merit: 1012


Beyond Imagination


View Profile
May 21, 2012, 10:39:54 AM
 #16

I want to see the transparent glass over liquid ultra Grin

rjk
Sr. Member
****
Offline Offline

Activity: 448
Merit: 250


1ngldh


View Profile
May 21, 2012, 03:18:40 PM
 #17

I want to see the transparent glass over liquid ultra Grin
I wanna see transparent glass followed by a heat gun. Or else Chip + compound + glass + heatsink + full burn for a while, then remove heatsink.

Mining Rig Extraordinaire - the Trenton BPX6806 18-slot PCIe backplane [PICS] Dead project is dead, all hail the coming of the mighty ASIC!
Inaba
Legendary
*
Offline Offline

Activity: 1260
Merit: 1000



View Profile WWW
May 21, 2012, 05:14:16 PM
 #18

If i knew how to cut glass, I would try to make a small piece of glass I could use, but bugger all if I know how to do it!

If you're searching these lines for a point, you've probably missed it.  There was never anything there in the first place.
Dalkore
Legendary
*
Offline Offline

Activity: 1330
Merit: 1026


Mining since 2010 & Hosting since 2012


View Profile WWW
May 21, 2012, 05:16:28 PM
 #19


Looks like the cross method was the best overall in the demonstration video.

Hosting: Low as $60.00 per KW - Link
Transaction List: jayson3 +5 - ColdHardMetal +3 - Nolo +2 - CoinHoarder +1 - Elxiliath +1 - tymm0 +1 - Johnniewalker +1 - Oscer +1 - Davidj411 +1 - BitCoiner2012 +1 - dstruct2k +1 - Philj +1 - camolist +1 - exahash +1 - Littleshop +1 - Severian +1 - DebitMe +1 - lepenguin +1 - StringTheory +1 - amagimetals +1 - jcoin200 +1 - serp +1 - klintay +1 - -droid- +1 - FlutterPie +1
rjk
Sr. Member
****
Offline Offline

Activity: 448
Merit: 250


1ngldh


View Profile
May 21, 2012, 05:17:20 PM
 #20

If i knew how to cut glass, I would try to make a small piece of glass I could use, but bugger all if I know how to do it!
Score it with a diamond bit and tap lightly. Non-tempered glass should break along the score line. Be sure to get a good scratch in, 2 or 3 passes of the bit.

Mining Rig Extraordinaire - the Trenton BPX6806 18-slot PCIe backplane [PICS] Dead project is dead, all hail the coming of the mighty ASIC!
Pages: [1] 2 »  All
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!