Lowyat.NET Forums

Welcome Guest ( Log In | Register )

7 Pages  1 2 3 > »  
RSS feedReply to this topicStart new topicStart Poll

Outline · [ Standard ] · Linear+

> Xigmatek S1283 & thermal paste, Come and discuss!

gsan
post Mar 29 2008, 09:57 AM
Show posts by this member only |Rating BETA (0+, 0-) | Post #1


Freak
******

Group: Senior Member
Posts: 1,681
Joined: January 2003





Below is how does the thermal paste seperate when I installed the S1283 vertically. I wonder that how you guy put the thermal paste when using the S1283 cooler. Normally, I put a small drop on center of the proc as what I did for few previous cooler. Somehow, I remount the cooler recently but the temperature is higher a lot than before, 3.3GHz full load @ 60~61'c with surrounding temperature 32'c in the morning. I recheck the mounting and there is nothing wrong. Maybe someone can enlighten me?

Attached Image

This post has been edited by gsan: Mar 29 2008, 10:00 AM
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
sHawTY
post Mar 29 2008, 01:42 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #2


░ I'm Single Now :'( ░
*******

Group: Senior Member
Posts: 6,534
Joined: July 2005
From: ░ Shah Alam ░ Status: Dying Slowly






IMO, You can't use the rice grain drop method on HDT technology.
Try spreading a thin layer of TIM on the IHS.
See either it perform better or worse.

But still, i don't think the rice grain drop method is useful on HDT technology.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
AllnGap
post Mar 29 2008, 05:54 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #3


[ Modding with Passion(tm) ]
*******

Group: Senior Member
Posts: 2,823
Joined: January 2003
From: Penangites






QUOTE(gsan @ Mar 29 2008, 09:57 AM)
Below is how does the thermal paste seperate when I installed the S1283 vertically. I wonder that how you guy put the thermal paste when using the S1283 cooler. Normally, I put a small drop on center of the proc as what I did for few previous cooler. Somehow, I remount the cooler recently but the temperature is higher a lot than before, 3.3GHz full load @ 60~61'c with surrounding temperature 32'c in the morning. I recheck the mounting and there is nothing wrong. Maybe someone can enlighten me?

Attached Image
*




owh, no wonder like that.........use a paper to spread the thermal paste evenly........btw, i already banked in the balance to you
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
sukhoi37
post Mar 29 2008, 08:09 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #4


Into Computer Since 2002
*******

Group: Senior Member
Posts: 3,757
Joined: January 2003
From: City of Lion






I got the higher temperature with that method as well.
This is because of the drop wont be wide enough to let the heatpipe get in touch with processor IHS
Thermalpaste should be applied to all heatpipes in order to dissipate the heat faster.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
sHawTY
post Mar 29 2008, 08:18 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #5


░ I'm Single Now :'( ░
*******

Group: Senior Member
Posts: 6,534
Joined: July 2005
From: ░ Shah Alam ░ Status: Dying Slowly






The reason why the rice grain method won't work on HDT based heatsink is because HDT tends to have some "longkang" on each side of the heatpipes, thus, it reduces the thermal transfer power.

If you use rice grain method, the TIM layer will be so thin, it won't go into those "longkang" which will result to a weaker thermal transfer.
On HDT, you need to spread the TIM all over the processor's IHS, with a lil bit more TIM on the IHS, some of the TIM will be pushed into those "longkang's".

Despite the high performance of HDT technology based heatsink, i found two weakness in HDT:
1. You can't use HDT based heatsink on naked core CPU & GPU.
2. You have to use more TIM compared to non HDT heatsink.

PS:
Sorry, i don't remember what's "longkang" in english. Sorry! blush.gif

This post has been edited by sHawTY: Apr 10 2008, 05:33 PM
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
ah_khoo
post Mar 29 2008, 08:34 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #6


^Sumthg In Life Never Chg^
*******

Group: Senior Member
Posts: 3,127
Joined: January 2003
From: Tanah Lot @ Dungun Status: Workin' @ Singapore





QUOTE(sHawTY @ Mar 29 2008, 08:18 PM)
The reason why the rice grain method won't work on HDT based heatsink is because HDT tends to have some "longkang" on each side of the heatpipes, thus, it reduces the thermal transfer power.

If you rice grain method, the TIM layer will be so thin, it won't go into those "longkang" which will result to a weaker thermal transfer.
On HDT, you need to spread the TIM all over the processor's IHS, with a lil bit more TIM on the IHS, some of the TIM will be pushed into those "longkang's".

Despite the high performance of HDT technology based heatsink, i found two weakness in HDT:
1. You can't use HDT based heatsink on naked core CPU & GPU.
2. You have to use more TIM compared to non HDT heatsink.

PS:
Sorry, i don't remember what's "longkang" in english. Sorry! blush.gif
*


lol... drain if i'm not mistaken bro... wink.gif
User is online!Profile CardPM
Go to the top of the page
+Quote Post
gsan
post Mar 30 2008, 09:19 AM
Show posts by this member only |Rating BETA (0+, 0-) | Post #7


Freak
******

Group: Senior Member
Posts: 1,681
Joined: January 2003





QUOTE(AllnGap @ Mar 29 2008, 05:54 PM)
owh, no wonder like that.........use a paper to spread the thermal paste evenly........btw, i already banked in the balance to you
*




QUOTE(sukhoi37 @ Mar 29 2008, 08:09 PM)
I got the higher temperature with that method as well.
This is because of the drop wont be wide enough to let the heatpipe get in touch with processor IHS
Thermalpaste should be applied to all heatpipes in order to dissipate the heat faster.
*



so, the thermal paste should apply evenly on the IHS only and both the IHS and the base of the s1283?
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
sukhoi37
post Mar 30 2008, 01:05 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #8


Into Computer Since 2002
*******

Group: Senior Member
Posts: 3,757
Joined: January 2003
From: City of Lion






QUOTE(gsan @ Mar 30 2008, 09:19 AM)
so, the thermal paste should apply evenly on the IHS only and both the IHS and the base of the s1283?
*



No, but i'm not really sure the best apply method though.
I apply a thin line along the IHS (AS5 method for quadcore) then add a bit at each side of the line(this is to ensure there is thermalpaste in between the IHS and heatpipes).
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
AllnGap
post Apr 1 2008, 12:28 AM
Show posts by this member only |Rating BETA (0+, 0-) | Post #9


[ Modding with Passion(tm) ]
*******

Group: Senior Member
Posts: 2,823
Joined: January 2003
From: Penangites






QUOTE(gsan @ Mar 30 2008, 09:19 AM)
so, the thermal paste should apply evenly on the IHS only and both the IHS and the base of the s1283?
*




not sure about the ultimate method, but usually i'll spread it all even on the IHS, but no too thick....

this method so far considered successful as i've even removed a heatsink with AM2 proc stuck on the base(came off together when i remove the heatsink, heck, the clip is still locked) sweat.gif
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
LittleLinnet
post Apr 1 2008, 02:34 AM
Show posts by this member only |Rating BETA (0+, 0-) | Post #10


Iophobia
*******

Group: Senior Member
Posts: 3,379
Joined: February 2005
From: ***Penang***




QUOTE(AllnGap @ Apr 1 2008, 12:28 AM)
not sure about the ultimate method, but usually i'll spread it all even on the IHS, but no too thick....

this method so far considered successful as i've even removed a heatsink with AM2 proc stuck on the base(came off together when i remove the heatsink, heck, the clip is still locked) sweat.gif
*



this is quite often to happen when using AS5, for me though
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
fesick
post Apr 1 2008, 04:31 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #11


Freak
******

Group: Senior Member
Posts: 1,295
Joined: October 2006
From: ::Sri cempaka Kajang::




i apply tim at each heatpipe
User is online!Profile CardPM
Go to the top of the page
+Quote Post
kucingfight
post Apr 2 2008, 08:09 AM
Show posts by this member only |Rating BETA (0+, 0-) | Post #12


I Want My B00bs back!!
Group Icon

Group: Elite
Posts: 3,162
Joined: January 2003
From: Sitiawan / USJ / Putra Heights / PJ







QUOTE(sHawTY @ Mar 29 2008, 08:18 PM)
The reason why the rice grain method won't work on HDT based heatsink is because HDT tends to have some "longkang" on each side of the heatpipes, thus, it reduces the thermal transfer power.

If you rice grain method, the TIM layer will be so thin, it won't go into those "longkang" which will result to a weaker thermal transfer.
On HDT, you need to spread the TIM all over the processor's IHS, with a lil bit more TIM on the IHS, some of the TIM will be pushed into those "longkang's".

Despite the high performance of HDT technology based heatsink, i found two weakness in HDT:
1. You can't use HDT based heatsink on naked core CPU & GPU.
2. You have to use more TIM compared to non HDT heatsink.

PS:
Sorry, i don't remember what's "longkang" in english. Sorry! blush.gif
*



agreed, most of the TIMs tend to be 'pushed' to the grooves in between the HDT pipe & the alu base. Droop method isnt effective on HDT. need to spread a thin line adjacent to the pipes
User is online!Profile CardPM
Go to the top of the page
+Quote Post
justin78
post Apr 2 2008, 10:46 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #13


Regular
**

Group: Junior Member
Posts: 150
Joined: July 2005





QUOTE(sHawTY @ Mar 29 2008, 08:18 PM)
The reason why the rice grain method won't work on HDT based heatsink is because HDT tends to have some "longkang" on each side of the heatpipes, thus, it reduces the thermal transfer power.

If you rice grain method, the TIM layer will be so thin, it won't go into those "longkang" which will result to a weaker thermal transfer.
On HDT, you need to spread the TIM all over the processor's IHS, with a lil bit more TIM on the IHS, some of the TIM will be pushed into those "longkang's".

Despite the high performance of HDT technology based heatsink, i found two weakness in HDT:
1. You can't use HDT based heatsink on naked core CPU & GPU.
2. You have to use more TIM compared to non HDT heatsink.

PS:
Sorry, i don't remember what's "longkang" in english. Sorry! blush.gif
*


thumbup.gif What a good explanation I must say thumbup.gif

I totally agree with brother sHawTY in the ways to apply the TIM on HDT. tongue.gif
User is online!Profile CardPM
Go to the top of the page
+Quote Post
Nidz
post Apr 3 2008, 11:45 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #14


Regular
**

Group: Junior Member
Posts: 204
Joined: December 2006
From: Subang Jaya, Selangor





based on earlier discussions in ckhoong's thread, a rep from xigmatek had already explained that the best method to apply TIM on s1283 is using 'rice grain' method... even some users like kmarc 'invented' his own method in applying TIM, its not that effective than applying it normally...
therefore, i apply TIM as usual, using rice grain method... so far, no problems... it has been 8months... huhu
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
sukhoi37
post Apr 3 2008, 11:59 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #15


Into Computer Since 2002
*******

Group: Senior Member
Posts: 3,757
Joined: January 2003
From: City of Lion






QUOTE(Nidz @ Apr 3 2008, 11:45 PM)
based on earlier discussions in ckhoong's thread, a rep from xigmatek had already explained that the best method to apply TIM on s1283 is using 'rice grain' method... even some users like kmarc 'invented' his own method in applying TIM, its not that effective than applying it normally...
therefore, i apply TIM as usual, using rice grain method... so far, no problems... it has been 8months... huhu
*



it should be fine if your rice grain size is big enough to cover all three heatpipes.
did you ever take the cooler out and have a look on the heatsink base?
were the TIM got contact with all the heatpipes?
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
kmarc
post Apr 7 2008, 07:40 AM
Show posts by this member only |Rating BETA (0+, 0-) | Post #16


Technology is too slow!
Group Icon

Group: Elite
Posts: 3,421
Joined: May 2006
From: Kuching






I've tried the rice grain method and it doesn't work well. Like what our forumers has said, a lot of TIM will just go into the grooves. You'll find that it doesn't spread out nicely, and the side heatpipes don't get much TIM at all.

Furthermore, if you attempt to put more TIM in the middle, you'll end up with a thick paste in the middle, causing less contact....

For me, I just apply a thin line of TIM on all 3 heatpipes. Maybe a blob on the aluminium plat too. When I remove the HSF after a while, I noticed that it does cover the proc and based of the HSF evenly.....

This post has been edited by kmarc: Apr 7 2008, 07:41 AM
User is online!Profile CardPM
Go to the top of the page
+Quote Post
glock88
post Apr 10 2008, 03:40 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #17


Regular
**

Group: Junior Member
Posts: 167
Joined: March 2008





guys. just wanna ask. if not all of the push pins are not in place, can it affect the temp? b'coz only 3 out of 4 pins is properly secure... will my proc temp increase higher than a stock intel cooler?
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
sHawTY
post Apr 10 2008, 05:35 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #18


░ I'm Single Now :'( ░
*******

Group: Senior Member
Posts: 6,534
Joined: July 2005
From: ░ Shah Alam ░ Status: Dying Slowly






QUOTE(glock88 @ Apr 10 2008, 03:40 PM)
guys. just wanna ask. if not all of the push pins are not in place, can it affect the temp? b'coz only 3 out of 4 pins is properly secure... will my proc temp increase higher than a stock intel cooler?
YES.
Indeed it is hard to secure all of the 4 pin.

What should you do is take out the motherboard and fix the heatsink to the board before you put it back into the casing.

PS:
This is why most hardcore users prefer bolt through compared to push pin method.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
glock88
post Apr 10 2008, 07:30 PM
Show posts by this member only |Rating BETA (0+, 0-) | Post #19


Regular
**

Group: Junior Member
Posts: 167
Joined: March 2008





oh then i am in trouble. i just broke the pin. haha. now i have to purchase a bolt thru kit...
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
gsan
post Apr 11 2008, 01:32 AM
Show posts by this member only |Rating BETA (0+, 0-) | Post #20


Freak
******

Group: Senior Member
Posts: 1,681
Joined: January 2003





QUOTE(kmarc @ Apr 7 2008, 07:40 AM)
I've tried the rice grain method and it doesn't work well. Like what our forumers has said, a lot of TIM will just go into the grooves. You'll find that it doesn't spread out nicely, and the side heatpipes don't get much TIM at all.

Furthermore, if you attempt to put more TIM in the middle, you'll end up with a thick paste in the middle, causing less contact....

For me, I just apply a thin line of TIM on all 3 heatpipes. Maybe a blob on the aluminium plat too. When I remove the HSF after a while, I noticed that it does cover the proc and based of the HSF evenly.....
*



got pic? it worth billion words laugh.gif
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

7 Pages  1 2 3 > » 
Bump TopicReply to this topicTopic OptionsStart new topic
 

Lo-Fi Version Time is now: 5th September 2008 - 06:13 PM
All Rights Reserved 2003-2008 Vijandren Ramadass (~why so serious?~)
Hosting powered by IPServerOne - a Malaysian Web Hosting Company