Sign in with
Sign up | Sign in
Your question
Solved

I put new thermal paste... and no results

Last response: in Graphics & Displays
Share
August 17, 2012 2:51:04 AM

So i added new mx-4 thermal paste on my gigabyte 550 ti card and the temperature is higher than it used to be.

I didn't get the arctic silver 5 cause I was like HERP DERP when I chose the thermal paste.

So the average temperature before the mx-4 application was about 35c degrees on idle. Now, it's at a 40C degrees average after the mx-4. Anyone know why it's such a big difference compared to the stock paste... T__T

Ontop of that.. My CPU temperatures with the mx-4 is pretty stable. So... I don't know. :S



I'm thinking of....
Should i just get the arctic silver 5 and test that out....?
a b Î Nvidia
August 17, 2012 2:54:07 AM

Idle temps do not matter, what are the differences in load temps?
m
0
l
August 17, 2012 2:55:45 AM

5C degrees higher than it's supposed to be.
m
0
l
Related resources
a b Î Nvidia
August 17, 2012 2:59:45 AM

if its on the gpu, recheck the contact between the heatsink and the actual core.
m
0
l
August 17, 2012 3:04:10 AM

are you sure you weren't HERP DERP when you put the heat sink back on the GPU...?

...just sayin
:/ 
m
0
l
August 17, 2012 3:05:17 AM

o_o i hope not cause i followed the white line diagram and put the heat sink where aligned against the line on the gpu board.
m
0
l
August 17, 2012 3:12:27 AM

xDivineCC said:
o_o i hope not cause i followed the white line diagram and put the heat sink where aligned against the line on the gpu board.

ok

did you tighten the screws clockwise, or counter-clockwise, slowly each screw a couple turns only - or did you tighten one screw and then move to the next, etc...?

the 2 surfaces (GPU and heatsink) need to be parallel with each other, with only a small thin film of paste between
m
0
l
a b Î Nvidia
August 17, 2012 3:13:09 AM

5C difference between Idle before and after?

Load temps the same?
m
0
l
August 17, 2012 3:14:56 AM

@Dingo07: ... I think i know what i did wrong. I screwed one tightly and then did the rest. i screwed it on clockwise of course. I kinda get what you mean now by screwing them.

@amuffin: :o  5c temp avg difference under load while playing bf3.
m
0
l
August 17, 2012 3:17:22 AM

xDivineCC said:
@Dingo07: ... I think i know what i did wrong. I screwed one tightly and then did the rest. i screwed it on clockwise of course. I kinda get what you mean now by screwing them.

let's hope you didn't crack it from that first screw :( 
m
0
l
August 17, 2012 3:33:27 AM

i didn't. I only did it finger tight and had it all aligned and everything... the card is still working perfectly fine.
m
0
l
August 17, 2012 3:39:10 AM

it's actually best if you go across from screw to screw... from 1 to 4 then 2 to 3

1 - 2

3 - 4
m
0
l
August 17, 2012 3:39:53 AM

I shall try that after i am done something! thanks!
m
0
l

Best solution

August 17, 2012 4:00:55 AM

if you still have more mx-4 give it another try before spending more money. All these thermal compounds shouldn't differ by more than a couple degrees.
You may also want to let it cure for a couple days and see if the temps drop. Yes, the instructions often say 200hours for full cure, but people who tested this said realistically after a couple days you're pretty close to cured.

Be sure to do the cleanup with rubbing alcohol and the tinting technique.
AS's instructions are pretty good no matter what you're applying.
http://www.arcticsilver.com/intel_application_method.ht...

as far as the dot/line or whatever, supposedly shouldn't make a huge difference, but try a dot instead.
Share
August 17, 2012 4:31:59 AM

hey raytseng! thanks for the wonderful info. i did the dot method on my cpu and it works perfectly. I cleaned off the cpu. It's just the gpu that's been reacting. The cpu temperature dropped 4-5 degrees from the pervious idle (added on a nh-d14 + with the mx-4 paste). I'll give it another few days for the GPU temp. I don't think it turned out really well.
m
0
l
August 28, 2012 8:41:39 AM

Best answer selected by xdivinecc.
m
0
l
!