Overclock.net banner
1 - 16 of 16 Posts

·
Registered
Joined
·
466 Posts
Discussion Starter · #1 · (Edited)
Hi guys,
I just sold my 8086K (sigh!) and I am experimenting cooling solutions with a Celeron G4930.
I de-lidded it using liquid metal and I'm still using my AIO x3 360 fan water-cooled loop.
Then I placed on the CPU a COPPER FOIL OF 3 MICRONS, instead thermal paste.

Font Circuit component Hardware programmer Rectangle Electronic component


Font Eye shadow Electronic device Gadget Auto part


I'm running P95 Small FFTs from 30 minutes and the temperature peak is 47 Celsius degrees, with an average of 44 Celsius degrees, room temperature of 26 Celsius degrees.

Rectangle Azure Font Screenshot Software


Font Material property Screenshot Parallel Pattern


I can not say that is an ideal testing situation, a Celeron G4930 is basically a "cold CPU" especially using liquid metal mods and water-cooling solutions.

Another consideration, is not possible to create copper foils with copper particles on it, on a similar way of graphite pads?
It could be a cooling solution?

Let's go on experimenting ;)
 

·
Registered
Joined
·
112 Posts
This is an interesting topic. I remember awhile back someone on [H] doing experiments with Indium foil, IIRC it performed a bit worse than good thermal paste but foils could have advantages in stability as drying and pump-out isn't a concern as it is with paste.

I think lapped IHS + lapped cooler with copper or Indium foil between could be interesting.

Curious what your findings are with higher-power CPUs and this method...
 
  • Rep+
Reactions: Bride

·
Not a linux lobbyist
Joined
·
2,925 Posts
I think they sell gold foil somewhere.
 
  • Rep+
Reactions: Bride

·
H₂O Aficionado
Joined
·
5,673 Posts
Do you have a baseline temp recorded prior to the installation of copper foil to categorize the impact of the results?

I'm a bit hesitant to believe it will surpass paste. If there are voids in the IHS or heatsink, the foil may not penetrate and air (insulator) will likely still be present. I've never used copper foil before but I imagine you may get air trapped easily without a good install of the foil.
 

·
Dark and Dropped SUV
Joined
·
4,143 Posts
  • Rep+
Reactions: Bride

·
Registered
Joined
·
466 Posts
Discussion Starter · #6 ·
I know that with a foil is impossible to "fulfill" the surface irregularities, but the thickness of this copper foil is 3 microns, so to run some additional tests can be interesting...
unluckily I don't have a "power hungry" CPU right now.

This is a P95 made previously with a thermal paste:

Rectangle Font Screenshot Parallel Software


Font Material property Screenshot Parallel Pattern
 

·
Registered
Joined
·
466 Posts
Discussion Starter · #7 ·
Do you have a baseline temp recorded prior to the installation of copper foil to categorize the impact of the results?

I'm a bit hesitant to believe it will surpass paste. If there are voids in the IHS or heatsink, the foil may not penetrate and air (insulator) will likely still be present. I've never used copper foil before but I imagine you may get air trapped easily without a good install of the foil.
yep, I have your same doubts. I would like to understand if copper foil can be produced with particles on it of the same material, but I am not sure if in any case the problem can be solved. Graphite physical characteristics and fiber structure are different from a metal as copper (maybe what i just said is not correct)
 

·
Registered
Joined
·
112 Posts
I would like to understand if copper foil can be produced with particles on it of the same material, but I am not sure if in any case the problem can be solved.
What immediately comes to mind here is metal foil coated in thin layer of sintered metal. Not sure if the grain size of readily-available sintered copper would be fine enough to meaningfully fill voids on the scale needed though.
 
  • Rep+
Reactions: Bride

·
Registered
Joined
·
466 Posts
Discussion Starter · #9 ·
What immediately comes to mind here is metal foil coated in thin layer of sintered metal. Not sure if the grain size of readily-available sintered copper would be fine enough to meaningfully fill voids on the scale needed though.
it could be a micron copper foil with a micron layer of sintered copper, but i don't know if the industrial process justify the application... in the end if it's expensive, there are a mulitude of other solutions
 

·
Registered
Joined
·
466 Posts
Discussion Starter · #11 · (Edited)
I tried to push the core voltage at 1.45v and with an high quality thermal silicone, the temperature was almost at throttling point (100C) after few minutes, room temperature around 22 degrees. Then I applied again another layer of micron copper foil and I was a little surprised about the results:
after 30 minutes of P95 Small FFTs, the average temperature was sitting around 57 Celsius degrees, with a minimum of 34 and a maximum of 59...

Rectangle Azure Font Screenshot Parallel

Font Material property Screenshot Parallel Pattern
 

·
Registered
Joined
·
466 Posts
Discussion Starter · #12 · (Edited)
Few considerations after these quick tests...

PROS
  • copper foil of 3 microns it's affordable
  • it's easy to remove
  • decent thermal results, especially in not very stressful scenarios
  • long term durability and performance stability (to be verified, but if we look at the metals characteristics...)
  • can solve some problem in modding side, as for example the chemical reaction between liquid metal and copper heat-sinks
  • can be found also in powder of different particles sizes and mixed with other thermal silicone based compounds

CONS
  • difficult application on the surfaces, it's a micron thickness foil
  • performance and stability results have to be verified in multiple different scenarios
  • electrically conductive but should not be a problem in foil format
 

·
Registered
Joined
·
466 Posts
Discussion Starter · #13 ·
core voltage at 1.50v , room temperature around 21 degrees.
After 2 hours of P95 Small FFTs, the average temperature was sitting around 60 Celsius degrees, with a minimum of 35 and a maximum of 61...

Rectangle Font Screenshot Software Parallel

Font Material property Screenshot Parallel Pattern
 

·
Registered
Joined
·
466 Posts
Discussion Starter · #14 · (Edited)
another test on my R9 290... First of all I down clocked it, then I applied at the corners of the die few drops of thermal paste, in order to keep the copper foil down for an easier application of the heat-sink,
Brown Rectangle Font Wood Cosmetics

Brown Eye Font Cuisine Rectangle

then I run a Furmark 1080p,
with thermal paste, 36 - 45 - 62 Celsius, room 21 Celsius
Font Rectangle Screenshot Parallel Number

with micron copper foil, 33 - 53 - 68 Celsius, room 21 Celsius
Font Rectangle Screenshot Parallel Number

900MHz core clock 1250MHz mem clock, GPU voltage core 0.975 - 1.005 - 1.044,
in-game temps with World of Tanks, 47 - 53 - 57 Celsius, room 21 Celsius
Font Material property Screenshot Parallel Pattern
 

·
Registered
Joined
·
466 Posts
Discussion Starter · #15 ·
1100MHz core clock 1600MHz mem clock, GPU voltage core 0.981 - 1.165 - 1.244,
in-game temps with World of Tanks, 44 - 68 - 80 Celsius, room 21 Celsius

Font Material property Screenshot Parallel Pattern

1100MHz core clock 1600MHz mem clock, GPU voltage core 0.988 - 1.106 - 1.244,
Furmark 1080p, 38 - 65 - 87 Celsius, room 21 Celsius

Font Material property Rectangle Screenshot Parallel


my GPU is in bottleneck with my CPU, but the results are consistent in any case
 

·
Registered
Joined
·
466 Posts
Discussion Starter · #16 · (Edited)
Few considerations after these quick tests...

PROS

  • copper foil of 3 microns it's affordable
  • it's easy to remove
  • decent thermal results, especially in not very stressful scenarios
  • can be used for CPUs and GPUs
  • long term durability and performance stability (to be verified, but if we look at the metals characteristics...)
  • can solve some problem in modding side, as for example the chemical reaction between liquid metal and copper heat-sinks
  • can be found also in powder of different particles sizes and mixed with other thermal silicone based compounds
  • can be mixed or grind with any thermal paste, due the micron size consistence
  • can be melted on heat-sinks with a flame torch

CONS

  • difficult application on the surfaces, it's a micron thickness foil
  • performance and stability results have to be verified in multiple different scenarios
  • electrically conductive but should not be a problem in foil format
 
1 - 16 of 16 Posts
Top