Replaced stock thermal pad with copper plates on Gpu

[silver2002]

i finally got around to replacing the thermal pads with copper plates according to plans left by Pascal_TH. Although the sill goes to 100C during DoD:S it does by far cool down faster after. it now idles at 60C with powerplay enabled and stays with about 70C with max preformance.

i did take pics of my install and will get around some time to posting them. I replaced the thermal pads with copper plates from a pair of Vantec Ram heat spreaders i cut them into sqaures and placed them where the pads were with Artic Silver on both sides, and screwed them down tight so the plates dont move. so thanks to all who provided information on this topic.

[DAH]

Well I for one would like to see what you have done. I wonder if perhaps a solid sheet of aluminum would not have worked better?

[silver2002]

Well I for one would like to see what you have done. I wonder if perhaps a solid sheet of aluminum would not have worked better?

i thought about using aluminum 2 but didnt really want to use different metals if the stock cooling was made of pure copper.

ill get those pics up once i get some batteries for my camera.

ps. u have the same thinkpad i do except mine is 2613-ESU

[BradS]

It now idles at 60C w/ powerplay and 70C with max performance, but where was it before. What amount of temperature change did you observe?

[erik]

Well I for one would like to see what you have done. I wonder if perhaps a solid sheet of aluminum would not have worked better?

never, ever mix copper and aluminum.   they are dissimilar noble metals (even though technically both are alloys) and galvanic action will corrode them after gathering moisture in the air.   not a good idea.

besides, copper is a better thermal conductor than aluminum.

[silver2002]

BradS:
It used to idle w/powerplay in the high 60's and with max it was around 78-80C. so there was quite an improvement because now the whole assembly including the metal keyboard base is now a big heatsink. it should improve further upon the arctic silver doing its "burn-in(200 hours)"

erik:
thank for the info iam glad i only used copper (looks like some Chem 101 is still in my memory )

is using a different puirty of copper matter? the fan assembly looks like its pure because fingerprints get stuck to it and the heat spreader ones i used werent that quality?

[DAH]

Well I for one would like to see what you have done. I wonder if perhaps a solid sheet of aluminum would not have worked better?

never, ever mix copper and aluminum.   they are dissimilar noble metals (even though technically both are alloys) and galvanic action will corrode them after gathering moisture in the air.   not a good idea.

besides, copper is a better thermal conductor than aluminum.

Of course you are entitled to your opinion, at 140 degrees Fahrenheit my guess is for most users there is not going to be a major problem with water between the plates. That a solid sheet is more effective that multiple sheets. That oxidation will occur around the perimeter, but within five years the likely hood of the sheet failing is slim, IMHO. A lead based paint would be effective, but as we are not trying to run electricity through these plates corrosion would not be a major problem. At least not in the sense it is for memory devices using different metals.

[GomJabbar]

never, ever mix copper and aluminum.   they are dissimilar noble metals (even though technically both are alloys) and galvanic action will corrode them after gathering moisture in the air.   not a good idea.

besides, copper is a better thermal conductor than aluminum.

Sorry, but copper and aluminum are not nobel metals. To quote Wikipedia: "Noble metals are metals that are resistant to corrosion or oxidation, unlike most base metals. They tend to be precious metals, often due to perceived rarity. Examples include gold, silver, tantalum, platinum, palladium and rhodium." And further: "Copper is considered a base metal as it oxidizes relatively easily, although it does not react with HCl. It is commonly used in opposition to noble metal."

Pure copper and pure aluminum are elements, not alloys. But I will agree that many times these will be alloyed with other metals to enhance certain desired properties. You are correct that copper transfers heat better than aluminum, and that dissimilar metals tend to react with one another. Water can enhance this reaction, but is not necessary. The wrong thermal paste may cause this reaction too. See following thread.

http://forum.thinkpads.com/viewtopic.ph ... 968#159968

[erik]

all metals have nobility... that's from chemistry 101.   i may have worded my response above poorly (sorry, it was late) but copper and aluminum will react with one another.

while a processor may hit high enough temps to essentially "burn off" moisture, the cooling process can form condensation.   that's where water can be introduced into the formula.

[erik]

erik:
thank for the info iam glad i only used copper (looks like some Chem 101 is still in my memory )

is using a different puirty of copper matter? the fan assembly looks like its pure because fingerprints get stuck to it and the heat spreader ones i used werent that quality?

no problem.   i just saw this post after writing my last response.

copper is getting expensive and is sometimes alloyed with other metals to save money.   for example, oxidized copper flashing (like what's seen on roofs) used to turn green where these days it turns brown.   so, if you use copper purchased commercially it can be alloyed.   however, copper used in the electronics industry is usually as pure as it gets.

[erik]

Sorry, but copper and aluminum are not nobel metals. To quote Wikipedia: "Noble metals are metals that are resistant to corrosion or oxidation, unlike most base metals. They tend to be precious metals, often due to perceived rarity. Examples include gold, silver, tantalum, platinum, palladium and rhodium." And further: "Copper is considered a base metal as it oxidizes relatively easily, although it does not react with HCl. It is commonly used in opposition to noble metal."

you need to look at the galvanic series info linked from the above wiki article on noble metals.

[GomJabbar]

you need to look at the galvanic series info linked from the above wiki article on noble metals.

I looked, but that is only a rating (or gradient) from Nobel at the top to Base at the bottom. No distinct division is made between the two. Wikipedia wrote: "A "Galvanic series" is a hierarchy of metals (or other electrically conductive materials, including composites and semimetals) that runs from noble to active, and allows designers to see at a glance how materials will interact in the environment used to generate the series."

My first thought of an analogy was the color wheel. At what point does one color become another? Perhaps a better analogy is income level. There is the super wealthy (Nobel), wealthy (less Nobel), middle class (between Nobel and Base), lower class (more Base than Nobel), and the impoverished (Base). Copper and Aluminum would would compare to the lower-middle class to lower class income level.

No, Aluminum is definitely a Base Metal, and Copper generally so (it depends on which list you look at).
http://en.wikipedia.org/wiki/Metal

[rbena]

May I make a suggestion? In my work in over 30 years as a corrosion engineer, I find it useful to look at:

- the conditions and environment in which the metals exist
- any connection between the metals, and
- the properties / alloys / composition of the metals.


In many cases, metals will behave in a certain manner, except sometimes. And that's when an investigation is conducted, to identify what other real world factors cause an 'expected behaviour'.

So I would suggest rather than focussing too much on the metal properties, including noble, classic oxidation behaviour, etc - to simply focus on the immediate example stated at the beginning of this thread, and identify the environment in which the metal(s) will exist. Also look at any interconnection, and then consider specific metal properties in order to choose the best arrangement.

While aluminum heat sinks can perform quite well in certain arrangements, in the confined space above the GPU, copper may be the most efficient choice.

And a copper-to-copper interconnection is more desirable than a copper-to-aluminum. The 'cyclic' conditions of heat/dry and cool/ 'possible' moisture can introduce a path for bimetallic corrosion, where the aluminum would deteriorate (and risk a reduction in its cooling ability), while the copper would receive the aluminum by-product (and possibly affect its cooling ability).

So the ‘safe option’ appears use only copper - preferably a single strip of adequate thickness rather than sandwiching multiple strips with conductive paste. Then compress the assembly, as has been mentioned elsewhere, to ensure optimum heat transfer and prevent gaps/separation from occurring over time.

I'm not saying variations to this could also not work, but this is my suggestion.

[Pascal_TTH]

i finally got around to replacing the thermal pads with copper plates according to plans left by Pascal_TH. Although the sill goes to 100C during DoD:S it does by far cool down faster after. it now idles at 60C with powerplay enabled and stays with about 70C with max preformance.

i did take pics of my install and will get around some time to posting them. I replaced the thermal pads with copper plates from a pair of Vantec Ram heat spreaders i cut them into sqaures and placed them where the pads were with Artic Silver on both sides, and screwed them down tight so the plates dont move. so thanks to all who provided information on this topic.

So we both enjoy the coolest T60p !


R61 have mixed fan assembly with copper and aluminium :

http://forum.tt-hardware.com/fichiers/u ... 719426.jpg