Beavinator
1 kW
- Joined
- Aug 17, 2010
- Messages
- 362
Where is a good place to buy that white thermal paste that is on the mosfets on the Infineon controllers?
Thermal paste
- Thread starter Beavinator
- Start date
spinningmagnets
100 TW
"Frozen CPU.com", computer cooling component super-store...
http://www.frozencpu.com/cat/l1/g8/Thermal.html?id=Kqm28odX
http://www.frozencpu.com/cat/l1/g8/Thermal.html?id=Kqm28odX
AussieJester
1 TW
Any automotive store also stocks thermal paste i forget now what/where its used
in the cars underbonnet electronics, but they have it, comes in lil plastic sashes cost a couple bucks... If you go for PC thermal paste grab Arctic Silver its the ducks nuts IMO.
KiM
in the cars underbonnet electronics, but they have it, comes in lil plastic sashes cost a couple bucks... If you go for PC thermal paste grab Arctic Silver its the ducks nuts IMO.
KiM
Thermal paste is used, among other things, between ignition control modules and distributor base plates in automotive applications. The RC crowd also uses this paste between electric motor cases and "slip on" or "clip on" heat sinks.
It's always a good idea to have some of this stuff around.
It's always a good idea to have some of this stuff around.
liveforphysics
100 TW
I use Artic silver, because I bought lots of it a couple years ago.
I saw a pretty elaborate test done with a lot of different types though. Water was absolutely the best thermal paste, but of course it evaporates away quickly (impractical obviously)... Artic silver was the best of them, but by a very very small margin over the cheapest bulk china goop they sell in big tubs for almost nothing on sites like dealextreme. All the thermal compounds were inside of a percent or two of each other in performance, then a drop of water was a large leap ahead.
I saw a pretty elaborate test done with a lot of different types though. Water was absolutely the best thermal paste, but of course it evaporates away quickly (impractical obviously)... Artic silver was the best of them, but by a very very small margin over the cheapest bulk china goop they sell in big tubs for almost nothing on sites like dealextreme. All the thermal compounds were inside of a percent or two of each other in performance, then a drop of water was a large leap ahead.
Also be aware that you may not just need paste, since things may not fit very closely in your controller, and you could probably improve it significantly by fixing that.
This thread is about doing just that, although it is just getting started:
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=21829&start=0
If you fix the gaps and such, and apply only *just* enough paste to fill the scratches/etc, then thermal paste degradation over time will be far less relevant as it won't be what is conducting most of the heat out of your FETs; the direct metal contact will (where applicable).
This thread is about doing just that, although it is just getting started:
http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=21829&start=0
If you fix the gaps and such, and apply only *just* enough paste to fill the scratches/etc, then thermal paste degradation over time will be far less relevant as it won't be what is conducting most of the heat out of your FETs; the direct metal contact will (where applicable).
oldhaq
100 W
I've been using Ceramique instead of Arctic Silver 5 mainly because AS5 is slightly capacitive.
I also use kapton tape instead of the much thicker silicone insulator sheet that is prevalent in controllers, along with keeping the thermal compound smear as thin as possible helps to get the heat from transistors to the heatsink quicker.
I have also had far less problems with burnt out mosfets using Ceramique.
I also use kapton tape instead of the much thicker silicone insulator sheet that is prevalent in controllers, along with keeping the thermal compound smear as thin as possible helps to get the heat from transistors to the heatsink quicker.
I have also had far less problems with burnt out mosfets using Ceramique.
liveforphysics
100 TW
I just stumbled across this test, if you check this extremely elaborate test of 80 different compounds tested against each other, the results are a little surprising.
http://benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=62
http://benchmarkreviews.com/index.php?option=com_content&task=view&id=150&Itemid=62
BatteryMooch
10 kW
That matches my testing. Your links shows only a 4.6 degree-C temp difference between the lowest and highest performing compounds and there are so many other variables with his setup that the results essentially show that the least expensive compound is just as good as the most expensive. This matches the results of other posted tests AND the published thermal resistance numbers for many of the compounds.
- Pick a compound based on the application and components being used.
- Do not use silicone (oil) compounds at higher temperatures as the silicone evaporates and the compound's thermal resistance increases.
- Be aware that many compounds can be "pumped out" from between certain surfaces if there is a lot of thermal cycling going on. Be sure the replace the compound as necessary. Think of it as an oil change.
- If using phase-change compounds, do not tighten unless you preheat the compound so it changes to "water" and you can get less than a 0.1mil layer of the stuff left in between the two metal surfaces (my favorite: Aavid's UltraStick).
- Never exceed the compound's long-term max rated temperature. Don't know it? Don't use that compound.
- Pick a compound based on the application and components being used.
- Do not use silicone (oil) compounds at higher temperatures as the silicone evaporates and the compound's thermal resistance increases.
- Be aware that many compounds can be "pumped out" from between certain surfaces if there is a lot of thermal cycling going on. Be sure the replace the compound as necessary. Think of it as an oil change.
- If using phase-change compounds, do not tighten unless you preheat the compound so it changes to "water" and you can get less than a 0.1mil layer of the stuff left in between the two metal surfaces (my favorite: Aavid's UltraStick).
- Never exceed the compound's long-term max rated temperature. Don't know it? Don't use that compound.
liveforphysics
100 TW
Some aspects I thought were particularly interesting is that the heat transfer particle choice didn't seem to make much difference.
You find some pure silver particle compounds like arctic silver 2 performing as badly as any compounds, while other arctic silver compounds perform with the best. Yet the best performer uses alumina, a ceramic with poor heat transfer, and others use various different types of exotic materials like diamond or carbon nano-tubes, and yet perform worse than the compound using dirt cheap ceramic.
What I gathered from this testing, was that the materials used in the compound seem to matter the least, meaning the particle size (and maybe shape if packing effects are a factor?) I'm thinking the fluid the particles are suspended in are likely to be the most important aspect of performance.
You find some pure silver particle compounds like arctic silver 2 performing as badly as any compounds, while other arctic silver compounds perform with the best. Yet the best performer uses alumina, a ceramic with poor heat transfer, and others use various different types of exotic materials like diamond or carbon nano-tubes, and yet perform worse than the compound using dirt cheap ceramic.
What I gathered from this testing, was that the materials used in the compound seem to matter the least, meaning the particle size (and maybe shape if packing effects are a factor?) I'm thinking the fluid the particles are suspended in are likely to be the most important aspect of performance.
Given how well water works, that doesn't surprise me.
Remember that a fluid is able to fully contact all the surfaces it is in contact with, and circulate (convect) heat within it's volume. A solid doesn't do that, even a particulate suspended in a fluid, because often the particles are as large as the gaps they are to fill (or nearly so).
Remember that a fluid is able to fully contact all the surfaces it is in contact with, and circulate (convect) heat within it's volume. A solid doesn't do that, even a particulate suspended in a fluid, because often the particles are as large as the gaps they are to fill (or nearly so).
BatteryMooch
10 kW
I lean towards almost none of this having much of an effect if 80 different compounds have less than a 5C spread from best to worst.liveforphysics said:
Some compounds though have very specifically sized particles (typically glass spheres) that help maintain a specific spacing between the two surfaces the compound is applied to in order to keep the surfaces insulated (electrically) from each other. The thermal resistance of these compounds can be a lot higher than others depending on the size of the spheres.
