Thermal compound W/mk rating BS

rg12

100 kW
Joined
Jul 26, 2014
Messages
1,584
Am looking for a good thermal paste with high conductivity for my heatsinks in order to replace my chinese one that is rated as 1.8W/Mk.

I went through amazon and found loads of brands and I sense a total bs market selling 1g at like $10 while each brand declares a higher and higher conductivity rank (W/mk) like 10.2, 13.4, 16.5 and all those uneven numbers that sound accurate while no one can really tell and no one really tests the rated conductivity in a lab.

My question is if anyone here has a recommendation for a good thermal paste that has been really tested for it's conductivity ranking at a reasonable price?

I use large heatsinks so 1g for 10usd isn't for me.
 
Not everything is suitable for Amazon shopping.

mouser, digikey, allied, newark, arrow, element14, avnet

even McMaster
 
rg12 said:
Am looking for a good thermal paste with high conductivity for my heatsinks in order to replace my chinese one that is rated as 1.8W/Mk.

I went through amazon
<snip>
If you want actual specs for real products, you'll need to look at big distributors (mouser, digikey, etc).

Amazon sellers (and any other giant marketplace like that) are often only there to make fast money, not to get you a product that will do what you want or need, so they have no reason to provide real specifications, or even to provide real products (meaning, ones actually tested for real to do what they say). Some of the sellers do actually have "real" stuff with "real" specs...most of them don't care; no reason to.


As for whether the difference between thermal compounds will make a significant difference to your projects' heatsinking...it's much more likely that your application method will make a bigger difference than that.

And get compounds that don't use volatile oils (or creeping fluids like silicone oil, etc) as solvents, because those that do will dry out over time and fail to do the job they are there to do, with the incorrect way that most people apply them.
 
Good thinking, thanks alot!

Just went through mouser and digikey...
One brand that is showing in both called Chipquik has a rating of 79W which I call total BS, then they have a syringe with 8.5W, then another can with 4.3W (which they claim is ultra conductive)...

Another brand that looks legit (just a plastic industrial looking thing) has 0.63W in the datasheet (which I believe but seems mega low).

Many of them with silicone..

Got even more confused :lol:
 
Beware many of the ultra high thermal conductivity pastes these days are liquid metal alloys that may not be suitable for what you're doing.

The "tech ingredients" youtube guy makes and sells some surprisingly good paste now.
 
I buy Ford. Ford thermal paste. Motorcraft.

Lol.

Probally a waste of money though. Probally supplied by the cheapest bidder. WaddoIknow.

However, when comparing specs, I do trust the printed documentation of the likes of BMW, or Ford, or GM, in lieu of the specs from an " Amazon Seller". Dont get me wrong, I got five little cans of the cheapo stuff, that overstate the ratings for sales, for sure... Fake W/ numbers...

....but for the important stuff i would want something OEM with the liability of operation behind it. For cheap. Ford, Chevy, Kia, BMW, ect, all require a heat sink paste in their parts inventory for repairs to be enacted.


For the expensive stuff? I would try to get something like a Dow Corning 340. I do trust those specs. Them numbers.

https://www.ellsworth.com/globalassets/literature-library/manufacturer/dow-corning/dow-corning-brochure-enhancing-the-reliability-of-your-electronics-designs.pdf
 
liveforphysics said:
Beware many of the ultra high thermal conductivity pastes these days are liquid metal alloys that may not be suitable for what you're doing.

The "tech ingredients" youtube guy makes and sells some surprisingly good paste now.

Im using it for mounting an alloy heatsink to an alloy plate.
Why wouldn't a liquid metal paste work?
and why you guys don't like silicone? I see most use silicones
 
thermal paste is supposed to fill the micro voids between 2 flat smooth surfaces with as thin a layer as possible. so viscosity and particle sizes have to be low to allow it to spread and for excess material to squeeze out if necessary

don't be deceived by w/mk ratings. a block of silver has high w/mk but would be terrible as a thermal paste as-is :lol:

silicones tend to creep over time thus drying out the interface
 
rg12 said:
liveforphysics said:
Beware many of the ultra high thermal conductivity pastes these days are liquid metal alloys that may not be suitable for what you're doing.

The "tech ingredients" youtube guy makes and sells some surprisingly good paste now.

Im using it for mounting an alloy heatsink to an alloy plate.
Why wouldn't a liquid metal paste work?
and why you guys don't like silicone? I see most use silicones

Look into the effects liquid gallium compounds have on aluminum. You might find some combination of materials thats safe with it, but in testing thermal compounds, it's your surface flatness and clamp load that is ~98-99% of the joints thermal transfer performance, and the best compound performance comes from the types that are thin enough they can get out of the way from holding your flat surfaces spaced off from each other.
 
liveforphysics said:
rg12 said:
liveforphysics said:
Beware many of the ultra high thermal conductivity pastes these days are liquid metal alloys that may not be suitable for what you're doing.

The "tech ingredients" youtube guy makes and sells some surprisingly good paste now.

Im using it for mounting an alloy heatsink to an alloy plate.
Why wouldn't a liquid metal paste work?
and why you guys don't like silicone? I see most use silicones

Look into the effects liquid gallium compounds have on aluminum. You might find some combination of materials thats safe with it, but in testing thermal compounds, it's your surface flatness and clamp load that is ~98-99% of the joints thermal transfer performance, and the best compound performance comes from the types that are thin enough they can get out of the way from holding your flat surfaces spaced off from each other.

Makes perfect sense...
Any recommendations for a specific brand?
 
Been mentioned once already, but here is a link:
https://www.youtube.com/watch?v=Glfak2B9J5U

Might also be good to look at the other side of the problem too. Why are you pumping so much waste heat into the air? Possibly better components, mosfets, etc would be more efficient and could get by with normal heat sink compound?
 
4πr^2 said:
Been mentioned once already, but here is a link:
https://www.youtube.com/watch?v=Glfak2B9J5U

Might also be good to look at the other side of the problem too. Why are you pumping so much waste heat into the air? Possibly better components, mosfets, etc would be more efficient and could get by with normal heat sink compound?

Massively over powering a flipsky 75100 pushing 110A battery current and 170A phase current while maintaining 90c temp in the mosfets via a custom copper heatsink that acts as the lid of a biffy scooter.
It has the same mosfets as the Nucular 6F but am cracking it way over that but not crossing the max 192A mosfet limit or 100c temp limit.
 
What do you guys think of this one?
https://www.digikey.co.il/en/products/detail/techspray/1978-1/7914539

Thats the datasheet:
https://www.techspray.com/content/msds/TS-1978_ENG_TDS.pdf

It's non silicone and industrial looking which seems better to me than fake ratings of over colorful tubes with stupid branding names like "sub zero, crazy, extreme"...
 
rg12 said:
What do you guys think of this one?
https://www.digikey.co.il/en/products/detail/techspray/1978-1/7914539

Interestingly pricey.... 300$/lb.... 4oz. tube is 135$. Very wide range of temperature flexibility. I wonder why its so expensive and/or if there is a cheaper "clone" alternative ( without all the fancy language marketing).
 
Hi everyone. :)
I'm basically lurking and learning atm, but as a computer overclocker I can speak on thermal compound:
What you want to do is search for:
Thermal compound review roundup CPU
(Look at the dates: Things have improved since 5 years ago)

Reason being; CPUs have a very high Thermal Density. (lots of watts to get out of a small contact area)
So the properties you are looking for in a paste are all well reviewed and catalogued on computer hardware sites.
Where whatever kept the CPU coolest will likely work best.

NB that the paste with the best W/mk doesn't always win:
Sometimes those pastes are too thick and don't squeeze out to leave the thinnest layer possible.
So basically the better your 2 surfaces mate/fit together; the better the runnier pasts work.
If the surfaces don't match well and there would be large voids; go for the best W/mk paste.


Here's a review eg:
https://www.tomshardware.com/best-picks/best-thermal-paste

NB that the Gallium based liquid metal 'compounds' mentioned, with very high W/mk will do VERY nasty things to aluminum surfaces.
As in Dissolve the aluminum to the point it was used to sabotage aircraft during the 2nd world war!!

Gallium based liquid metals may well work very well between the steel laminates and the pressed steel stator center
(An overlooked area of improvement IMHO..?)
BUT
You are probably better off with a good normal paste and an aluminum center, due to the higher Thermal conductivity of alu vs steel and and the greater volume of alu used, acting as a larger heat 'soak'.

I would also look at Arctic Silver Thermal Adhesive or Arctic Alumina Thermal Adhesive to glue the magnets in if using Statorade...
IIRC 7.5 to 9 W/mk vs 1 to 1.5 for the adhesives currently used.

BTW;
How come no one machines fins into the steel magnet ring when they turn them down for weight savings..??

And where are the side covers turned out of old Viscous Fan Couplings/clutches already!???
You guys all asleep at the handlebars!?? :D
https://www.bing.com/images/search?q=FAN+VISCOUS+CLUTCH&form=HDRSC2&first=1&tsc=ImageHoverTitle
 
Back
Top