Edit 15/01/2025: I have an updated version of the Thermal Pad Tests pdf, which includes some extra stuff. See post 18.
Thought Id share the results from some tests I did recently after getting myself into trouble!
I hope it may be of some use.
Thought Id share the results from some tests I did recently after getting myself into trouble!
I hope it may be of some use.
Last edited:
Very nice report, thank you!
I did something similar but not as scientific:
I did something similar but not as scientific:
Excellent work, thanks for sharing.
Surprised how much difference the clamp makes, like you say possibly due to it being made from a heatsink itself - I wonder if a more normal clip like these (scroll down) would make the difference less marked.
I've also wondered for a while why more established/branded sellers of alumina insulators always sell them much thicker than what you see on Ali Express etc - perhaps the thinner ones tend to crack over time with repeated expand/contract cycles?
Surprised how much difference the clamp makes, like you say possibly due to it being made from a heatsink itself - I wonder if a more normal clip like these (scroll down) would make the difference less marked.
I've also wondered for a while why more established/branded sellers of alumina insulators always sell them much thicker than what you see on Ali Express etc - perhaps the thinner ones tend to crack over time with repeated expand/contract cycles?
I purchased Alumina Al203 TO-247 insulators a few years ago after experiencing some Keratherm failures. I love those ceramics -- yes, they're thick (0.6mm), but they are robust and work very, very well. Relatively inexpensive and highly recommended.
Oh, and these are the ones I purchased: https://www.aliexpress.us/item/2251...1b721802uepO86&gatewayAdapt=glo2usa4itemAdapt
Regards.
Oh, and these are the ones I purchased: https://www.aliexpress.us/item/2251...1b721802uepO86&gatewayAdapt=glo2usa4itemAdapt
Regards.
My experience of failures with thermal pads is that they were caused by poor flatness of the heatsink, burrs at the mounting hole or over tightening.
Nice work @MistyBlue! Thanks for sharing. Kapton and goop works surprisingly well!
I purchased some aluminum nitride insulators from AliExp a while back. They worked really well. Probably overkill and more expensive than aluminum oxide. But why not?
Current option link
Other post on these
Other other relevant post
I purchased some aluminum nitride insulators from AliExp a while back. They worked really well. Probably overkill and more expensive than aluminum oxide. But why not?
Current option link
Other post on these
Other other relevant post
Thank you for the kind comments guys.
@von Ah - Thanks for the heads up! I didn't see that one. I have just ordered some of those and am keeping the test rig active to test them when they turn up.
Should be interesting!
@von Ah - Thanks for the heads up! I didn't see that one. I have just ordered some of those and am keeping the test rig active to test them when they turn up.
Should be interesting!
Definitely. Alumina, Al2O3, is also known as ruby or saphire or corundum, extremely hard oxide ceramic material. I think pads from it are sintered powder, which is less brittle and cheaper than pure crystals. Oxides of various elements are named like this, see beryllia, silica, germania, zirconia, ceria, yttria, titania...
Magnesium oxide has better performance then alumina, yet I have never seen silpads using it as filler. I made some myself and tested them with my thermal ohmmeter: they were consistently better than commercial types
MgO is hygroscopic and absorbs atmospheric moisture turning to magnesium hydroxide, so no use as an engineering material except perhaps in space applications...
It shares this property with most alkali earth oxides (other than BeO which is highly toxic but an excellent thermal conductor).
It shares this property with most alkali earth oxides (other than BeO which is highly toxic but an excellent thermal conductor).
Last edited:
That's the sound you make when you catch a whiff of that magnesium hydroxide in your amp circuit...
"MgOH!!!"
"MgOH!!!"
Excellent data! Thank you.
May I suggest that you add a table or column graph that shows the thermal resistance of the various pads? I can calculate it from the data provided, but it would make for a one-stop comparison for others.
Thanks,
Tom
May I suggest that you add a table or column graph that shows the thermal resistance of the various pads? I can calculate it from the data provided, but it would make for a one-stop comparison for others.
Thanks,
Tom
Thanks, Tom.
I have a few more things in addition to the AlN pads that I want to test, so I will update the article once I have some more data.
I have a few more things in addition to the AlN pads that I want to test, so I will update the article once I have some more data.
Impressive effort. I investigated some of the data sheets on the thermal insulators and came to the same conclusion that testing was the only way to find the best.
Probably, at a low level, but certainly not when it is sequestered/inactivated in a hydrophobic binder like silicone resin or oil
Moisture can diffuse through a wide range of materials... I suspect its easier cheaper to use stable materials in practice.
I have updated the Thermal pad tests to include Aluminium Nitride, a Bergquist phase change pad, some Boron Nitride thermal paste and some interesting initial experiments with copper heatpipes.
I'm exhausted now and have got the answers I needed.
Hope this helps others.
I'm exhausted now and have got the answers I needed.
Hope this helps others.
Neither will I. I knew the performance was probably not as good as the traditional mica/paste, but I didn’t expect them to be that bad.
The test jig has been decommissioned and the amplifier is now built.
Although, I do have some ideas for revisiting my heatpipe experiments in the future.
The test jig has been decommissioned and the amplifier is now built.
Although, I do have some ideas for revisiting my heatpipe experiments in the future.