Thermal Paste Tests!
- Thread starter BEn_2600+
- Start date
V|per said:Confirms what i've been telling people for ages
MX-1 ALL THE WAY - especially for the price...
I'm happy with the temps I got from MX-1. However the tube doesn't seem to last long. But then again it doesn't matter a great deal to most people. Arctic Silver could come in a 10KG bucket for the same price, and we'd still only need a couple of pea sizes of it.
Jokester said:
I've yet to see any tests that I like apart from the one you did Jokester... and with all due respect I'd have liked to have seen more compounds tested, for longer, and with more tests.
From my own experience there is a significant and noticable change in temps by wiping off a block and reapplying it with the same paste.
I suspect if we had 10 different people test 10 different paste we'd get a huge variety of results.
I've been thinking about how I apply the paste myself, and I'm coming to the conclusion though logic - unfortunately I don't have the time to test - that the best way to do it would be to coat the block and IHS, wipe them lightly so they are really just dirty with the paste rather than coated in it, and then apply a very small amount on the center and let pressure do the rest.
That way all the micro valleys get filled initially, and the deep ones stay filled even when it's wiped. This would avoid air being trapped in those when a bead of paste is applied by pressure from the heatsink.
I also think we don't put enough attention on pastes. We're all concerned about a few degrees drop in temps through water or air cooling, but when applying the paste properly can make the same or greater an effect, I wouldn't mind knowing more about it.
The less used the better, this is why Liquid Pro is so good as you can't realistically mess it up. As you say add to much of a thick paste and your temps will be a few degrees worse than they would have been otherwise.
Jokester
Jokester
Jokester said:
On the XS thread there was a bit of criticism about the liquid pro forming a near permanant bond between the IHS and the heatsink. I believe someone posted a picture after he'd removed the IHS and finally got them separated.
IHS/Heatsink "bonded" by Liquid Pro post.
Then again with a lapped waterblock and a bare core I lifed the CPU right out of the locked ZIF socket (no damage thankfully) when I tried to remove the block last time. Basically a vacuum had formed - same way as a mug of tea sometimes lifts the mat.
But as we seem to be approaching the limits of what watercooling can do (brought on by a huge degree from IHSs) then we need to look elsewhere for the odd degree or two.
Jokester said:
Any ideas if the results are good because it's a great thermal interface, or just because it allows a very thin layer?
I've a waterblock on order (new XT DI) and I'm thinking I might try putting MX-1 on, and then scraping it back off with a credit card (as much as I can scrape off keeping the card flat) and then putting a dot in the middle of the cores and seeing how that goes.
Then again if the Liquid Pro is going to be better, I might go for that - if I can find it.
The Halk said:
I would say it's a combination of the both, as it seems to bond directly into the surface (it's pretty weird as it seems to get absorbed into the surface of a copper block). And becuase you can spread it ultra thin if you follow the instructions (this makes it a bit of a problem if the IHS isn't flat) you don't need to worry about overdoing it.
Jokester