Discussion in 'Hardware' started by jizwizard, 6 Sep 2011.
i could swear i read something about high-k breakdown with high voltages...must research
have you tried lapping the h60 the finish on that doesn't seem that great might give a rediction in temps
It is much safer and as beneficial to lap your heat sink, plus you can wet sand your heat sink and that is how you really get is smooth.
However if your CPU is concaved you may gain nothing or even gain temps, same goes for CPU. It is best to do them both, so you know both are flat. 99% of the time neither are flat.
have thought of lapping the h60 but it could be awkward to do will have a look later to see if its feasable
To think, i'm worried about accidentally scratching mine when building tonight. Suddenly, i don't feel the need to be quite so delicate.
Deep scratches aren't good, they'll adversely affect cooling performance.
Not to worry you or anything
Well, i'm hoping to get it in with no scratches tbh.
Can't see how you would scratch it to be honest.
Unless you practise being a sword wielding ninja as you insert the CPU and get one of your moves wrong and slice straight through it
just dont drop it!
Back on topic. I wouldnt get too hung up on a mirror finnish when lapping your cpu n cooler. The point is to make it flat to increase metal on metal contact with less TIM. why polish something you wont ever see? unless they invent see through copper lol
Sent from Bittech Android app
Surely a mirror finish would increase heat dissipation though? The flatter a surface the more light bounces back in the same direction it hit the surface. Meaning that a flatter (thus more mirror like) surface should give a better, if marginal dissipation increase.
(none of what I just said might be correct, it's just how I applied my A Level physics knowledge )
A none mirror surface would bounce light in more directions. Black line at top shows light source/base of cooler which in my overly simplified example is totally flat.
More mirror like, bounces light generally more towards source.
As you can see the cooler would fit better on a mirror like surface. Oh and if all of the above is complete hogwash, I apologise
Not fat from hogwash
Theres no light involved so it doesnt work that way.
The princaple is for both the cooler surface and cpu surface to be as flat as possible in order to reduce areas in the mating surfaces where there would be more TIM than others.
The ideal scenario is to have the cooler and CPU both PERFECTLY smooth so that they are 100% in contact without the aid of TIM, this would provide the most efficient heat transfer from CPU to cooler.
You cant get 100% smooth, but lapping gets you close.
The reason people post pictures of those shiny mirror copper surfaces is to show everyone how flat the CPU is, not how shiny they made it. The reflection has no visible distortion/warping, so is therefore flat. Flatter = Improved heat transfer.
There cant be much difference between a CPU with a mirror finish and a CPU where only the grey metal coat has been lapped away... both have removed the majority of surface imperfections.
Shiny is not important as long as there is flat.
But surely the shinier it is the fewer surface imperfections there are?
Wow that is really surprising how effective it is.
But surface imperfections dont make any difference in the real world.
Dont forget, TIM is not liquid, it's metallic paste.
My point is that you don't need to lap as far as a mirror finish because the TIMs metallic particles are far bigger than any scuffing caused by 1500g sand paper.
But if the scuffing is smaller than the TIM particles, then you'd be left with spaces where the TIM can't get, hence why the mirror finish would help.
It might not make a negligible real world difference, but if you're gonna void your warranty, you may as well do it properly
noob question here but what exactly is lapping?
edit: nevermind. i read the thread
Lapping is perfectly mating two (usually metal) surfaces to each other so they either sit perfectly flat to each other or slide over each other with the minimum resistance.
This is usually accomplished with an abrasive paste.
The idea here was to smooth the metal surface of the CPU to eliminate any microscopic irregularities in the surface which can trap air bubbles and hinder heat transfer.
In purely engineering or gunsmithing terms (my experience of lapping), the CPU has not been truly lapped unless the same has been done to the contact plate on the CPU cooler - ie both surfaces have been matched to each other.
But it looks nice and seems to have worked even better with temp drops of 5C!
right thats it i will do the h60 over the weekend and post the results and i will also pick up some 2000 and 2500 grit to get them both to a mirror finish only problem is the h60 has some allen key bolts in the bottom does anybody know if the plate can be removed so i can lap it?
Separate names with a comma.