• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Aww shoot. Forgot and used "pea" method for TIM on my OCZ Vendetta.

VirtualLarry

VirtualLarry

No Lifer
So damn used to regular AMD and Intel heatsinks.

The OCZ Vendetta is like the CM 212+ kind of, it has three direct-contact heatpipes, with an aluminum base. However, there are some grooves where the base meets the heatpipes.

Normally, to apply these kinds of heatsinks, I apply a line of TIM along each of the heatpipes themselves.

But for some reason my mind blanked, and I just put the "pea" of TIM onto the CPU itself, not the heatsink.

How bad do you think that would affect it? I hesitate to re-seat the heatsink, it was kind of a PITA to get on.

In UEFI BIOS, CPU temp at idle was 35C.
 
It's hard to say. You're getting better thermals simply by use of the aftermarket heatsink with heat pipes. However, given that the bottom surface of the Vendetta is not flat, and you applied only a pea sized amount, it's possible that the TIM isn't completely covering the contact area.

Kind of hard to check and see with it seated. Although, the temps are ok, so maybe no need to worry?
 
You might be causing hot points too, a well cooled heat pipe area with hot spots in between. Take it off and do it again just to be safe.
 
Hmm, another question. The board I'm using, the ASrock 990FX Extreme4, has the AM3+ heatsink mounts such that the heatsink that I'm using points vertically, when the mobo is in the case. Thinking that I would want the airflow to exhaust through the top fan, I set up the Vendetta such that the fan side was on top, and the air is flowing out through the fan that way.

Is that ok to put the fan pulling air through the heatsink, rather than pushing it?
 
I did the pea method with an A70 which has the same grooved heatpipe/base that you are talking about. Temps were ok, but the TIM spread was less than ideal. All in all I believe people blow the whole TIM application proceedure out of proportion. Only a truly horrific mounting attempt will show terrible temps. I have performed no TIM tests along with lots of TIM tests. I would say 3-7c is the difference between a good application and a truly terrible one.

If it were me I would re do it just to have the spread uniform and not create hot spots.
 
Are "hot spots" really an issue here? Isn't that what the IHS supposed to help with?

If load temps are fine, then maybe the pull configuration is ok too. XD
 
I read this and start questioning myself as well. Then I thought about the IHS... Pea sized method should of been fine.

I took my heatsink off after 2 months to find that the TIM didn't even spread much from the center. Temps were fine still.
 
The important part is to cover the area of the ihs right over the silicon, which is of course a smaller patch than the whole ihs. As long as its doing that, temps will likely be passable.
 
VirtualLarry said:
But for some reason my mind blanked, and I just put the "pea" of TIM onto the CPU itself, not the heatsink.

How bad do you think that would affect it? I hesitate to re-seat the heatsink, it was kind of a PITA to get on.

In UEFI BIOS, CPU temp at idle was 35C.
Click to expand...
I'd leave it as is unless you're feeling uncomfortable with your load temps. What you did isn't catastrophic but will certainly affect the temps some. Those heatsinks are tricky and difficult to get done perfect without previous experience. I wouldn't bother to clean/re-apply unless load temps are high. What TIM did you use by the way?
 
Last edited:
I'm running OCCT 4: 64-bit linpack, and temps according to "CPUTIN" are 48, and slowly creeping up.

Core temps went from 6C at idle, to 27C.

In my other machine, with the 4-heatpipe AMD heatsink, and an Antec 300 case full of fans, and four 9600GSO 96SP graphics cards (78C, 72C, etc.), CoreTemp reports my temp as 45C.

Edit: CoreTemp 1.0 also reports 27C, so I think I'm good.
 
Last edited:
A pull fan setup might work better if there was a shroud or something to help draw the airflow through the fins from the back side of the cooler; without anything to direct where the airflow is coming from, the fan draws air from everywhere around the fan itself.

After I pulled my 212+ off and was fooling with it, I noticed a gap both top and bottom where air being pushed into the cooling fins was leaking out of... quite a bit, I could feel it with my hand. That was one of the reasons I went to a push-pull setup on my 212+, just to try to get more air through the exchanger rather than blow much of it out the top and bottom.

In the photo below I outlined the gap between the fans and the heat exchanger... if you could seal that up, the cooling efficiency would be far greater.

HAF922004-1.jpg


Also, even installed, my 212+ has some 'wiggle-room...' it will rotate a few degrees either direction. If yours still moves a bit, you can wiggle it some and try to seat the base better, working the TIM out across the surface with movement.
 
Charlie98 said:
A pull fan setup might work better if there was a shroud or something to help draw the airflow through the fins from the back side of the cooler; without anything to direct where the airflow is coming from, the fan draws air from everywhere around the fan itself.

After I pulled my 212+ off and was fooling with it, I noticed a gap both top and bottom where air being pushed into the cooling fins was leaking out of... quite a bit, I could feel it with my hand. That was one of the reasons I went to a push-pull setup on my 212+, just to try to get more air through the exchanger rather than blow much of it out the top and bottom.

In the photo below I outlined the gap between the fans and the heat exchanger... if you could seal that up, the cooling efficiency would be far greater.



HAF922004-1.jpg


Also, even installed, my 212+ has some 'wiggle-room...' it will rotate a few degrees either direction. If yours still moves a bit, you can wiggle it some and try to seat the base better, working the TIM out across the surface with movement.
Click to expand...

I like your cable management, very simple, but clean looking. Doesn't look like the case is setup for cable management either.
 
SLK said:
I like your cable management, very simple, but clean looking. Doesn't look like the case is setup for cable management either.
Click to expand...

Thank you... that's actually the 2nd time around, after I added a 2nd HDD and a Lamptron fan controller. Last but not least will be when I install the GPU, but I planned for that. There is actually gobs of room in this thing, more importantly, plenty of room behind the mobo bulkhead to stash the unused PSU cables.

I'll be posting complete pics in the GH section when I have time to write it up.

...now we return to the regularly scheduled OP's question... :whiste:
 
Well, I upped the voltage to 1.4v in BIOS, left Load-line on "AUTO", bumped the FSB from 200 to 230 (3.1Ghz), and tried OCCT 4:64-bit linpack (90%). Temps got up to 63.5C before I stopped it.

That was with the old TIM application, which I remounted. This time, I put a small line on each heatpipe, but I may have put actually too much TIM this time, as during the mounting process, the CPU got covered, and it seemed like there might have been some excess.

I'll have some pics of the old coverage, and the new application, as soon as I can find my SD card reader.

Pics!

IMG_0811.JPG


IMG_0812.JPG


Edit: Interesting. It gets to the EXACT SAME TEMP, even with the new application of TIM, across each heatpipe. Seems unlikely to me, because I didn't have much TIM on the two side heatpipes originally, but perhaps, the primary heat load is on the center heatpipe regardless?

Or did I put too much TIM on, and that's why the results are the same.

Scratch that, temps are 64.5, they are actually worse, although I think I stopped the prior run of OCCT when temps hit 63.5.

Edit: All those temps were at 230 FSB, 13.5multi, close to 2000 HT/NB, DRAM slightly slower (1333-1400), 1.4v, but with LLC on AUTO, OCCT measured vcore getting as high as 1.45.

I tried a different OC, this time with FSB 260, vcore 1.35v, which with LLC on AUTO, seems to get as high as 1.4v in OCCT.

Temps are at 65.5, which is about as high as I would ever want them to be.
 
Last edited:
I kind of wonder if metal to metal contact isn't as good as using TIM. There is quite a bit of pressure in the mounting application, it may be enough to transfer the heat just as efficiently as the same mount using TIM... Granted, the 212+ isn't the most uniform contact area, but the bearing surface is what it is.

Does thermal paste actually absorb heat or does it conduct heat?
 
fffblackmage said:
Conduct. Where would the heat go if it absorbed it all?
Click to expand...

Well, I guess 'insulate' would be better than 'absorb.' And, no, it wouldn't absorb it all... it would hold the heat in the appliance (in this case, the CPU.)

If it was a heat conductor, more would be better? Or, I guess in the case of TIM, it doesn't conduct with the efficiency of the copper heat pipes, hence a very thin layer vs gooping it on.

Hmmmm.... 😕
 
I put a very thin layer across the entire processor with an old credit card and make sure I get any gooby extras off. I put the heatsink on with no thermal paste and then rub it around a little in the thermal paste and works like a charm.
 
Charlie98 said:
Well, I guess 'insulate' would be better than 'absorb.' And, no, it wouldn't absorb it all... it would hold the heat in the appliance (in this case, the CPU.)

If it was a heat conductor, more would be better? Or, I guess in the case of TIM, it doesn't conduct with the efficiency of the copper heat pipes, hence a very thin layer vs gooping it on.

Hmmmm.... 😕
Click to expand...
Well, TIM is supposed to conduct better than air, but adding too much ends up being counterproductive. The best results is achieved with the thinnest layer, just enough to fill the gap between the heatsink and the cpu.
 
i doubt any reasonable TIM application results in a difference of more than 1 or 2 degrees. unless the temps are high i wouldn't worry about it. and no way your idle was ever at 6C unless you were outside on a cold day.


Charlie98 said:
A pull fan setup might work better if there was a shroud or something to help draw the airflow through the fins from the back side of the cooler; without anything to direct where the airflow is coming from, the fan draws air from everywhere around the fan itself.
Click to expand...
this is why my case has quite a bit of cardboard in it.

push pull, cardboard duct so that the air exits the heat sink and heads straight to the exhaust holes. no fan mounted directly on the exhaust.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top