Overclock.net › Forums › Cooling › Air Cooling › Hyper 612 PWM = VERY hot temps?
New Posts  All Forums:Forum Nav:

Hyper 612 PWM = VERY hot temps? - Page 6

post #51 of 55
Originally Posted by GeneO View Post

Here is what I found after removing mine.:


You can claim you did it properly all you want. You own supplied picture proves you did not... or your CPU/cooler base have a huge low area though the middle of them leaving a gap the TIM could not fill.

There are only 2 possibilities:
*You fubared the install.
*Cooler base and/or CPU are fubared.

My money is on you.
Several of us have tried nicely to explain the how and why. You refuse to listen.
You've made up your mind and refuse to see the facts.

I'm not wasting any more time.
Edited by doyll - 1/26/13 at 12:10am
post #52 of 55
post #53 of 55
I'm going to tag onto this thread because I'm having the same experience, sans the fan failure issues.

This is the first really hot processor I've ever used and so I decided on getting something bigger to handle the heat.

AMD Phenom II X4 975 BE on a Gigabyte GA-890FXA-UD5 in a mid tower case.

Finding a cooler that will fit and clear the memory on this board was a challenge.

This one seemed great. All the dimensions worked out well and being able to snap the fan off to install memory is nice. Unfortunately my case does NOT have a cut-out on the back side when the CPU backing plate mounts so it's all the cables off and board out each time you want to futz with this cooler.

Here's the problem though, it just doesn't cool and I do think the issue is that the heat is not making it's way into the cooler effectively.

I'm on my 4th attempt now here's how it went down:

1st It wouldn't survive the Gigabyte ET6 self overclocking attempts without getting too hot. So I found prime 95 and speed fan plus a few others to help figure out what was going on. I went back to stock settings for the cpu. Running the bench mark or torture test with only two threads it climbed pretty quickly, 2 minutes, and exceeded the 62C limit given by AMD. So I thought hey too much thermal paste.

I called Coolermaster and said I thought I had spread it thinly but your instructions weren't exactly clear so what do you recommend?
They told me to put a 2mm ball of paste in the center and let the clamping force spread it out.

2nd I got even worse temps so off it came and what I observed was a circle of paste center and spread perfectly but about the size of a dime.
Now the way this thing mounts on an AMD board you tighten nuts on the four posts on the back side of the board until they run out of thread and dead end. That's supposed to cause enough spring loaded force to make proper contact and spread the paste.

3rd I decided to give it another shot at spreading the paste much thinner than the first time but it did cover the entire surface.
Well it was a lot better now than either the 1st or 2nd times but with stock settings the prime 95 bench mark test got it into the mid 55C and a two thread torture test would run for maybe 10 minutes before it hit 62C again. A full torture test was off the scale in 2 - 3 minutes.

So I thought maybe this is an airflow issue? I have the side of my case open and the back side of the cooler (fanless side) sits about 2 inches away from the rear exit fan (noctua NF-s12B) running at full speed. I have one of those in front of the hard drives as well and that one is being speed controlled by the board.

The CPU fan does get to full speed during the stress tests and then eventually quiets back down to the same speed as my front fan. The system temp barely changes which is why the front fan never gets very high.

So I took an extra fan and held it in various places during all of this to see if that would help. It made no difference so I have to conclude that the airflow is adequate but the heat is just not making it into the heat sink all that well.

I even closed the entire case up and tested again and that made hardly any difference.

4th So I called back and got someone who really didn't want to hear what I was telling him. He insisted that the Prime95 test is way too much for any system and then asked me what my temps were at idle and when I was surfing the web. rolleyes.gif Then I told him that even with limited stressing, 50%, it was exceeding 62C and he said who told you that was limit ... Try AMD sir. So I asked him how he suggested I make certain that I'm applying the paste properly. He said a thin line across one side of the cpu and use a credit card @ 45 degrees to spread it thin enough so that you could barely read the writing on the CPU. I can barely read it with NO paste! But OK I tried this today.

It's much worse again. Now when I pulled the cooler off this morning before this latest attempt the paste was well distributed unlike the 2mm ball exercise.

I have a Noctua cooler on my HTPC. They recommend a 5mm ball of paste spread by the cooler technique but that's a very quiet rig with a lower power processor and I never stress tested that one. I don't think a Noctua cooler will fit this board and case?

I think I'm done screwing around with this cooler. I've heard enough issues with it not cooling and while I understand that making very good contact for maximum heat transfer is critical I think I've really tried and that this cooler just doesn't allow that to happen very easily. In particular I believe, with AMD mount at least, it just does not spread the paste well so unless you get it perfectly applied it's not distributing what is applied very well and thus the heat never makes it to the heat sink.

I'm not sure whether Newegg will take it back so is there another model that Coolermaster makes that is fool proof compare to this one? The V6 & V8 models look too big and I guess I'm looking for a different mounting system?

Some other anecdotal evidence:


Edited by staatsof - 2/7/13 at 5:33am
post #54 of 55
OK so I DID make one more attempt. I can post some photos of what I found is required but basically here it is.

1. When I pulled the cooler of of the Coolermaster recommended method of applying the thermal paste as of the last call to tech support I found that it had barely made contact with the cooler plate and just along the edges. The past on the cpu was for the most part undisturbed. Hmmm .

2. So I looked at the cooler base with a straight edge and discovered that the think is a bit cupped!

3. I decided to make one more attempt with this approach. Now I realize this isn't how things should be but this is how they made the darned thing.
I put a very thin layer on the cpu as before but I also put a layer, somewhat thicker, on the cooler base in an attempt to get the entire plate covered and fill in the concave void. Again, not like it should be but this is what I have.

4. I also used a set of thin washers slipped over the threaded posts that protrude through the backside mounting plate so that there's a bit more tension applied before the nuts dead-end on the threads of the posts.

Well this improved things greatly.

If I rerun the full torture test eventually it does get too hot but it takes about 35 minutes.

A two thread version seems to be stable in the low to mid 50C range but I only ran that for an hour.

THIS IS a huge improvement.

BUT ... in my investigations I ran across this:


Just look at those results and one can only conclude that this cooler is pathetic.

The question is have they fixed these issues on any of the replacement coolers?

The cooler they use in that link won't work in my case is it only blows vertically and I need to exit to the rear on my mid tower case. Damn!
post #55 of 55
Sent it back and got a Noctua NH-D14 Se2011.

Can't say I didn't try.

And I can now say the Noctua worked just fine. I was even able to eliminate my rea case fan by shifting the fans on the NH-D14.
Edited by staatsof - 2/28/13 at 1:56pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Air Cooling
Overclock.net › Forums › Cooling › Air Cooling › Hyper 612 PWM = VERY hot temps?