Did I put too much paste?

Yes, it would probably be best that the 'X' was not broken, that way the paste would just spread outwards instead of partially inwards to cover the broken areas.

As for all of heat generation being in the center, T4 is a perfect example that it just does not work for me. T4 shows probably the recommended application style which also corresponds to the picture you attached.

I think only two of my test scenarios peaked at 80C and T4 was pretty much all the time at 80C from the 17min mark. I almost expected it to go to 81C.

Interesting thing I noticed. Appears that AS5 curing is not happening for me. After 3 weeks there were no temp drops, may have even increased by 1C. (T3) Or maybe it is, since the LiquidT does increase, potentially signifying better heat transfer?

I'll see how this application fares after 3 weeks. I have also refined my testing process a bit to make it even more consistent.

My case is medium tower of a >5 years now (PSU at the top). I do have too few intake fans compared to outtake. Extra air is forced into the case through openings. I may try to add one at the top of the case blowing in.

Also, the Devil's Canyon chips have reworked heat transfer under the casing, so maybe they like it when the entire chip is covered?
 
In theory the problem is a line of paste around a void forms a barrier when cooler is set onto CPU. If the air trapped in this void does not leak out it is a space of very low heat transfer much bigger than the microscopic one in metal surfaces the TIM is being used to fill.

Another problem with these air bubbles is the 'pump' the TIM out as the expand an contract when system is started or runs under heavy load.

The thickness of the IHS means very little heat movement laterally. It's just too thin to do much.
 
Getting some more data...

My place is very dusty. I noticed that the radiator had quite a bit of dust buildup so I cleaned it.

A safe conclusion I think, is that this action has brought the temps down by around 2C.
Prior to HS dusting: 24.6C ambient lead to 74C temp with 75C peak.
After HS dusting: 26.1C ambient lead to 73C temp with 75C peak.

After running the "After HS dusting" test for 23 minutes, I continued with it but removed the side cover for the next 10 minutes to see how case airflow impacts CPU cooling.

Case open: 26.2C ambient lead to 69-70C temp with 71C peak.

I've ordered Noctua NF-S12A PWM to replace my Noctua NF-F12 industrialPPC-3000 PWM as the single front case intake fan. (Outside of testing (day to day operation) the F12 is running at 1200 RPM only. I'm not a fan of loud PCs when all I'm doing is browsing the net.) I'll put the F12 as the push fan on the radiator and will see what kind of difference extra 500RPM and maybe better fan design does.

I've also gotten a Silverstone FF143B ultra fine mesh filter to test against the currently installed quality foam filter. I'm looking to see if the mesh will maybe allow more air through while keeping dust buildup close to the foam. I do expect the foam to be better at collecting dust, but will see firsthand soon.

I looked at a larger mid tower case and Lian-Li PC-A61WX is a total beauty. I currently own the Lian-Li PC-8 (window version). Unfortunately the A61 is just slightly bigger but enough so to be just slightly too big. I could probably live with a bigger case but it wouldn't look as good in the space I have for it as the PC-8.

I would of course get the 240mm radiator for the new case which would bring the temps down almost 10C in comparison.

I'll see how the fan shuffling fares. Clearing out the dust was pretty significant change in C just for removing dust. Shortly I'll probably get the H80i v2 (newest revision) which I'm hoping (saw a single comparison) will bring the temp down another 2C.

As long as I'm not over 75C (not in the red in Asus software), ever, then I'm satisfied. For the CPU, 5C +/- probably doesn't mean much, so it's mainly for my own satisfaction of achieving a goal/desired CPU temp :) When gaming, I'm thinking the CPU will be steady at max 70-80%, which would lower the so far observed 75C to about 72C.

Getting a new H80i unit would also allow me to compare the two. See if maybe my current one is defective somewhere. Though I'm beginning to think that the temps I'm seeing is pretty much what can be expected from my cooler class.
 
NEVER ask people on a forum how to apply thermal paste. :D

It's like asking people what their favorite flavor of ice cream is.... you'll get eighteen million different answers including that one guy who swears you need to remove the IHS and shove the CPU up your ass for an hour before applying thermal paste.

I think lately the general consensus is to spread a very thin slightly opaque layer over the entire CPU die, but really If your cooler is mounted correctly just about any method is going to work equally... except for maybe not applying any paste whatsoever. :p
 
I don't remember seeing what all you system is. Could you tell us what it is? Case, vent and fan placement, all components, how fans are all controlled, etc. Without knowing what the air temp going into component is makes it near-impossible to know if case is flowing cool air to component intakes

While this guide was written with air cooling in mind, it is still applicable to your system. a basic but by no means complete guide to case cooling and how I use a temperature sensor inside of case to monitor air temps. Mostly cooler intake temps.
Case airflow and cooling is not hard to do but can be time consuming.
Setting up the case for optimum cooling is often the hardest and most time consuming part of a build... And often the most neglected.
  • There is much more to cooling than good cases and good CPU / GPU coolers. Add the fact that many GPU's make more heat than CPU means getting that heat out of the case and keeping a cool airflow to components can be a challenge.
  • There must be more air flowing through case than component fans are using. If the CPU cooler uses 70cfm and GPU uses 80cfm fans we need 180-220cfm.
  • Cases, especially those with filters, usually benefit from fans with higher static pressure ratings than stock fans... "cooler" fans instead of "case" fans.
    Intakes are typically have more restricted than exhaust because of air filters, more restrictive grills, HDD cages, etc.
  • I prefer more intake than exhaust. And don't confuse number of fans with amount of airflow... or airflow with airblow
    • airflow is flowing cool air from intake to component and flowing hot air from component out of case without the hot air mixing with the cool air.
    • airblow is lots of fans blowing air with some of hot air from components mixing with cool air making it warmer resulting in warm air not cooling components as well as the cool air will.
  • Putting fans in case as intake and/or exhaust is only the first step. These fans only move air in and out of case.
  • This does not mean heated air is not mixing with cool air.
  • Nor does it mean cool air is going to where it is needed.
  • Getting the air to flow inside of case properly is even more important. We still need to manage where the air flows inside the case. We can do this several ways; deflectors, more intake fans.. & exhaust fans, removing PCI slot covers, removing vent grills, removing HDD cage, using fans with higher pressure/airflow, building ducts to or from CPU/GPU cooler, etc.
  • Using a remote temperature sensor to monitor what air temps are is the key to finding out where the cool air is flowing and knowing heated air is not mixing into it. By monitoring this we can than make changes to get airflow the way we want it.

How to monitor air temperature different places inside of case:
  • A cheap indoor/outdoor thermometer with a piece of insulated wire and a plastic clothspin works great.
  • Made up with floral wire and tape. We don't want anything to short out with metal. ;)
    e65759c3-fbf9-414f-9f1f-f5941d540397_zps2fe90c77.jpg
  • Clip and position sensor where I want to check the temp. Make it easy to see what the air temp going into components actually is relative to room temp. ;)
  • When system is working air temps going into coolers will be 2-3c warmer than room.. up to about 5c is okay.
Click to expand...
Feel free to ask questions / add suggestions after reading.
 
Air flow is far from perfect.

In:
Noctua NF-F12 industrialPPC-3000 PWM @ 1200 RPM when idle. Bottom front. The fan is restricted on the intake and blow side. On the blow side there is a hard drive cage right next to it with two hard drives. Basically I barely feel any air reaching past the hard drive cage. There is more of it on the sides, flowing along the case walls.

Passive:
At the top there is an opening for 120mm fan but I only have a grill with foam filter there. The only way to install a fan there, due to PSU length, is if I made a 120 to 92mm adapter and placed a 92mm fan at the top. But it is tricky and I'm not thinking much about it at this time. I do feel cold air when I move my hand over the opening, on the outer side, and I believe the air is being drawn into the case, as opposed of it being pushed out.

There's also a grilled opening at bottom rear, the the right of the expansion slot covers. Cold air is being pulled in through there. I see that by dust that builds up around the opening and on the expansion slot covers.

In (continued):
At the front top of the cage I have IcyDock 5in3 cage full of hard drives. Noctua NF-A9 PWM 92mm @ 2000 RPM (max, always) cools those down. While technically it draws cold air, the air coming in after it passes the hard drives is quite warm. It doesn't get far into the case either. I barely feel it on my hand, partly because it is warm. It keeps the drives very nicely cooled, though. Ideally, this air should be immediately removed from the case via the top opening. I was thinking of putting a fan there that would remove air from the case but since it seems that the case is drawing air inside through that opening I will leave it as is for the time being.

Out:
Noctua NF-P14r PWM 140mm @ 1100 RPM (usual, the quiet setting) when idle. Occasionally I might turn it up a bit. Sits on the glass side right at the level of the GPU. I had to make an opening for the fan in the glass. I made round circle that is just the size of the fan and in the end I feel it looks better then when there was glass (acrylic) only. I used to have it blowing in, but I noticed that it's better for it to be blowing out. Blowing in was just pushing the hot GPU air into the case. Did not test how that affected the GPU, but there is a component right below the GPU that gets very hot as well - Areca RAID card. It's processor usually runs from 66-70C. It spews out more heat than the idle GPU. When the fan was blowing in Areca was hotter than when the fan is blowing out. Plus, I don't have to worry about putting on a filter on the fan and cleaning it once a week. When I do stress the GPU, the fan removes probably at least 70% of the heat that the GPU generates. The air being moved by the fan during the GPU test is very warm against the skin, might even approach hot. This configuration is extremely helpful in removing hot air from the case. Also, when I run this fan at 1250 RPM it rather quickly drops Areca's temp down by 2-3 C and at higher rotations this delta increases additional 1/2C (based on my observations).

When I turn the front fan up to 2000+ RPM it has negative effect on Areca's temp (increases it). That is (partly) the reason why I'm replacing it with NF-S12A shortly. A higher RPM has the reverse effect of what one would expect. The side fan is crucial at keeping Areca temps down and getting rid of majority of the GPU generated heat.

Out (continued):
Then there is the Corsair CPU cooler fans and radiator at the top back. Those are stock, but the push side will soon be replaced with the Noctua Industrial fan that will be taken from the front of the case.

Finally there's the PSU drawing air out of the case around the area of the CPU heatsink. This also removes some of the hot air created by the VRM components around the CPU socket.

The case is severely lacking intakes of cold air but I can't really do anything about it. In my setup I believe the side fan is key in removing heat from the biggest producers of it in that region (GPU and Areca).

Fan speed control. All are PWM and all are connected to MB headers controlled through Asus AI software. Very satisfied with it. I have all my fans set to Fixed RPM. The CPU cooling fans are connected to the pump and are managed through the Corsair software, those are dynamic based on CPU temp.
 
Last edited:
Can you give us a simple list of what your system is? With no list of components and not knowing what case you have it is hard to help.

'Passive' means nothing in case airflow .. unless it is in a system with no fans .. none at all. Even an idling fan's airflow will overpower any passive flow in a system.

Intake and exhaust flow are equal. If they were not the case4 would blow up like a balloon or get sucked flat like a crushed beer can. :D I know that is extreme example, but a case only flows the maximum of the smaller of the two. I know, that sounds confusing. What comes in must go out / what goes out must come in.

Asus has lied about having PWM headers when they were not. They claim 4-pin fan headers are PWM when their is not PWM signal on pin-4 and the only speed control is on pin-2, the power pin. Real PWM has constant 12v power on pin-2 and requires a PWM signal from pin-4 to fan for fan's internal PWM PCB to pulse the 12v power to the fan motor. Sigh... I know, even more confusing stuff. :p
 
You must log in or register to reply here.
Back
Top