H80i V2 Temperature Issues

[Deptacon]

Hello,

 

New to water cooling and the Corsair H80i V2 I installed yesterday is my first go. Just curious if any of this is normal or should I check seating, etc???

 

Fractal Design R5 Case

I5 6600K 3.5GHZ OC'd to 4.28

 

H80i mounted on the back, fans set to intake air (both). Also, on the raditor, Hoses are on TOP, not bottom. Had to do this to make it fit (curious if it really matters)

 

Idle temp (several small programs running 35% CPU Usage in Task Manager) at about 42 C (CPU Package Temp in CL, 37 C in Real Temp). Cpu Package temp bounces between 40 and 48 and in between alot at this standard idle. H80Vi2 Temp showing 38C at the same time. Fans and Pump Max (2700 and 2800 rpm)

 

Max Temp During Prime95 30 mins was 75 C on CL (RealTemp shows 2 cores at 77 and 80). H80Vi2 Temp never exceeded 41C. (2700 and 2800 rpm)

 

 

 

Seating issue? Switch radiator to pump hoses on bottom? Or am I parnoid and good?

 

Thing is i am still hitting high 60's maybe 70 when playing Battlefield 1 for an extended time..... SAME temp I would reach with a standard Hyper Evo 212 heatsink with SAME overclock...... but idle does seem to be slightly cooler.

[c-attack]

You don't have a contact issue or the Prime test would have been untenable within seconds. You ran for 30 minutes which is enough to be sure it functions. Predicting idle temps is somewhat complex. Your room and case temperature play a part and so does your C-state, EIST, and voltage settings. It's hard to give an exact range, but if you are using 35% of your CPU resources, you aren't at idle anyway.

 

Now that you know it works, don't waste a lot of time trying to set it up for synthetic stress tests and the like. Collect your data from the PC's actual use. The game data is fine, but you can't just take a min/max to make comparisons. Almost all games are going to have to same max CPU temp regardless of the cooler and it likely comes at specific point or stage in the program. It has nothing to with sustained thermal load and is more likely related to a specific instruction that causes a large spike. Most games I have reach their peak temperature the moment I launch the application. The actual use temps tend to peak 10C below that. Line graphs make when and where the spike occurs very apparent. Absent that, use a monitoring program that also keeps "average CPU temp". Clear it before you start the program. That is a much better way to gauge the amount heat in the system, compared to a single data point peak.

 

As for the physical set-up, the hoses on top should not matter. Paranoid people live in fear a giant air bubble will crush their cooler. No. As for the fan orientation, a lot of people get lead astray by the recommendation in the instructions about using "cooler external air". That is not a firm rule and disadvantageous for many. The key player in that decision is your GPU. When you turn the H80i fans to exhaust, all the CPU waste heat is dumped out of the case. The intake air will be warmer when gaming because of the GPU waste heat below. Open air designs put more heat in the case. Reference blowers put more out the back, but all make a difference. What really matters is the coolant temperature inside the H80i (v2?) or H80i Temp in Link. No matter which way you point the fans, it sits right above the GPU and the radiator will absorb some of the radiant heat. Your coolant temperature will be warmer than someone running a single or dual cooler in the roof, farther away from the GPU. So which is better? My preference is for exhaust. Dumping extra heat directly out of the case is always advantageous. The coolant temp is affected in both positions, but when it is intake your motherboard, RAM, and all other internal components share in the warmth. You also need to take case position and location into account. If you push the back of the case near a wall, that is not going to be cooler external air at all and you take a double penalty. In your data shot, your motherboard temp is 41C. Not dangerous, but that effectively is the internal temperature of your case. That is something to use as a measuring stick. The best thing to do is turn the fans around and find out. I don't think it will make much difference in CPU temps. I do think your internal temps will go down. Either way, it is an easy experiment. You can run a stress test if you like, but in reality your actual use patterns should determine the decision and be the test set.

[Deptacon]

Ok, Thanks.

 

Here is what I did:

 

Left the Radiator mounted on the back, but switched the fans to EXHAUST instead of intake like they were before.

 

Also, while you said I probably dont have a contact issue, I reapplied it with new paste anyway. ALSO, I DID HAVE the standoff gap/looseness/back plate looseness issue others have reported. I went to Lowes and got 4 rubber grommets and added those between the back plate and the back of the board. Standoffs were snug now so I assume when I tightened everything back down it should, in theory, have given me a better seating. I also applied MX-4 since I had to clean off the stock paste.

 

Im going to run some tests, and will post the comparison. Idle load temp (well, my idle, just browser and some monitoring software running) Appears to be similar but just slightly cooler than before. Before it hovered around 41-47 now its hovering 38-42.

 

The pump shows a lower RPM now, 1950 instead of 2850 like before.... but, that previous screenshot was POST load test not before and the water temp is lower since I just booted the machine.

 

Again, will run another load test, then idle test afterwards (like I did last time) to see what I get.

[c-attack]

Pump runs on a low/high switch called Quiet/Performance. It can be set just like a fan. The 1850 is the low and ~2800 the high. It doesn't make a huge difference and I tend leave it stepped down when working at the desktop and turn it up when gaming, etc. when the noise won't be heard anyway.

 

The backplate is supposed to be loose by design. The washers are technically not necessary, but in most cases they are harmless. Don't get too carried away looking for maximum pressure. That will not help and can have long term repercussions for your motherboard. When you don't have good contact, you will see large jumps in CPU temperature and very erratic general patterns. Unfortunately, that is what it sometimes looks like when you launch any application on Windows 10, so don't go crazy anytime you see a temp spike.

 

I think the real measure will be your overall system temps when gaming for 2 hours plus. This is when you get he most built up heat from all components. Take note of CPU, GPU, motherboard, Drive temps, H80i Temp., or anything else you have to indicate internal temperature. You can use the rear as intake, but it does require a ton of top fan speed since that is now your only exhaust point. For cases on the floor below you, that can be irritating. Also, most people can give up a few degrees in CPU temperature in exchange for lowering general case temps. Since the lowest possible coolant temperature is also the lowest case temperature, not dumping more heat into the case has multiple benefits. Also remember the coolant temperature is affected by room temperature too. You are always going to have a lower reading in the morning than in a sun splattered room late in the afternoon.