Jump to content
Corsair Community

H80i High Idle Temps. (40C Idle)


Archdregs
 Share

Recommended Posts

Hello.

 

I have H80i with very high idle temps. By default (on 23C ambient) it shows 50C. After I change the thermal paste from stock Corsair to Noctua NT-H1 it dropped to 40C. What is wrong with my cooler? This is far worse than average air cooler on idle. Even increasing the fans to 100% doesn't even help a bit.

 

Please help. Thanks.

Edited by Archdregs
Link to comment
Share on other sites

Can you confirm the H80i Temp (coolant temperature) is in this 40-50C range at idle? The coolant temperature is very useful for separating cooler issues from all the other kinds. Normally in a relaxed idle state, it sits between 4-7C above your room temp (assuming you haven't just finished heating up the case doing something else).

 

Also, can you confirm you have this H80i that is SATA powered and not the GTX/v2 versions that are motherboard header reliant?

Link to comment
Share on other sites

You do have the SATA powered H80i. Does not matter for what we are talking about, except the GTX/v2 versions can easily be waylaid by cutting the voltage to the unit in the BIOS. Easy to forget when you update to set it back to 100%. Your pump speed looks normal so you do seem to be getting power.

 

 

33.4C coolant temp is not terrible for a H80i in Summertime. It also seems to be corroborated by motherboard and GPU temps. Best guess the case interior is about 33C and thus that is as low as you can go.

 

I am going to assume the H80i is in the rear exhaust slot. Not sure where else you would really put it in a 600T. Do you have it set to intake or exhaust?

 

Can you run a short (10-15 min only) CPU only stress test? Doesn't really matter which on 3770. I would probably skip Prime95 unless you regularly use it. I prefer mild tests when investigating cooler issues in case there is a real problem. Prime can be hot for any number of reasons. When something mild like Intel XTU or AIDA64 gets hot, you know there is a problem. Run the stress test for 10-15. Note the starting H80i temp and then the final temp. After you stop the test, what happens with the coolant temp? It should start dropping at about 1C per minute. The last 2C to get back to 33C will be slow and can take 20 min. I would expect a coolant change of +6C, but there might be some variance depending on your voltage/BIOS settings.

 

Before after Link shots would be great if you can. If you use AIDA, you can save the entire test line graph after it has finished.

Link to comment
Share on other sites

OK, always Summer. What happened to the H80i Temp during the tests? There are a dozen reasons why CPU temps can be high. If the coolant temp is too high in a CPU test, there is usually only one reason.

 

Also, don't use small FFT right now. It's too aggressive, particularly if this is a contact issue or there is a cooler problem.

Link to comment
Share on other sites

I think +10C in 5 min is a bit much, but not rock solid conclusive. Do two things:

 

1) Let it cool down for 15 minutes. Watch it the first couple of minutes. Is the H80i temp coming down (about 1C/min)? Or is it stuck at 44C?

 

2) If it does cool down, run the same 5 min blend test but this time with the fans set at a fixed speed. Something fast like 1900-2000. That will be annoying, but I want to see if there is a difference. That will be telling in itself and also nullifies any future complaint "your fan speed is too low".

Link to comment
Share on other sites

OK, so temps return to baseline quickly and fan speed does have an effect on the coolant temperature. That suggests the cooler is working normally. The coolant delta is a bit higher than I was expecting. Is there anything preventing the H80i was getting rid of its heat like an dust filter or pinned against the wall?

 

Idle temps: There really should be no difference at idle between a small air cooler, a 280mm AIO, and triple radiator custom loop. Your idle temperatures are almost exclusively determined by your room/case temperature and your basic voltage and power settings. Obviously if it is 20C in your room and the CPU idles at 30, then when its 30C in your room it will idle at 40C. Changes in your internal power saving features like EIST, Speed Step, Adaptive vs Manual voltage all affect the baseline. Assuming no changes there, your OS level Windows Power Plan can alter the minimum processor state as well. The default balanced will let the cores relax if the BIOS will as well.

 

If all those things check out, you should look for signs of abnormal activity. A background process that is using 10% when it should be less than 1%, background OS operations can temporarily alter the load without being obvious.

 

The last point is cold plate to CPU contact. Normally, this would be higher on the list, but since you already re-seated the pump with success, doing it again is not likely to help. Obviously if you don't have good contact from the start, it trumps all of the above and heat is not effectively transferred into the cooler. Given the clear rise in coolant temperature for your tests, this seems unlikely to be a current issue.

Link to comment
Share on other sites

1) If you pour water through it, you obviously have to remove it from the case, carefully rinse, and then wait for it to completely dry before putting in back in. One stupid drop that was hiding in the between two fins can do damage to something else. I would probably try to avoid that unless really necessary. While on a technical level, a faint coating of dust will inhibit heat transfer more than a pristine piece of metal, you are not likely to see a temp difference between "I think I see dust" and "water washed clean". If you have piles of dust building up, that is another matter.

 

If compressed air is available, that can work. My preference would be for a the hose end of a vacuum cleaner (if you have that type). The nozzle end can pull it off efficiently enough without having to rip it all apart. Be careful with static discharge or touching the nozzle to other components.

 

2) Any time you remove the block, you most likely should re-paste. You may made the TIM fit and create a seal between the two. When you pull it off, you alter where the pieces were and it is never quite the same. Can you get away with not doing it? Sure. It may have no temperature difference. If you put it back on all of sudden things are 10C hotter, you know what you need to do. The other risk is contamination. Dust or hair getting stuck in the TIM will not be helpful.

  • Confused 1
Link to comment
Share on other sites

 Share

×
×
  • Create New...