Jump to content
Corsair Community

H80i V2 Fans Won't Ramp Up.


Borna
 Share

Recommended Posts

Don't use package temp. That's not how this cooler or any other works. Using package temp will only put your fans into a constant state of flux as they try to play follow the leader with the ever changing CPU temperature. The default is coolant temperature and that will give you the most efficient cooling for normal use. If you are going to stress test for whatever reason, the better thing to do is set a fixed of RPM of your choosing, whether that is minimum, maximum, or somewhere in between. That will ensure your results are comparable to prior/future runs.
Link to comment
Share on other sites

Don't use package temp. That's not how this cooler or any other works. Using package temp will only put your fans into a constant state of flux as they try to play follow the leader with the ever changing CPU temperature. The default is coolant temperature and that will give you the most efficient cooling for normal use. If you are going to stress test for whatever reason, the better thing to do is set a fixed of RPM of your choosing, whether that is minimum, maximum, or somewhere in between. That will ensure your results are comparable to prior/future runs.

 

If I set It to coolant temp the fans are still quiet and the coolant temperature Is still stuck at 30C under load Is that Intended?

 

Do the fans ever reach 100% under load with the default profiles? since my cpu temp hits 70C and I still can't hear the fans.

Edited by Borna
Link to comment
Share on other sites

Yes, that is intended. Your CPU is cooled by conducting heat away from the CPU lid and the parts there in. This is the same no matter if you have a simple air cooler or a 10 ft tall standing external water cooling apparatus. What the cooler does with the heat after that is what distinguishes various cooling types and sizes. Each +1C of coolant temperature will add +1C to the CPU temperature. When your coolant only goes up 3-4C, that means the cooler is removing heat from the system almost as fast as it comes in and the most you can reduce the CPU temp by is 3-4C at any fan speed. However, there is nothing you can do to increase the conductivity or heat transfer at the CPU pins where the voltage creates the heat. The heat must pass through the CPU. On a any modern CPU, you will be limited by the voltage and the CPU design. If you put a small enough cooler on, you can make the margins even thinner. A larger cooler gives you more leeway, but the differences are always going to be small in comparison to the voltage based heat on the CPU.

 

When you initialized the stress test, there was an instant jump in CPU temp of around 35-50C (depending on your voltage and settings). That is the voltage based heat and you can never eliminate that without lowering voltage or changing the physical properties of the CPU. It's what happens after that that is under your control. You can control the rate of heat evacuation (radiator fan speed) and to some degree to can control the environment, but ultimately that initial temp spike is the largest factor. Even if I give you 10m long radiator panel with 100 fans, all it can do is get rid of the heat in the coolant stream. It can't make the CPU cooler, only prevent it from getting hotter than it already is. This is also why extreme overclockers result to liquid nitrogen. They cannot make the CPU conduct heat away any faster either, but they can lower the initial temperature of the CPU to something extremely low so it can survive that 100C+ jolt that occurs when fed 2 volts of Vcore.

 

Your specifications aren't listed, but for a typical 4-6 core CPU you are going to see a +6C rise or so in coolant temperature on 100% stress test. It usually takes the coolant 4-10 minutes to reach that mark and then it will level out (assuming the CPU test is fixed and not variable). That is your balance of watts in versus watts dissipated. After that, any rise in CPU/coolant temp is most likely the result of an increase in case temp. A +1C rise in case temp will also raise the coolant +1C and thus the CPU +1C as well.

Edited by c-attack
Link to comment
Share on other sites

Yes, that is intended. Your CPU is cooled by conducting heat away from the CPU lid and the parts there in. This is the same no matter if you have a simple air cooler or a 10 ft tall standing external water cooling apparatus. What the cooler does with the heat after that is what distinguishes various cooling types and sizes. Each +1C of coolant temperature will add +1C to the CPU temperature. When your coolant only goes up 3-4C, that means the cooler is removing heat from the system almost as fast as it comes in and the most you can reduce the CPU temp by is 3-4C at any fan speed. However, there is nothing you can do to increase the conductivity or heat transfer at the CPU pins where the voltage creates the heat. The heat must pass through the CPU. On a any modern CPU, you will be limited by the voltage and the CPU design. If you put a small enough cooler on, you can make the margins even thinner. A larger coolers gives you more leeway, but the differences are always going to be small in comparison to the voltage based heat on the CPU.

 

When you initialized the stress test, there was an instant jump in CPU temp of around 35-50C (depending on your voltage and settings). That is the voltage based heat and you can never eliminate that without lowering voltage or changing the physical properties of the CPU. It's what happens after that that is under your control. You can control the rate of heat evacuation (radiator fan speed) and to some degree to can control the environment, but ultimately that initial temp spike is the largest factor. Even if I give you 10m long radiator panel with 100 fans, all it can do is get rid of the heat in the coolant stream. It can't make the CPU cooler, only prevent it from getting hotter than it already is. This is also why extreme overclockers result to liquid nitrogen. They cannot make the CPU conduct heat away any faster either, but they can lower the initial temperature of the CPU to something extremely low so it can survive that 100C+ jolt that occurs when fed 2 volts of Vcore.

 

Your specifications aren't listed, but for a typical 4-6 core CPU you are going to see a +6C rise or so in coolant temperature on 100% stress test. It usually takes the coolant 4-10 minutes to reach that mark and then it will level out (assuming the CPU test is fixed and not variable). That is your balance of watts in versus watts dissipated. After that, any rise in CPU/coolant temp is most likely the result of an increase in case temp. A +1C rise in case temp will also raise the coolant +1C and thus the CPU +1C as well.

 

That makes sense thanks for the Info.

Link to comment
Share on other sites

 Share

×
×
  • Create New...