Jump to content
Corsair Community

H150i not so great results


cristofer50
 Share

Recommended Posts

Can you show us the performance tab for the H150i?

 

Coolant Temperature (H150i Temp) is the best measurement of radiator/cooler functionality. You should idle about 4-7C above room temp and a 9700K at 5GHz should be a max load increase of +6-7C above that. Unfortunately, everything else is connected to things other than the cooler.

 

Your voltage is reasonable and I am not really surprised to see a 41C Package idle temp if your coolant is 28-30C. The difference between max CPU temp and coolant temp can help identify where the problem lies. I would expect +45C for that voltage and CPU, but there is variability from silicon to silicon.

 

I'll wait to address the fan thing until we see the Performance tab, but make sure all fans are "ringed" yellow with the fan curve you want to use. You have to tag all three.

Link to comment
Share on other sites

Can you show us the performance tab for the H150i?

 

Coolant Temperature (H150i Temp) is the best measurement of radiator/cooler functionality. You should idle about 4-7C above room temp and a 9700K at 5GHz should be a max load increase of +6-7C above that. Unfortunately, everything else is connected to things other than the cooler.

 

Your voltage is reasonable and I am not really surprised to see a 41C Package idle temp if your coolant is 28-30C. The difference between max CPU temp and coolant temp can help identify where the problem lies. I would expect +45C for that voltage and CPU, but there is variability from silicon to silicon.

 

I'll wait to address the fan thing until we see the Performance tab, but make sure all fans are "ringed" yellow with the fan curve you want to use. You have to tag all three.

 

I have it selected OK.

150i.thumb.jpg.ba6cacdbf9e6670d57d49b4462206b88.jpg

Link to comment
Share on other sites

Was that picture taken at idle or load? 37C is pretty warm, unless you room is near 30C.

 

In the Performance Tab, click + to create a new "cooling mode" (you can rename it). This is your custom curve. A graph will appear at the bottom. In the top right corner are three shape tools. Click the one on the very right. This is a copy of the "Extreme" profile you are currently running. The difference is you can now see the graph points and know what speed should equal what temperature. Leave the sensor control variable on H150i Temp. You can also tweak it to tame things at idle. Fan speed is not overly sensitive on a 360mm radiator. You don't go up 5C because you drop the fan by 200 rpm. That is also why I am a bit concerned about the current levels.

 

Can you tell us anything about the case you are using, where the cooling is located (front intake, top exhaust, etc)? Aside from room temperature, that is the other looming factor. Radiators don't breathe very well through glass, heavy dust filters, or anything else that add significantly the airflow restriction the radiator already creates.

Link to comment
Share on other sites

Was that picture taken at idle or load? 37C is pretty warm, unless you room is near 30C.

 

In the Performance Tab, click + to create a new "cooling mode" (you can rename it). This is your custom curve. A graph will appear at the bottom. In the top right corner are three shape tools. Click the one on the very right. This is a copy of the "Extreme" profile you are currently running. The difference is you can now see the graph points and know what speed should equal what temperature. Leave the sensor control variable on H150i Temp. You can also tweak it to tame things at idle. Fan speed is not overly sensitive on a 360mm radiator. You don't go up 5C because you drop the fan by 200 rpm. That is also why I am a bit concerned about the current levels.

 

Can you tell us anything about the case you are using, where the cooling is located (front intake, top exhaust, etc)? Aside from room temperature, that is the other looming factor. Radiators don't breathe very well through glass, heavy dust filters, or anything else that add significantly the airflow restriction the radiator already creates.

 

It's at 34.10 C at the moment in iCUE no load. It's a 570X Corsair case and it's all pull 3+3 fans. It's at the front too. Core Temp was the software I was using to read the temps and it says 41 right now. Why are they different?

Link to comment
Share on other sites

[CUT]

Coolant Temperature (H150i Temp) is the best measurement of radiator/cooler functionality. You should idle about 4-7C above room temp and a 9700K at 5GHz should be a max load increase of +6-7C above that. Unfortunately, everything else is connected to things other than the cooler.

 

[CUT]

 

 

I’ve read this quite often and always wanted to ask... +4-7C above room temperature for the coolant as “normal” is intended on which pump speed and which fans profile?

Just to compare to my data.

 

Baio

Link to comment
Share on other sites

I’ve read this quite often and always wanted to ask... +4-7C above room temperature for the coolant as “normal” is intended on which pump speed and which fans profile?

Just to compare to my data.

 

Baio

 

It is a very general estimate based on multiple cooler types and a lot of different cases over the years. I always assume the user is employing a quiet fan speed at idle, something like 750 rpm for 120s or 600 rpm for 140s. Regardless, heat dissipation is not normally an issue at idle, so if you max out your fans, that might slide you down 2C on the scale or if you set them to super low 350-400 RPM it might add 2. Of course, a 10980X has a lot more idle heat than 6700K. The greater factor is usually internal ambient temp and radiator location. Top exhaust is going to be on the higher end of the scale because it is warmer in that top layer. I could keep it at 2C in my 740 with wall to wall fans and a lot of mesh. Not such luck in a glass case. There are a lot of factors, but if you don't feel that your in the top category in several of those factors and still are running +10C over ambient, then there may be a some type of set-up issue or an actual physical problem.

Link to comment
Share on other sites

It's at 34.10 C at the moment in iCUE no load. It's a 570X Corsair case and it's all pull 3+3 fans. It's at the front too. Core Temp was the software I was using to read the temps and it says 41 right now. Why are they different?

 

Coolant temperature will only match CPU temp when the voltage is zero. Unless you are in an S3 sleep state or off, that doesn't happen much and the current group of CPUs all modulate their own P-states, but underneath the monitoring level. Your CPU idle temp is normal for that coolant temp. It is the coolant temp that appears to be higher than expected.

 

So the H150i is front intake. Do you still have the dust filter on? If so, try removing it and see how much difference this makes. Also, you haven't mentioned the room conditions. If it is 28-29C in your room, then 34C really isn't too far off base. The problem I am having is you are running the fans and pump at near maximum and it is still at 34. Any kind of dust filter or restriction should not be much of a factor at those fan speeds.

Link to comment
Share on other sites

Coolant temperature will only match CPU temp when the voltage is zero. Unless you are in an S3 sleep state or off, that doesn't happen much and the current group of CPUs all modulate their own P-states, but underneath the monitoring level. Your CPU idle temp is normal for that coolant temp. It is the coolant temp that appears to be higher than expected.

 

So the H150i is front intake. Do you still have the dust filter on? If so, try removing it and see how much difference this makes. Also, you haven't mentioned the room conditions. If it is 28-29C in your room, then 34C really isn't too far off base. The problem I am having is you are running the fans and pump at near maximum and it is still at 34. Any kind of dust filter or restriction should not be much of a factor at those fan speeds.

 

I messed with iCUE and was able to get the temperature lower to 90 C in RealBench @ 5 GHZ.

 

Can someone please tell me why one of the fans sticks in the 1440's RPM when i have it set to max flow? Am I missing something? The other 2 are hovering around 1500.

Link to comment
Share on other sites

I messed with iCUE and was able to get the temperature lower to 90 C in RealBench @ 5 GHZ.

 

Can someone please tell me why one of the fans sticks in the 1440's RPM when i have it set to max flow? Am I missing something? The other 2 are hovering around 1500.

 

Typically a fan has a +-10% variance at max speed. For the ML120 "quiet", that would be 160rpm. With three independent fan channels, there are not likely to be in perfectly in step. However, this is irrelevant for cooling and +-100 rpm makes no difference.

 

Are you using a custom fan curve set to 100% PWM? Or something else?

Link to comment
Share on other sites

Typically a fan has a +-10% variance at max speed. For the ML120 "quiet", that would be 160rpm. With three independent fan channels, there are not likely to be in perfectly in step. However, this is irrelevant for cooling and +-100 rpm makes no difference.

 

Are you using a custom fan curve set to 100% PWM? Or something else?

 

Thanks, c-attack. I have it set @ 100%. I tried fixed RPM also but no dice.

Link to comment
Share on other sites

You can swap the "low" fan with one of the others on the H150i splitter and see if the lower speed travels with the fan (fan characteristic) or it stays on the same "fan number" in iCUE. That could be hardware or software.[/quote

 

Sounds like a good idea. I might be awhile but I'll report back when I can. I am waiting for a response from Corsair too about the problem. Thanks again! :D

Edited by cristofer50
Link to comment
Share on other sites

It is a very general estimate based on multiple cooler types and a lot of different cases over the years. I always assume the user is employing a quiet fan speed at idle, something like 750 rpm for 120s or 600 rpm for 140s. Regardless, heat dissipation is not normally an issue at idle, so if you max out your fans, that might slide you down 2C on the scale or if you set them to super low 350-400 RPM it might add 2. Of course, a 10980X has a lot more idle heat than 6700K. The greater factor is usually internal ambient temp and radiator location. Top exhaust is going to be on the higher end of the scale because it is warmer in that top layer. I could keep it at 2C in my 740 with wall to wall fans and a lot of mesh. Not such luck in a glass case. There are a lot of factors, but if you don't feel that your in the top category in several of those factors and still are running +10C over ambient, then there may be a some type of set-up issue or an actual physical problem.

 

Thanks c-attack.

I'm using a very low rpm profile because in this moment I prefer silence over performance. Pump speed is on silent, AIO fans are on a custom curve that keeps them at 30% untill coolant temperature is under 35C. All other fans (7) are set to 300 RPM, AIO is on top of the case in push/pull configuration.

 

With 20.1C in my room the coolant temperature starts from 26C and sits on 34C after half an hour of browsing, so I think my temperatures are a little higher than they could be, but the all system is working fine.

 

Baio

Link to comment
Share on other sites

Thanks c-attack.

I'm using a very low rpm profile because in this moment I prefer silence over performance. Pump speed is on silent, AIO fans are on a custom curve that keeps them at 30% untill coolant temperature is under 35C. All other fans (7) are set to 300 RPM, AIO is on top of the case in push/pull configuration.

 

With 20.1C in my room the coolant temperature starts from 26C and sits on 34C after half an hour of browsing, so I think my temperatures are a little higher than they could be, but the all system is working fine.

 

Baio

 

With the 9900k increase pump speed. I have same cpu and cooler and on silent temps are much higher. It doesn't push liquid fast enough on low settings. I personally cant hear pump on extreme, so I leave it there.

Link to comment
Share on other sites

You will pay a definite coolant temp penalty for using the ultra low 1100 rpm Quiet speed on the pump. It makes no meaningful difference for standard desktop work, but I would bump up to balanced for extended load - work or play.
Link to comment
Share on other sites

To give you a sense of this, a few months back my cpu was crashing on a normally stable overclock. Turned out one of the icue updates had reser my profile and pump was on quiet. Turning it to extreme dropped coolant temperatures by almost 5c instantly. With the heat the 9900k produces, the low flow rate of quiet pump operation means liquid sits there longer than im.comfortsble with. You could try a custom pump profile too.
Link to comment
Share on other sites

Afraid no custom pump speeds on Corsair AIO coolers. The Quiet 1100 speed only exists on the Pro series and was made for desktop work. Most other coolers will have minimums around 1800-2000 rpm. You might also compare the Balanced to Performance for your use. Most other 9900K owners report a 0-1C coolant temp difference between them. You can be too slow, but you don't get a continual cooling benefit from ramping up the flow rate.
Link to comment
Share on other sites

With the 9900k increase pump speed. I have same cpu and cooler and on silent temps are much higher. It doesn't push liquid fast enough on low settings. I personally cant hear pump on extreme, so I leave it there.

 

I have a 3700X, but I’ll give it a try!

 

Baio

Link to comment
Share on other sites

You will pay a definite coolant temp penalty for using the ultra low 1100 rpm Quiet speed on the pump. It makes no meaningful difference for standard desktop work, but I would bump up to balanced for extended load - work or play.

 

Thanks c-attack!

 

Baio

Link to comment
Share on other sites

 Share

×
×
  • Create New...