H115i PRO Sensors

Hello there!

I recently read a topic from DevBiker and got a bit confused. So here’s my question which sensor(s) shall I use for more accurate cooling? I’m using a H115i Pro RGB AIO and I usually choose “CPU package” as the temp sensor but maybe that’s not the best sensor to pick?

 

Also an off topic question:

Shall I use iCUE only or do I also need Corsair LINK to be installed as well?

My Corsair hardware is the following:

Vengeance LED Ram

H115i Pro RGB AIO cooler

ML RGB Fans with Lighting node PRO

Wow, this is the question of the day. No, absolutely do not use the package temp. The 8700K is way too dynamic and more importantly, this isn't how the cooling process works. Your CPU temp is the result of the electricity in the socket and the CPU load and materials, less the amount of heat conducted away through the cold plate. The fans do not directly factor into this at all. The radiator and liquid loop is really the waste disposal part of the cooling system. Pick up waste heat at the cold plate, dump it at the radiator. Now the fans come into use. They aid in the rate of heat dissipation from the radiator. Obviously this only effects the coolant temperature and not the CPU directly. You put 1.60 Vcore on your CPU and no cooler can save you, no matter how many fans or how large the radiator. All of heat must be conducted through the CPU to get rid of it. You can't bypass and that is how the coolant temp fits back into CPU temperature. The conductor pathway between coolant stream and CPU across the cold plate is a two way street. Heat goes both ways, so while it will conduct out, it conducts back to the CPU. Your coolant temperature (H115i PRO Temp) is effectively the lowest possible CPU temperature at idle (or theoretically with zero voltage). Coolant temperature goes up +5C. That makes the baseline CPU temp +5C and everything shifts upward by that value. It also means fan speed does not have a huge effect on end CPU temperature wit the larger 280mm and 360mm coolers. If the coolant change is only 4C, then that is the most you can reduce CPU temp at any fan speed. I only need 600-800 RPM for my 8700K at 5.0 GHz and I could certainly get away with less.

Wow, this is the question of the day. No, absolutely do not use the package temp. The 8700K is way too dynamic and more importantly, this isn't how the cooling process works. Your CPU temp is the result of the electricity in the socket and the CPU load and materials, less the amount of heat conducted away through the cold plate. The fans do not directly factor into this at all. The radiator and liquid loop is really the waste disposal part of the cooling system. Pick up waste heat at the cold plate, dump it at the radiator. Now the fans come into use. They aid in the rate of heat dissipation from the radiator. Obviously this only effects the coolant temperature and not the CPU directly. You put 1.60 Vcore on your CPU and no cooler can save you, no matter how many fans or how large the radiator. All of heat must be conducted through the CPU to get rid of it. You can't bypass and that is how the coolant temp fits back into CPU temperature. The conductor pathway between coolant stream and CPU across the cold plate is a two way street. Heat goes both ways, so while it will conduct out, it conducts back to the CPU. Your coolant temperature (H115i PRO Temp) is effectively the lowest possible CPU temperature at idle (or theoretically with zero voltage). Coolant temperature goes up +5C. That makes the baseline CPU temp +5C and everything shifts upward by that value. It also means fan speed does not have a huge effect on end CPU temperature wit the larger 280mm and 360mm coolers. If the coolant change is only 4C, then that is the most you can reduce CPU temp at any fan speed. I only need 600-800 RPM for my 8700K at 5.0 GHz and I could certainly get away with less.

 

So correct me if I’m wrong but if I got that right, the optimal way for the best cooling possible is going with the PUMP’s sensor? And if yes, which is the recommended temp limits (min-max) for the custom fan curve? I always picked the package cause according to the temp sensors was the one running hotter but probably I was wrong. Thanks for your enlightening answer by the way.

 

For now I run the 8700K in stock settings btw.

YES, use the Pump's sensor. This is the coolant temperature. The fans are cooling the coolant, not the CPU. (The coolant cools the CPU.)

 

You don't need Link. You can control everything with iCue.

 

Finally, the temp range ... it's hard to say and it may vary by season (depending upon your ambient). Typically, you'll want to look at somewhere between 7-10C above ambient as your MAX temperature. The cooler itself will certainly handle warmer ... I'd say that 50C is an absolute upper, panic mode limit.

 

Again, this is just a generalization as your overall airflow and, in particular, your GPU can have a pretty significant impact on this. I would start with one of the built-in curves (like Balanced) and see how that works for you. From there, you can create a custom fan curve that tweaks the "Balanced" curve more to your liking and environmental conditions.

I live in Greece so in the winter room temperature sits about 20-22 celcius, while during the summer can get up to 30-32 room temp always. So in order to have a starting point shall I boot up my PC, leave it for about 20-30 minutes in idle to check the temp of the PUMP which I will be running on Extreme mode by the way and then try to balance the fans speed during a stress test/gaming environment? I would like my CPU temp to reach about 50 celcius during gameplay

I think with a pump at Balanced and the fans at Quiet (fixed) you'll have enough cooling performance for gaming.

 

Here for the most part I don't need anything higher than that even on a hot summer day (again, only gaming) and the PC stays ridiculously quiet.

 

You'll need the pump at extreme and fans at balanced or extreme only if you plan to stress your CPU to temps above 70c (rendering video, stress test)

 

Yes, is better t choose coolant as your sensor. CPU could be a solution too but iCUE's 60c upper limit does not make it practical for most scenarios.

I used to set max fan speed on 51 celcius while using the CPU sensor. But I get it, I will be using the PUMP's sensor from now on, I only need to find the "sweet spots" for the fans curve and I guess I'm ready to go. I'll do some expiremental fan curves I guess

 

I wouldn't say that I notice the noise from the PUMP while running at 2800 RPMs, at least not until I come really close to my case.

 

Anyway, no I don't do rendering, I will do a stress test though to check the cooling performance and that's it. If I would render or even stream I'd go with a R7 2700X instead of 8700K

I wouldn't say that I notice the noise from the PUMP while running at 2800 RPMs, at least not until I come really close to my case.

 

True, this pump is very quiet. (kudos to Asetek for that)

 

I do hear it at extreme speed, but is barely noticeable. I run a sound studio, that's why I want no noise at all.

 

At balanced speed is just 500RPM less than extreme speed and the (small) difference will only be noticeable as I said during stress tests.

which is the recommended temp limits (min-max) for the custom fan curve? I always picked the package cause according to the temp sensors was the one running hotter but probably I was wrong.

 

 

Maximum safe temperature for coolant is somewhere in the 50-60C range. You should never get there. That is not a fail point for the cooler, but if your baseline temp is 55C (coolant temp is minimum CPU temp), then you will hit 95C the moment you open a browser. And there is the problem with using package temp or any other CPU based temp as a control variable. On recent CPUs (especially Kaby and Coffee), the behavior has been designed to provide very responsive light duty work. When you open a browser, launch a game, play music with iTunes, a lot of CPU resources are used to make this open and respond quickly. My 8700K will use one core at maximum to play music, strolling along at 55C, whereas my 5930K would sit there is casual idle state with clocks relaxed. You don't want your fans to perk up when open a folder and it doesn't do anything for the CPU anyway. The heat is created at the pins and cores. Blowing air out the top of the case does nothing for that. All heat must pass through the CPU before the cooler gets rid of it. Coolers are really the waste heat removal system. There is no prevention, except by lowering Vcore.

 

Actual or expected coolant temp ranges vary with case, layout, CPU, and power settings. On average, most people will see a minimum coolant temperature of +4-7C above their room temp. Also keep in mind the real minimum is not the room, but the case. If you play a heavy GPU game for an hour and the case heats up to 35C, your coolant, board, RAM and everything else will have a minimum temp of 35C. While this will also make the fans speed up and not make the CPU cooler, don't forget the cooler is part of the intake/exhaust system and warmer case temp means you need more air exchange to bring it back down. Or open the door.

 

For maximum, I would expect around +4-6C rise for a 100% CPU load. The trick is combination CPU/GPU loads like gaming. As mentioned above, the GPU is going to heat up the case. Rather than set up your fan curve for some theoretical thing like running Prime95, take note of your coolant temp peak for your normal activities. If you baseline typically increases from 24-27C from morning to late afternoon, make sure you leave those temps as quiet desktop speeds. You can start to increase the fan speed +3-4C down the line. Remember +2C coolant temp only equals +2C CPU temp, so there is no reason to fuss about small changes. If you mostly do a lot of gaming, figure out the peak coolant temp when doing a long session of that. Set that temp with a comfortable and acceptable fan speed. If that is 37C, then make a usable point there. Then +5C down the line ramp the fans up to higher and noticeable level. If for whatever reason you get out of your normal zone, the fans will alert you. This can serve as an early warning system for trouble.

Thank you for all of your answers guys. I will take them all into consideration. For now as I see it, I'm going to setup my cooler like that:

 

H115i PRO

-Sensor: PUMP Temp

PUMP Rate: Extreme

 

First, I will wait for 20-30 minutes to get the idle minimum temp so I can use it as a base. Then I will setup the fan curve to Balance preset and try a stress test. Then according to my results I will try to adjust a custom fan curve based on the results with 3-4C difference for every fan level and repeat the test. If everything works fine I will proceed to some gaming to check the temps from that point of you as well.

 

If i'm forgetting something or something doesnt make sense feel free to correct me.

 

Last but not least what about the case fans? Which sensor shall I use for them? The motherboard's one, VRMs or what? I got the ML120 PRO RGB fans (3) installed on Carbide 400C one on the back and 2 on top. My AIO is front mounted pulling air in with x2 140mm ML140 PRO White LED

You won't see any appreciable performance difference between Balanced and Extreme pump speeds. You will see a difference between Quiet and Balanced.

 

For the case fans, you need a temperature that will give you an indication of the interior case temperature - that's what you are cooling, after all. The Maximus X actually has a connector for a 10K temperature sensor (like this: https://www.amazon.com/XSPC-XS-10KSEN-Wire-Sensor-10K/dp/B00CMR38LC/ref=sr_1_4?ie=UTF8&qid=1537115633&sr=8-4) and the Commander Pro comes with 4 of these. My case fans run from just such a sensor that's placed just over the pump head - this provides a good reading on the heat thrown off by the GPU.

Isn't the Motherboard readings enough for monitoring it and configuring the fans based on it?

 

Unfortunately I don't have enough space to even mount a commander pro on the back of my case. Things are pretty tight in 400C. The plastic 2.5 inch mounting kit takes too much space

The motherboard does have sensors ... but where are they on the motherboard? What are they measuring? Are they appropriate for fan speeds? Maybe ... maybe not.

 

With your motherboard, a $7 sensor will connect right to your motherboard. You'll be able to place it exactly where you want it and control your fans based on its temperature.

And where would that value appear? Probably I will buy one for better measurements then!

It'll appear in the BIOS ... I think as "Temp Sensor 1" or "Ext Temp Sensor". And it is select-able as a source temp for a fan curve - I had mine set up this way before I got a Commander Pro.

Alright then, that's all I guess for now at least. Thank you once again all of you for taking the time to help me out with my questions. I'll follow all the advices that I received and see how things turn out.

 

Kind regards!

Hello there,

 

Well during the afternoon where room temp is 26, my PUMP Sensor idles at 30-30C. Also there are the results from after a Cinebench test that I tried to do. Are my temps fine? Note that my XMP Profiles is activated only and I pressed "NO" when the dialog about improved performance from ASUS pop up once I activated the XMP Profile. CME is on AUTO as well. (Don't know if it's better to disable that).

 

http://i.imgur.com/he1IxnY.png

Hello there,

 

Well during the afternoon where room temp is 26, my PUMP Sensor idles at 30-30C. Also there are the results from after a Cinebench test that I tried to do. Are my temps fine? Note that my XMP Profiles is activated only and I pressed "NO" when the dialog about improved performance from ASUS pop up once I activated the XMP Profile. CME is on AUTO as well. (Don't know if it's better to disable that).

 

Yes, they look good. If your radiator is configured as exhaust, the internal case temperature is more important than the room ambient temperature (though they are related). But a 4C delta at idle is absolutely within expectations.

 

How warm does the coolant get during the load test?

Yes, they look good. If your radiator is configured as exhaust, the internal case temperature is more important than the room ambient temperature (though they are related). But a 4C delta at idle is absolutely within expectations.

 

How warm does the coolant get during the load test?

 

34.1-34.2 (most of the times 34.1) and during gaming sessions 30.9-31.3.

I tried to do a stress test with OCCT but exactly at 9 minutes and 18 seconds it crashed. (Timer stopped, despite the stress kept running) and my iCUE led profiles changed to default so I guess that was system instability. Weird thing is system runs on default settings. I will re-run the stress test later tonight to see if it crashes again, the CPU Freq was 4300 and the temps as shown below (got a screenshot, it didn't give me a full report as it usually does).

 

I just got the Maximus X Hero and got tons of settings and I find it really confusing to configure it to run just the Intel stock settings and just the Corsair's XMP Profiles. I decline the auto enchantment though once i activate the ram profiles. (I don't remember if i had system enchantment activated when the instability occured, so I reset to Defaults and activated just the XMP Profiles and pressed no). CME is on but max Freq is 4.3 on AUTO all cores.

 

here are the results of the crashed test. ( I know image's quality sucks) also my time got changed automatically :D :D :D

 

http://i.imgur.com/fz63svr.jpg

I think that you're in good shape. iCue does struggle to keep up when under heavy load but the fan curves should actually be running in the hardware so you'll be OK with that.

But there was system instability as it crashed. I will re-run it and comeback with a feedback. Temp side, I thought 70 was high for stock settings :D

But there was system instability as it crashed. I will re-run it and comeback with a feedback. Temp side, I thought 70 was high for stock settings :D

 

For a max temp under load on a stock CPU? Absolutely not. You might even get quick jumps into the 50s or higher while idle. You'll get better temps if you delid.

I won't risk a delid but i know it's the best solution for improved thermals. Though it runs on 4.3-4.4GHz, and the thermals are considered good? I'm aware that the voltages are not ideal on stock, maybe i will try to change that but other than that, I'm not that experienced to delid my CPU. On idle CPU temp is steady at 33-35, I haven't noticed any spikes on idle. But since thermals are fine under stress test (it stays steady at 4.3GHz (4.4GHz cache - 4.7 Turbo according to BIOS) and the only 'heavy task' that i do is gaming which is about 49-50 (spikes almost 60-61) i think it's fine, isn't it?

Yup, perfectly fine. The only other thing to look at is the vCore. Especially with the default configuration on the Asus motherboards (with "Multicore Enhancements" enabled), we've seen spikes in the vCore (voltage that the processor gets) which causes "hot flashes". But that doesn't seem to be the case here at all. You're good. :)

Yup, perfectly fine. The only other thing to look at is the vCore. Especially with the default configuration on the Asus motherboards (with "Multicore Enhancements" enabled), we've seen spikes in the vCore (voltage that the processor gets) which causes "hot flashes". But that doesn't seem to be the case here at all. You're good. :)

 

The thing is that with MCE disabled and SVID Support on AUTO, after a few seconds into the Stress test (AIDA64, Prime95, OCCT) the CPU underclocks itself even below the stock speeds (!).

 

I spent a lot of time today playing today and temps didn't go beyond 60 (MAX), Average 45~53-55) and freq was 4300-4400 Mhz. I mean even when I try to stress the CPU it won't reach anywhere close to 4.7, with or without CME, which seems odd. I remember being able to run a stress test with steady frequenices on my AORUS Gaming 7 which had hot VRMs compared to Maximus X. I don't think it's a MOBO issue, I just can't figure out what's keeping the CPU 'back"