Jump to content
Corsair Community

Corsair H100i V2 liquid is getting too warm


Rahmanoid

Recommended Posts

Hello,

recently I've finished my first build.

I have the feeling that the liquid of the cooler is getting too warm. When gaming, it usually reaches about 38°C, sometimes even raching 40°C.

My CPU, an i7 7700k then easily reaches easily 70-75°C while it's only under about 60% load.

I have the fan speed set to performance in Corsair Link and also set it to max fan speed all the time, nothing changes. The pump is also set on the performance preset.

My idle CPU temperatures are around 40°C while the liquid is at around 33°C.

Am I doing something wrong to reach such high temperatures? I reseated the pump multiple times, and I'll order new thermal paste soon.

Could it even be possible that my unit is faulty? Is it maybe bad airflow that's allowing such high temperatures?

Someone shared on Reddit with me that they would have way lower temperatures when gaming and when in idle, with the exact same CPU and cooler.

Link to comment
Share on other sites

A 5-7C rise in coolant temperature is pretty normal for a 95W TDP CPU (overclocked too). You coolant to max core temp delta also seems consistent with what others have reported. The detail that jumps out is the 33C idle coolant temp.

 

Generally, your coolant temperature (H100i v2 Temp in Link) should hover a few degrees over room temperature. Specifically, it should be around the same temperature as the internal case temp. You may see more variance depending on mounting location, local case environment (is it trapped under a desk, etc), or if you run fixed voltages with no C-States/EIST. However, the single biggest factor is going to be room temperature. You can't drop below that, so if you are Aussie and it's Summertime 29C in your room, the 33C idle is probably as low as you can go. Fast forward to Winter with 19C room temps and every reading you have will be 10C lower.

 

If you are in the middle of Summer with warm temps, this is likely the answer. If not, we need to start looking at the physical set-up of your case. The model, where is the radiator located, what direction to the fans blow, what GPU is in place, and location in the room.

Link to comment
Share on other sites

A 5-7C rise in coolant temperature is pretty normal for a 95W TDP CPU (overclocked too). You coolant to max core temp delta also seems consistent with what others have reported. The detail that jumps out is the 33C idle coolant temp.

 

Generally, your coolant temperature (H100i v2 Temp in Link) should hover a few degrees over room temperature. Specifically, it should be around the same temperature as the internal case temp. You may see more variance depending on mounting location, local case environment (is it trapped under a desk, etc), or if you run fixed voltages with no C-States/EIST. However, the single biggest factor is going to be room temperature. You can't drop below that, so if you are Aussie and it's Summertime 29C in your room, the 33C idle is probably as low as you can go. Fast forward to Winter with 19C room temps and every reading you have will be 10C lower.

 

If you are in the middle of Summer with warm temps, this is likely the answer. If not, we need to start looking at the physical set-up of your case. The model, where is the radiator located, what direction to the fans blow, what GPU is in place, and location in the room.

 

I live in germany and the room temperature is usually like 20-22°C. The fans of the radiator are pushing air through the radiator out of the top of the case. That is also the only way I can install them, due to my case. My case is the be quiet! Silent Base 600 and my GPU is an AMD Radeon RX480 from Sapphire. My case actually sits under my desk, so I guess that's an additional source of heat.

 

I still can't explain the CPU temp constantly jumping to well above 70°C, often very close to 80°C while loading an area in a game. Is that just how hot Kaby Lakes get? I got told that I have to expect warmer temperatures on Kaby Lake processors then on the previous gerneration CPUs.

 

Thank you for your help!

Link to comment
Share on other sites

With an update BIOS, Kaby Lake actually runs cooler than a Skylake at the same clock speed ... at least mine does.

I think a key question that you need to answer is what is the temperature of the air that is getting pushed through the radiator? As another poster mentioned, that is the coolest that your radiator liquid can possibly be. If your case has poor airflow, it can trap heat. Also, depending on the positioning of your components, you may have heat from your GPU warming up the liquid in the radiator.

Does opening the case cool things down? If so, that's a pretty clear indication that your case is trapping heat. What is your motherboard temperature? Most, if not all, motherboards have a temp sensors somewhere around PCIEX 1. Or ... even better, if your motherboard has an option for a additional temp sensor (mine does), get one and place it right in front of your cooler's fans to measure the temperature of the intake air. In my system, I've seen that my internal case temp is about 2-3C higher than the room temp and the liquid (at idle) is about 2C higher than the case temp. And this is under a desk.

Link to comment
Share on other sites

OK, then room temperature is not the answer and we need to look at other stuff. The Kaby's do seem to hit higher temp peak values and I suspect you can rein those in a little by setting a specific adaptive voltage and moderate LLC level. However, for now I would like to focus on the coolant temps. That is the place the make the largest gains and your current temps are higher than expected. Incidentally, contact between the CPU and cold plate does not appear to be a factor, so you don't need to keep remounting. A premium TIM might take 1-3C off the top end, but that is not always consistent or realized.

 

I would expect your idle H100i v2 Temps to be in the 24-27C range when at a room temperature of 20-22C. I'll start with the bad news. A good portion of this may be the case. I really like BeQuiet's stuff and have a truckload of their fans, however we do see users with SB600, 800, 900 cases experiencing higher than expected coolant temperatures. I don't have one to tear down and they are not very common in the US, but my best guess through 6 or 7 different studies is the the mostly closed top panel (for noise suppression) does not let the heat coming off the radiator escape the local area and creates a warm zone above it. This in turn leads to higher average water temperatures during load. A few possible solutions might be: 1) take the plastic top off the case. Since it is under the desk, this may not be such a visual aberration. See if the coolant temps decrease. 2) Move the radiator to the front intake. I am not sure this is possible on the SB600 or even the larger models. Obviously the first is the place to start.

 

You can try and distinguish a cooler flow problem from an environmental problem by looking at the rate of coolant rise when you first boot up/wake from sleep. If there is a cooler problem, the H100i v2 Temp will start climbing the moment you power on, no real load required. I do not think that is the issue, but it is worth mentioning. When the problem is environmentally created, you will stay at those mid 20's coolant levels after power on, but very slowly over the course of time it will creep up to that 30C+ level. If you are just web browsing, it might take an 30-60 minutes. With load present, probably just 1 or 2 minutes.

 

Room placement - You may not have a lot of option with placement. That's OK, as long as you know that is the cause. If the top plastic cover of the SB600 is causing the issue, then the coolant temps will be in the 30's but other component temps (mainboard, drives, etc) may stay in the 20's. If the entire area under the desk is heating up to 30C, then all temp sensors will be at that level. This is likely to only happen during longer gaming/load sessions and you will probably be able to feel the difference instinctively. If you are seeing internal temperatures into the mid and upper 30's while gaming, this may be something to think about. That is a good 10-15C temperature penalty applied to the CPU.

 

I think the place to start is taking off the top and see what or if anything changes. Maybe give it some time on the desktop first. While and after gaming, compare the coolant temperature to the reports for the motherboard and HDD/SSD temps (or anything else you have). That may give you an idea of how much is CPU waste heat versus GPU and other component waste heat.

Link to comment
Share on other sites

With an update BIOS, Kaby Lake actually runs cooler than a Skylake at the same clock speed ... at least mine does.

I think a key question that you need to answer is what is the temperature of the air that is getting pushed through the radiator? As another poster mentioned, that is the coolest that your radiator liquid can possibly be. If your case has poor airflow, it can trap heat. Also, depending on the positioning of your components, you may have heat from your GPU warming up the liquid in the radiator.

Does opening the case cool things down? If so, that's a pretty clear indication that your case is trapping heat. What is your motherboard temperature? Most, if not all, motherboards have a temp sensors somewhere around PCIEX 1. Or ... even better, if your motherboard has an option for a additional temp sensor (mine does), get one and place it right in front of your cooler's fans to measure the temperature of the intake air. In my system, I've seen that my internal case temp is about 2-3C higher than the room temp and the liquid (at idle) is about 2C higher than the case temp. And this is under a desk.

 

I just downloaded the latest BIOS update yesterday, nothing seems to have changed temmperature wise.

 

I opened the side panel of my case and left it open for about 20 minutes. I took a screenshot of HWiNFO64 before opening it up and after. Seems like the temperatures of the CPU have dropped for about 3°C. The liquid temperature, according to Corsair Link has dropped for about 2°C.

 

I can't tell which of the temperatures displayed is the temperature of the actual motherboard, maybe you know?

713869809_HWMonitorMainboard.PNG.18d8ee073105bf2104a18811de060a47.PNG

1398634119_HWMonitorMainboard2.PNG.9aef496a22d206263de203155e681fe5.PNG

Link to comment
Share on other sites

OK, then room temperature is not the answer and we need to look at other stuff. The Kaby's do seem to hit higher temp peak values and I suspect you can rein those in a little by setting a specific adaptive voltage and moderate LLC level. However, for now I would like to focus on the coolant temps. That is the place the make the largest gains and your current temps are higher than expected. Incidentally, contact between the CPU and cold plate does not appear to be a factor, so you don't need to keep remounting. A premium TIM might take 1-3C off the top end, but that is not always consistent or realized.

 

I would expect your idle H100i v2 Temps to be in the 24-27C range when at a room temperature of 20-22C. I'll start with the bad news. A good portion of this may be the case. I really like BeQuiet's stuff and have a truckload of their fans, however we do see users with SB600, 800, 900 cases experiencing higher than expected coolant temperatures. I don't have one to tear down and they are not very common in the US, but my best guess through 6 or 7 different studies is the the mostly closed top panel (for noise suppression) does not let the heat coming off the radiator escape the local area and creates a warm zone above it. This in turn leads to higher average water temperatures during load. A few possible solutions might be: 1) take the plastic top off the case. Since it is under the desk, this may not be such a visual aberration. See if the coolant temps decrease. 2) Move the radiator to the front intake. I am not sure this is possible on the SB600 or even the larger models. Obviously the first is the place to start.

 

You can try and distinguish a cooler flow problem from an environmental problem by looking at the rate of coolant rise when you first boot up/wake from sleep. If there is a cooler problem, the H100i v2 Temp will start climbing the moment you power on, no real load required. I do not think that is the issue, but it is worth mentioning. When the problem is environmentally created, you will stay at those mid 20's coolant levels after power on, but very slowly over the course of time it will creep up to that 30C+ level. If you are just web browsing, it might take an 30-60 minutes. With load present, probably just 1 or 2 minutes.

 

Room placement - You may not have a lot of option with placement. That's OK, as long as you know that is the cause. If the top plastic cover of the SB600 is causing the issue, then the coolant temps will be in the 30's but other component temps (mainboard, drives, etc) may stay in the 20's. If the entire area under the desk is heating up to 30C, then all temp sensors will be at that level. This is likely to only happen during longer gaming/load sessions and you will probably be able to feel the difference instinctively. If you are seeing internal temperatures into the mid and upper 30's while gaming, this may be something to think about. That is a good 10-15C temperature penalty applied to the CPU.

 

I think the place to start is taking off the top and see what or if anything changes. Maybe give it some time on the desktop first. While and after gaming, compare the coolant temperature to the reports for the motherboard and HDD/SSD temps (or anything else you have). That may give you an idea of how much is CPU waste heat versus GPU and other component waste heat.

 

I thought about those super small air intakes too, but expected that this wouldn't harm the temperatures since I never heard of it. On the other hand, I could find barely any info about it, which backs your claim that this case isn't very common in the US.

 

I'm pretty sure that the temperature under my desk is also heating up notanly. I'll change that when I get to game for a longer time again.

 

I'll take the top cover off and see if there's any difference in temperatures.

 

I'll now turn my PC off to allow the coolant to cool down and turn it back on in like 2-3 hours, then I'll know if the temperature is rapidly increasing or slowly over time. If I rember correctly from my previous times of monitoring the temps, the coolant is warming up slowly, not rapidly.

 

I have the possibilty to place the PC next to my desk, that shouldn't be a problem, hell, I could even put it on my desk, but I'd prefer to not do that since the case is pretty big and I don't want something that sounds like a launching space ship next to my ears while gaming (Thanks standard Corsair fans).

Link to comment
Share on other sites

30 minutes off should be enough for the coolant to naturally come back to room temp. Also, I do not think you have a cooler issue or the load coolant temps would be horrible and much higher. Since the load coolant delta (33 to 38C) is fairly normal, I think the cooler is functioning.

 

The 29C near the top is the motherboard temp sensor. In the current configuration, that 29C is going be about as low as you can go and is likely representative of the general case temperature. However, since the reduction in coolant temperature is applied at all levels, it is worth pursuing.

 

You'll have to experiment with placements and see what compromised you are willing to make. I do think we have isolated the cause and you may also now want to look at setting a specific voltage (I am assuming your at the stock 4.5Ghz) and getting off the Intel Auto marks. They are just too flexible and several other users have stated they hit 1.40V while at the default settings. Getting rid of that should take the top off those peak values and hopefully you can slide the whole scale down several degrees with a coolant temp reduction.

 

Oh, and yes the grey SP120L fans are not acoustically pleasing, especially if you are getting near 40C where they will max out on speed. Down the line, those are the first thing I would replace.

Link to comment
Share on other sites

Having my PC on for 30 minutes now. Coolant started out at 27,4C, so still 5 °C above ambient temperature. It started working it's way up to 29,5°C in the 30 minutes, only using my browser and watching videos, so I guess that should be pretty good.

 

I defintitely wanna replace thos horribly loud fans too. Been looking at some Noctua fans, heard they are pretty much the best reagrding temperature and performance, as of now. But I still need to get an SSD (waiting for the 250GB Samsung Evo 960 to be released over here), so money is the restraining factor right now.

 

I will try to get into this whole voltage thing, but I'm a bit scared to do it since I suppose messing up is a very effective way to kill the CPU and MB, and I've never done it before.

 

Thank you so much for helping finding out what causes those temperatures! :)

Link to comment
Share on other sites

Some things to think about:

1) You want cool air in.

2) You want cool air blowing through your radiator.

3) You want warm air out.

I looked at your case design and it looks like you have some really thin slats above the radiator, making it more difficult to get warm air out. I wonder if it's building up near/around your radiator, limiting your cooling capacity.

Also, the fan at the back of the case is typically configured for exhaust. This will pull warm air past your radiator fans on the way out. I have mine reversed - that rear fan is configured as intake on my system, pulling cooler air in and past my radiator fans. I have pretty high positive pressure and my only exhaust fans are at the top. Side, front, bottom and rear are all intake. All with screens, of course.

My motherboard (and yours as well) has pins for a temp sensor. I'm using this: https://www.amazon.com/XSPC-XS-10KSEN-Wire-Sensor-10K/dp/B00CMR38LC/ref=pd_rhf_se_p_img_3?_encoding=UTF8&psc=1&refRID=M4M4PG7BS15FZ2FCCZKS. I have it set to about 1" below my fans intake by lodging the wire between the two fans and I use this temperature (primarily) to control the case fans, not CPU temp. As a result, I don't think I've ever seen my coolant temp get to 40C. At idle, the coolant temp is 1-2C above that temperature.

I have my CPU overclocked to 4.8 Ghz and I've seen it hit 80 or so ... when stress testing. During gaming (Overwatch), it might hit 70 at the most but averages around 55-60C. The case does heat up a bit from the heat off of the GPU - this may also be impacting you. I do have a custom fan curve set that is a bit more aggressive than the default curve but the coolant just doesn't typically get warm enough for the fans to be a nuisance. It's not so bad that I can't live with the stock fans until I get around to putting a couple of ML120s on there. That said, my case is has a bit of mesh all around, including at the top, so air does move.

I'd be interested to hear results of some more experimentation, particularly taking off the top, on your coolant temps. You might also want to consider flipping that exhaust fan to intake and/or adding the side fan and opening the adjustable vent on the side panel. I'm just not sure you have enough escape routes for warm air. :(

temps.thumb.PNG.b2c4189e509dd757e7ddd75f292568d3.PNG

Link to comment
Share on other sites

Some things to think about:

1) You want cool air in.

2) You want cool air blowing through your radiator.

3) You want warm air out.

I looked at your case design and it looks like you have some really thin slats above the radiator, making it more difficult to get warm air out. I wonder if it's building up near/around your radiator, limiting your cooling capacity.

Also, the fan at the back of the case is typically configured for exhaust. This will pull warm air past your radiator fans on the way out. I have mine reversed - that rear fan is configured as intake on my system, pulling cooler air in and past my radiator fans. I have pretty high positive pressure and my only exhaust fans are at the top. Side, front, bottom and rear are all intake. All with screens, of course.

My motherboard (and yours as well) has pins for a temp sensor. I'm using this: https://www.amazon.com/XSPC-XS-10KSEN-Wire-Sensor-10K/dp/B00CMR38LC/ref=pd_rhf_se_p_img_3?_encoding=UTF8&psc=1&refRID=M4M4PG7BS15FZ2FCCZKS. I have it set to about 1" below my fans intake by lodging the wire between the two fans and I use this temperature (primarily) to control the case fans, not CPU temp. As a result, I don't think I've ever seen my coolant temp get to 40C. At idle, the coolant temp is 1-2C above that temperature.

I have my CPU overclocked to 4.8 Ghz and I've seen it hit 80 or so ... when stress testing. During gaming (Overwatch), it might hit 70 at the most but averages around 55-60C. The case does heat up a bit from the heat off of the GPU - this may also be impacting you. I do have a custom fan curve set that is a bit more aggressive than the default curve but the coolant just doesn't typically get warm enough for the fans to be a nuisance. It's not so bad that I can't live with the stock fans until I get around to putting a couple of ML120s on there. That said, my case is has a bit of mesh all around, including at the top, so air does move.

I'd be interested to hear results of some more experimentation, particularly taking off the top, on your coolant temps. You might also want to consider flipping that exhaust fan to intake and/or adding the side fan and opening the adjustable vent on the side panel. I'm just not sure you have enough escape routes for warm air. :(

 

Hmm ok, I'll look into that. Thank you!

 

I now left the side panel open while gaming for about an hour, and the coolant temps never went above 35°C. I still got those jumps to the mid-hight seventies, while loading in areas. Also saw it jumping to 82°C once. But the general temps were around 60-65°C

I then closed the panel, and in the duration of a few minutes the coolant temp jumped up to 39°C. As a result my CPU temps went up to the low-mid seventies, while playing normally and high seventies and occasioanlly into the low eighties, while loading an area. With about 60% load while loading and 30% load while playing normally.

 

So I guess my airflow might be a problem after all? If it comes to i'll probaly just cut out more air intakes at the top of the case and maybe install more fans all around. RIP money that I safed for an SSD.

Link to comment
Share on other sites

Hmm ok, I'll look into that. Thank you!

 

I now left the side panel open while gaming for about an hour, and the coolant temps never went above 35°C. I still got those jumps to the mid-hight seventies, while loading in areas. Also saw it jumping to 82°C once. But the general temps were around 60-65°C

I then closed the panel, and in the duration of a few minutes the coolant temp jumped up to 39°C. As a result my CPU temps went up to the low-mid seventies, while playing normally and high seventies and occasioanlly into the low eighties, while loading an area. With about 60% load while loading and 30% load while playing normally.

 

So I guess my airflow might be a problem after all? If it comes to i'll probaly just cut out more air intakes at the top of the case and maybe install more fans all around. RIP money that I safed for an SSD.

I think we have pretty much determined that you have an airflow issue. Now, before you start cutting that pretty case apart ... can you remove the top rather than the side? What difference does that make? Also, have you reversed that fan in the rear to be intake rather than exhaust? (Note: if you do this, you'll want to get a screen for it. Silverstone makes some nice ones.)

That jump to 82 seems a little high, especially if you aren't overclocking. Did you apply your own TIM or did you use the stuff that was already on the cooler block?

Link to comment
Share on other sites

I think we have pretty much determined that you have an airflow issue. Now, before you start cutting that pretty case apart ... can you remove the top rather than the side? What difference does that make? Also, have you reversed that fan in the rear to be intake rather than exhaust? (Note: if you do this, you'll want to get a screen for it. Silverstone makes some nice ones.)

That jump to 82 seems a little high, especially if you aren't overclocking. Did you apply your own TIM or did you use the stuff that was already on the cooler block?

 

I unfortunately won't get to do that till tomorrow, but I'll definitely try it. I'm using the pre-applied thermal paste, but I'll most likely get some new premium paste, maybe that will help to tickle out another 2°C less.

Link to comment
Share on other sites

I unfortunately won't get to do that till tomorrow, but I'll definitely try it. I'm using the pre-applied thermal paste, but I'll most likely get some new premium paste, maybe that will help to tickle out another 2°C less.

 

I can't confirm - or deny - the quality or lack thereof of the TIM that Corsair ships with. I use Noctua NT-H1 because it seems to have the best performance for the price based on reviews that I've seen. YMMV. However, 2 things if you replace the TIM: clean it off well(90% alcohol or ArcticClean) and don't use a lot. You want as much metal-to-metal contact as possible. There are also tools to make delidding simpler and safer ... I've not gone there but am considering it.

Link to comment
Share on other sites

It is interesting you dropped 3-5C with the door off. I am not sure if it is because of an easier pathway to draw in air for the radiator fans or because this allows some of the radiant heat trapped around the top of the case to drift away. You may again be able to try and make a comparison to the motherboard temp sensor in HWMonitor. I am curious if it also goes down 3-5C (suggesting case heat) or is more or less the same (suggesting a heat trap problem at the top).

 

To that end, if the temperatures are only out control while gaming, more rear exhaust may help get the GPU waste heat out. As is, the coolant temp is a little higher than expected at idle. Getting airflow through the radiator while in an idle state is rarely an issue. I can let my slightly larger 280mm sit at the bare minimum fan speed almost definitely without incurring that kind of penalty. This is what makes me suspect the top panel may be trapping some heat around the radiator. At idle voltages, there just should not be a lot of other case heat without doing something else to warm things up.

 

The dual front/back intake layout may work for some cases, however I am not sure about yours. The limitation with that set-up is it leaves the top as the only active exhaust point. The radiator will reduce your fans' total airflow potential to about half the free air specification when restricted by the fins. So, a 63 cfm SP120 likely only moves about 30-35 cfm when at maximum speed or the two together about 60-65cfm. Your rear 140mm can put almost that much into the case by itself, plus whatever comes through the front. All that air then stacks up waiting to get out somewhere. This is an utterly disastrous combination on my 540 and 740 cases. On yours, I worry that with the top dome restriction you can even move 60 cfm out there and this would be like standing on the weak point. That arrangement would also require you to run the top radiator fans at higher speed all the time, something I am not sure you want to do with the current set. Still, I think everyone should experiment and it is a very easy thing to try. You can certainly flip the rear fan around and have a go. Keep an eye on the internal temperatures. You certainly will not hit dangerous levels, but I saw some very eye opening numbers when running this way.

 

We are working on the coolant temps, but those spikes in game are the result of the voltage and CPU instruction received. You cannot address this with any cooler setting, air or water. I think we at the point where you should also set a specific adaptive voltage. I don't have a specific voltage to give you for the stock 4.5 frequency. You can probably find a starting point with a little Google and it will likely be (NO, 1.10v was far to low) a 45x100 multiplier. That is all you really need to change, although I might also change the Load Line Calibration (LLC) from Auto to say a level 6. That should take some of the extra swing out of the vcore variance. In your HW Monitor pictures above, you can see Vcore peaked out a 1.265v at some point (THAT MAY not be too far off the mark). The full AUTO voltage setting is just too flexible, too compliant on the Sky and Kaby Lake CPUs.

 

I would leave the existing thermal paste as is for now while you troubleshoot other steps and until after you have gathered some more data. It is better to change one variable at a time and you don't want to accidentally attribute a substantial cooling change to the TIM. There are very few miracles to be had with TIM and lower TDP CPU's.

Link to comment
Share on other sites

Ok, seems like we finally found what's the problem!

You were right with the top panel trapping the heat beneath it. I removed the panel and played games for about an hour and the coolant temps barely exceeded 35°C (formerly 39-40°C with top closed). My average temperature while gaming was in the high fifties and low sixtees. They even "only" topped out at 74°C while loading (formerly 83°C).

 

I'll send be quiet! an Email and see what they have to say on the matter. I don't want to get my hopes up, but in my opinion this is a no go. They even mention on their website that the case is supposed to have a perfect cooling efficency and air circulation.

 

Now I'll look into the whole voltage thing! Maybe I can reach that the temperatures top out in the low sixties, even while loading.

 

Thank you so much for your help! Without you I probably would have purchased all kinds of unneccesary stuff, because I'd never have expected that such a high quality case has such major issues.

 

Below are the screenshots of HWiNFO64, and a pic of the liquid temp.

1393693819_HWMonitorMainboardOpenTop.PNG.bf424ad7d0f41b0cbbf0aef5e6bda9c2.PNG

467864304_CoolantTemperature.PNG.86a2f8bddd1066972c7e0d7f3d74c23c.PNG

Link to comment
Share on other sites

Unfortunately, this seems to be a theme with the BeQuiet case series. They just weren't designed with water coolers in mind. Now that they have their own series of AIO's, I thought that might change a little. In fairness, that is the catch-22 with noise suppression and temperature. Any design intended to keep sound energy in, will also retain heat energy. The reverse is also true of open mesh cases that let heat and sound pass out of the case.

 

As for the game loading screen peak temperatures, I do think that will be somewhat addressed with putting a cap on the voltage allowance. However, this is also a byproduct of the programming and may not be completely solvable. I am not playing the game in question, but can give a comparative example. I played No Man's Sky the first month or so after release. The PC performance reputation was deserved and the game mostly rolled along on just 2 of my 6 cores. Steady gaming core temperatures were about 30-33C on the two active cores. Then I would hit a load point or some other map change and the the CPU usage would spike to 70% for a split second along with core temp changes to the 65C mark. That is a good 15C more than most other games and nothing I do ever uses 70% of my 5930K. Some games are just like this, but not being a coder I won't criticize any further. In spite of this and the annoyance factor, a momentary blip to 70C is not likely to be damaging to your CPU's longevity. You need sustained, elevated core temperatures to degrade the silicon.

Link to comment
Share on other sites

I've got a quick update.

The be quiet! support already answered to my mail (props for the fast response). The support guy told me, that the fans are supposed to be installed above the radiator, so they suck in air from outside of the case through the radiator.

The only problem with that is, is that the radiator physically won't fit if the fans are above it. The fan at the back of the case will block like 1cm more space, which leads to the radiator not fitting. I could install the case fan at the outside of my case, but I guess that wouldn't be too ideal?

 

I looked up the voltage thing, but couldn't find any proper guides. I'll keep looking though!

Link to comment
Share on other sites

Yes, I am not surprised they would say that. Of course, the majority of people in the world do the opposite for a reason. That moves the heat zone from above the case to inside the case. However, given the other restraints, it may be more practical to try and force all heat of the rear exhaust. To do this, you would likely need to put the radiator against the case (as it is now) and then simply flip the fans the other direction. This will make the label side of the fan visible, pointing down into the case. This is commonly referred to as the "pull position". Lots of people have a preference for one side other the other, but ultimately the difference in efficiency is likely less than the effect of the heat trap.

 

This will certainly be effective for CPU only loads, but the real test will be gaming with high GPU/moderate CPU loads. That puts a lot of heat into the case. Compare both coolant and motherboard temp sensors to the current set-up. You will need some higher rear fan speed when gaming. This solution has been tried with other Silent Base cases with mixed results. I do think it is worth a try.

Link to comment
Share on other sites

Yes, it is not whether the fans are above or below the radiator that makes the difference. It is the fans' direction that changes the equation. Right now, the label side is up and the fans blow out the top of the case. They have suggested turning the fans around to bring air into the case from the the top. You probably have to take the radiator down to turn the fans around, but it can go right back up top as:

 

Case

----------

Radiator

-----------

Fans

|||||

airflow down

 

 

As for the overclocking, you should certainly look at the Asus guide. It is really written as directions for overclocking to 5.0GHz. That is something you can do in the future, but also it has a lot key terminology for understanding the BIOS. I am not sure if there is a German copy and his language can be hard to follow. However, there it has pictures and there is a key element that is new for the 270 series Asus BIOS.

 

On this platform, you can set the IA DC Load Line to a specific value (AC/DC limits). According to Asus, this should reduce the voltage over-swing and may do wonders for your high peak temperature values. Their recommendation is to set the both to 0.01 when using adaptive voltage.

http://edgeup.asus.com/wp-content/uploads/2017/01/UEFI-3.jpg

 

 

It is also clear I was too conservative in my voltage estimate earlier. Using this table from AnandTech, you may need somewhere between 1.23 and 1.28 volts to run a 4.5GHz (45x100). It is also possible setting the IA limits alone, in combination with adpative auto voltage will be all you need at this time.

Link to comment
Share on other sites

I have flipped the fans around now!

Which leads to the coolant staying at 34°C(!) The only problem now is, as you mentioned, that it's blowing all the warm air into the case, which leads to my HDD getting to 41°C while gaming, which is still fine I guess.

The motherboard sensor also shows whooping 9°C more than before, but I suppose that should be fine too, maybe not ideal, but OK.

 

I'll be reading through the Asus guide later, thanks a lot :)

 

Now I can finally enjoy the new Overwatch season with cool CPU temperatures. :D

1824207264_HWMonitorFlippedFans.PNG.912f3ab80c5344324193cae636d86295.PNG

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...