Steamをインストール
ログイン
|
言語
简体中文（簡体字中国語）
繁體中文（繁体字中国語）
한국어 (韓国語)
ไทย (タイ語)
български (ブルガリア語)
Čeština（チェコ語）
Dansk (デンマーク語)
Deutsch (ドイツ語)
English (英語)
Español - España (スペイン語 - スペイン)
Español - Latinoamérica (スペイン語 - ラテンアメリカ)
Ελληνικά (ギリシャ語)
Français (フランス語)
Italiano (イタリア語)
Bahasa Indonesia（インドネシア語）
Magyar（ハンガリー語）
Nederlands (オランダ語)
Norsk (ノルウェー語)
Polski (ポーランド語)
Português（ポルトガル語－ポルトガル）
Português - Brasil (ポルトガル語 - ブラジル)
Română（ルーマニア語）
Русский (ロシア語)
Suomi (フィンランド語)
Svenska (スウェーデン語)
Türkçe (トルコ語)
Tiếng Việt (ベトナム語)
Українська (ウクライナ語)
翻訳の問題を報告
ルールとガイドライン
この投稿を報告する
#1 If your AIO cooler is installed correctly, I'm guessing the CPU_OPT is adjusting pump speed. Do you have a separate readout of the pump speed and/or voltage? I would try plugging it into a different header, maybe just a regular one on the mobo?
#2 I haven't used it before so I don't know how effective it is but having dynamic RPM case fans doesn't seem ideal.
And don't worry about the GPU temps. It is getting the maximum performance it can't while staying within it's safe temp/volt/speed range based on its current load.
Also, ambient room temp would help.
Couple of reviews seem to suggest 11-16C above ambient for ~150w load for the S24.
It’s quite confusing. So I’ve got the all the case fans plugged into the case fan header. The case fan header gets it’s power direct from a SATA power cable. There is also a PWM cable which is in the CPU_FAN header. I don’t think this will receive a signal unless I also plug a PWM fan into the case fan header.
So I figure this means the case fans being 3 pin will all just run at full speed? There are no fan readings showing up for CPU_FAN but they are definitely running.
The CPU cooler has it’s own fan header. The two CPU fans are plugged into that. Thus the only cable from the CPU cooler is to the CPU_OPT header on the motherboard. I can get a reading for this which shows flow rate of 22 L/min and RPM of 1900 in BIOS.
Yes I do. I can control this too. If I jump into Smart Fan 5 and turn this up to 100% I can hear the pump ticking over and the CPU fans running full speed.
Why's that? This case came with 3 of the Dynamic X2 GP-14 fans and I bought 2 extra (the bottom fans) to match.
It just seems so strange that a simple game like EVE Online can have this GPU at 82 C as it is at the moment after a solid hour of being in-game. This is a game without about 64% usage based on the settings I play on. The CPU on the other hand is sitting at about 40 C.
Maybe I'm just expecting to much. As soon as I minimise the game, the GPU is back to 40-50 on idle.
I can't give an accurate reading but using my mobile and aircon thermostat I'd guess about 20 C. It's about 12 C outside.
I've also found a lot of successful results from mounting the radiator at the front of the case. I've tried this and starting to see some better numbers. No change for the GPU though.
Fixed! Thanks! Keeping the CPU cooler on a seperate header seems to help. Having the case fans running in sync is what's causing so much noise too because all the fans are spiking with the CPU.
Strange. If it is reading the fans, that is about the max RPM of the fans (~2000RPM).
If it is the pump, it is able to do ~2800RP max.
Flowrate is also unexpected. S24's pump is rated at 40LPH.... So I don't know how it is showing 22L/min, unless it is .22L/min which would mean it is at ~1/3 its rated flow of ~.67L/min.
Maybe try this again and check the pump readings while its at 100%?
Sorry, not the fans themselves. I was still under the assumption that all the case fans were controlled by the CPU temp and were adjusting their speed based on a watercooled CPU, which generally would mean they stay at lower RPMs even though they need to still cool and exhaust air from the case for all the other components.
Don't know specifically about the load for Eve Online but your GPU is a big chunk of silicon able to pull ~300w. That is a lot of heat to be dumping into a case, which is why I was worried about the variable RPM of the case fans.
That is a bit high for that model though, I wonder if the 4 intake fans and 1 exhaust, plus the two variable RPM radiator exhaust fans is not enough to keep up. Maybe try running the two fans exhausting on the radiator to 75%-100% and see if that helps.
That is even more worrisome... CPU idling 20C above ambient and the GPU 35C above ambient... Another test would be to leave all the settings as is and take the side panel off, if things get a lot better, that would likely mean something is wrong with the airflow/fan setup.
That's interesting. Changing the rad to the front would generally show better CPU temps, and slightly worse GPU temps. If the GPU temps are staying the same, then likely something changed with the exhaust setup. i.e. removing the rad from the top opened up more airflow on the top exhaust which negated the extra 100-150w from the CPU being dumped into the case.
When you made that change, did you move the front intake fans to the top of the case to exhaust?
Edit- Ugh, this really is throwing me for a loop and I'm so interested because that is a really nice system with a good cooling setup... I too was planning on building a new rig and using that case.
2nd Edit- This review from GamersNexus breaks down the case's thermals pretty well-
https://www.gamersnexus.net/hwreviews/3177-fractal-define-r6-case-review-mid-tower
CPU cooler is now mounted on the front, intaking air with its two included 120mm PWM fans. This has seen a big improvement. I haven't taken any proper measurements but I was playing ME:A for a few hours and don't remember seeing the CPU above 55/60. This was in a warm room and with the GPU sitting steady on 80 C. Once I finished, after a short while it was back to 30 at idle, I'd guess about 10 above ambient. Right now as of this post, I've got EVE minimised, a few browser tabs open, office stuff and I'm seeing the same 30 C idle. Massive improvement. It's hard to gauge because BIOS/Smart Fan thinks it's idling on 20 C which is ambient. There still seems to be a 10 C difference in measurements.
The was no noticable fan ramping during the game either. Most of the noise was from the GPU which drowned out the case fans anyway.
One thing I could have tried as manually upping the case fans to see if that would improve the overall temperatures, since the CPU had no reason to increase its fans.
This was quite interesting and I got on to Fractal tech support who have explained everything, I was surprised how fast they responded to be honest, should have asked about it earlier.
So the Nexus Fan Controller needs a PWM fan connected (at least 1) in order to control the entire hub. If you plug in a CPU fan for example, it will allow the case fans and CPU fans to all sync as one. 100% CPU = 100% case fan. Likewise if I was to make one of the case fans PWM and plug that in, and then connect the Hub to a SYS_FAN header instead of CPU, then 100% PWM case fan = all other case fans 100% based on SYS_FAN temps.
If no PWM fans are connected, the whole hub will just run the 3-pin fans at 100% as expected.
So the solution at the moment is Cooler pump is plugged into CPU_OPT, CPU fans and case fans into CPU_FAN. Both those headers get the same temperature from the board so the CPU cooler and case fans are now synced.
I think the solution here will be to replace one of the case fans with a PWM fan so I can ramp up the case fans separate from the CPU or allow the fan hub to use PCI-E_16X temp control instead of CPU so that they ramp up with the GPU.
Sorry, not the fans themselves. I was still under the assumption that all the case fans were controlled by the CPU temp and were adjusting their speed based on a watercooled CPU, which generally would mean they stay at lower RPMs even though they need to still cool and exhaust air from the case for all the other components. [/quote]
At the moment they are, I've played around a bit and I think this is the best option as having the case fans at 100% on idle is just too load. Measuring it with my phone it was about 35-40 DB which was 15-20 above the room when quiet.
At least now though I can adjust the fan curve and have the case fans ramping up sooner which at worst means the CPU fans are running faster than they really need to, it's still going to be quieter than the GPU fans until I figure out a new solution.
That's interesting. Changing the rad to the front would generally show better CPU temps, and slightly worse GPU temps. If the GPU temps are staying the same, then likely something changed with the exhaust setup. i.e. removing the rad from the top opened up more airflow on the top exhaust which negated the extra 100-150w from the CPU being dumped into the case.
When you made that change, did you move the front intake fans to the top of the case to exhaust? [/quote]
Yep. The two front intakes are now mounted at the top. The rails on the front allowed me to mount the radiator at the same height as the fans, so I could keep my bottom intakes installed. It also meant I could keep my optical drive but the SSDs had to be moved to fit the hoses. They are now mounted on top of the PSU shroud, which I find looks better. I could have mounted them behind the motherboard but I prefer a little air flow for them. The PSU shroud covers 1 of the bottom intake fans anyway so I don't see it changing airflow too much. My main goal with that fan was to keep positive pressure.
I'll need to do some more tests, maybe try some gaming benchmarks and start writing numbers down and properly recording. I'm happy with it overall. Since relocating the CPU Cooler, I'm seeing better temperatures overall.
I haven't done any gaming yet at the time of this post and the CPU is idle at 30 C and the GPU at 35-40 C. According to my mobile it's 20 C (and aircon heating just turned off with thermo set to 21). Another issue with the GPU is that it's set to disable the fan until 55 C. Which I found pointless. It would turn off, temps could creep back up to 55 and then turn on again. Leaving it on idle instead, and with the CPU relocated and extraction fans up to, this thing will idle at 35-40 now instead.
The easiest way to check if it's airflow problem is by removing side panel and checking if the problem is gone
Interesting explanation from Fractal about the fan hub/fan controller. Is that behavior described at all in the manual?
Anyways, everything seems much better with temps at their expected levels with your new setup.
Glad you were able to get it all squared away.
First thing I did and that's what threw me off as I saw little improvement. I'm convinced the new arrangement is much better.
Yeah starting to see a lot of anecdotal evidence of this. It's just surprising considering the sort of numbers many reviews are pumping out.
I'd argue it's 'implied'. Then throw in a dose of my lack of understanding...
This is how it's presented in the manual:
https://www.techpowerup.com/forums/attachments/2018-02-25-15_05_24-ce46a531-d80e-4c10-82c9-ae19d42d53e7-png.97678/
My understanding was that that without any PWM fans in the PWM headers on that hub, I'd still be able to control the hub by voltage, just without readings. I don't care so much for RPM all I care about for the case fans is a general low/medium/high voltage input. And where it says "CPU_FAN" this can be replaced with any 4-pin header. So if I connect it to a SYS_FAN header, it'll use those readings instead of CPU.
Either way I've just ordered one of the same case fans in a PWM variant and will hook that up to the hub. Fractal support was very fast and very helpful with explaining that and answering my questions.
Will see how it goes. But for now things seem much better.
Thanks everyone for the replies and help!
I was a dumb ass at one point and was benchmarking but had somehow turned my fan curve to a very very shallow curve while overclocking in precision X1 software. I stopped the test when I saw GPU temps get into the mid 80s and realized what I had done after a while.
The fan is currently set to auto. I'm going to have a tinker with it though and maybe try something steeper. I have the feeling some of these games are demanding full GPU usage when they might not even need them. Same with using a frame limiter to see if that helps.