The Corsair User Forums  

Go Back   The Corsair User Forums > Corsair Product Discussion > Cooling

Reply
 
Thread Tools Search this Thread Rate Thread Display Modes
  #46  
Old 02-20-2019, 02:52 AM
elric75 elric75 is offline
Registered User
 
Join Date: Oct 2016
Posts: 59
POST ID # = 994951
elric75 Reputation: 10
Default

Thanks, actually for me I can hear both of the settings, the noise is different though and less annoying with balanced.

Could I use silent when not doing anything heavy and use balanced for gaming, video editing or the change of setting will affect performance and speed of cooling?
Reply With Quote


  #47  
Old 02-20-2019, 10:22 AM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 7,730
POST ID # = 994978
c-attack c-attack Reputation: 103
Default

You can use Quiet 1100 rpm for desktop work. There is a definite increase in coolant temp when doing moderate loads like gaming. Definitely not for video editing or other professional tasks. You will need to manually toggle it to Balanced or Extreme before the run.
Reply With Quote


  #48  
Old 02-22-2019, 02:08 PM
Asryan Asryan is offline
Registered User
Asryan's PC Specs
 
Join Date: Dec 2018
Posts: 6
POST ID # = 995253
Asryan Reputation: 10
Default

Thanks.

I've installed some ML 120 Pro and saw few degrees improvement. Thay have better static pressure than the NF F12 I had before. ( 2.61 mm H2O vs 4.2 mm H20)

However 0 rpm mode does not seem to work with those

Also i've decided to go with a H500M case which apparently has the best airflow, I just hope the H150i will fit but I suppose this won't be an issue?
Reply With Quote


  #49  
Old 02-22-2019, 08:59 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 7,730
POST ID # = 995299
c-attack c-attack Reputation: 103
Default

No, the zero RPM mode only seems to work with the stock ML120/140 Quiet fans on the Pro cooler. I thought I remembered a thread where it was confirmed other ML series fans would work with the cooler's controller, however if performance is an issue my recommendation is to leave them running. Even at the slowest possible speed (~400), you are keep the coolant temp down several degrees compared to zero mode. If you were to jump into a load (render/encode/game/etc), you would be starting at a higher penalty and the cooler is not going to be able to easily work off the excess while watts are being dumped into the coolant stream. It takes about 30 seconds or so to drop 1C with moderate speed, so you would almost need to run the fans again for 2-4 minutes just to get back where you started. Each user can make their own choice, but it seems like a weird concession to make if you care about temperature differences.

It certainly looks like the H150i should fit in the H500M. There are clear 3x120 rails front and top, plus the specs confirm support. However, sometimes there are funny compromises with this stuff and only another owner will be able to address that.
Reply With Quote


  #50  
Old 02-24-2019, 01:38 PM
Asryan Asryan is offline
Registered User
Asryan's PC Specs
 
Join Date: Dec 2018
Posts: 6
POST ID # = 995462
Asryan Reputation: 10
Default

The 0 fan mode was working with noctua but it's ok the ml 120 pro are quite silent i'm having them at min 1000 rpm and it does not bother me, at idle coolant is around 30c at idle

With my new 8086k at 4.8 all core @1.24 (5ghz was demanding too much vcore), i have 60c max in game and around 65 stress test

I hope to see the H500M improve those temps more :)

Last edited by Asryan; 02-24-2019 at 05:04 PM.
Reply With Quote


  #51  
Old 03-01-2019, 11:26 AM
elric75 elric75 is offline
Registered User
 
Join Date: Oct 2016
Posts: 59
POST ID # = 995996
elric75 Reputation: 10
Default

Quick update here.

I got a new case (cooler master h500m), new fans (ml pro 2500rpm) and I put the fans on the top of the case as intake, I noticed that the temps were lower than in exhaust, which is what I always did before.

The coolant in the old case was idling at 32c, now it's more between 27 and 29c.

At what coolant temps d I need to run the fan at 100% ? Does that even matter to run the fans that fast and that loud?

My curve is now like this, maxing at 80%





thanks
Reply With Quote


  #52  
Old 03-01-2019, 01:35 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 7,730
POST ID # = 996005
c-attack c-attack Reputation: 103
Default

Keep in mind idle coolant temp is predominately affected by the local in case temperature. There is little actual heat to dissipate at idle. Your causal idle temp will more or less be the thermal temperature layer where the radiator is located. More telling might be what happens at load, but these are still small changes in a matter of single degrees.

So you have 2400+ rpm fans on the H150i now? When do you need maximum? Never. I am trying to remember what CPU you are using. Anything under 200W (pretty much all 6 core processors and down) you won't need much beyond 1500 rpm even when benchmarking. An 8-12 core processor capable of 250W+ likely shifts the scale a bit, but we are still talking about single digit temperature differences. When you start getting into thicker medium or high density custom radiator pieces, then the potential for heat dissipation can be quite high at these 2000 rpm speeds. However, the medium density, thin (27-30mm) radiators used on AIOs and in many custom loops as well, don't really benefit a tremendous amount from high end speed. The larger the amount of heat to get rid of, the more beneficial, but the opposite is true as well. Down at 150W (a typical "heavy" CPU load for a 4-6 core Kaby/Coffee Lake), you may only see 1C. Mine is typically +4-5C with a 8700K@5.1 at weak fans at 900 RPM on a 280mm. 3C is about as low as you can go, so there isn't much point to blasting for a 1-2C reduction on end CPU temp.

The links below are to graphs from a well known set of reviews for 360mm radiators. One of the models I consider to be below average. The other is somewhat exceptional, given its 27-30mm frame. So this is your range from terrible to terrific. Note you get a pretty decent chunk of coolant reduction (5C) when going from 750-800 up to 1300 rpm. However, the reduction in temperature at 300W from 1300 to 1850 rpm is much smaller, 1-2C. This is about what I would expect on the H150i for a similar 300W load, but then your wattage is probably 150-200W. That may cut the expected change by 50%. The take away is your sweet spot is likely between 1200-1400 rpm on the radiator. You can experiement away and see how it turns out. There is always a measure of uniqueness for each case, radiator, CPU combo.

Result 1

Result 2
Reply With Quote


  #53  
Old 03-01-2019, 01:56 PM
elric75 elric75 is offline
Registered User
 
Join Date: Oct 2016
Posts: 59
POST ID # = 996008
elric75 Reputation: 10
Default

I have a 8086k,

So what curve would you suggest me to do for that? Is 80% good for a max? in game now, like Assassins creed, the coolant reached 32c max, before it was around 36 on my former case and with the fan on exhaust instead of intake.

But those fans even at 1200rpm i can't really hear them. Right now im at 29c on coolant, silent mode on the pump, 1000rpm on the fans



I have a non related question, i set up my cores to 48 except the first one to 50 to get the turbo, I tried to OC to all cores to 50 but it was demanding too much voltage to my taste (error in occt at 1.3)

thougn when i check Hwinfo, i never saw the core 1 reaching 5Ghz, any ideas?

Last edited by elric75; 03-01-2019 at 02:01 PM.
Reply With Quote


  #54  
Old 03-01-2019, 03:42 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 7,730
POST ID # = 996025
c-attack c-attack Reputation: 103
Default

I would set 1500 rpm to match the highest coolant temp you’ve seen - probably 36C. Save the 80% blast for 40C to let you know something is out of the normal range. You now know your efficient fan zone is 1100-1300, so I would set that in the range where you spend the most time - 29 to 32C. Remember, one really warm day that raises the room temp by +4C will offset everything by the same 4C. 32C is as high as I ever get in Odyssey and that is more than most other games. You must have had some kind of airflow restriction in the old case.

When you set one core to 50 and the rest to 48x, the turbo function will only kick in when a single core is active. No other core can be working or they will all run 48x100. This condition will never occur on a modern processor. You would need to kill everything, then run a 1 core program like Super Pi 1.5. Almost no desktop utility. Something else is always going on. You can set your first three cores to 50 (likely need a tiny bit more voltage), but this will only work at the desktop. All games will be using multiple core, although no promises on efficiency.
Reply With Quote


  #55  
Old 03-01-2019, 03:56 PM
elric75 elric75 is offline
Registered User
 
Join Date: Oct 2016
Posts: 59
POST ID # = 996028
elric75 Reputation: 10
Default

Thanks :)

Does that seems ok?



i will try to oc it to 5Ghz all core but not sure it worths it ..

Tried at 1.3 on occt but cpu packaged reached 70c, only for few sec but occt found some error so...

Will see at 1.35 if the temps stays ok but 48 all cores at 1.24 is quite good I think

Last edited by elric75; 03-01-2019 at 05:01 PM.
Reply With Quote


  #56  
Old 03-01-2019, 06:16 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 7,730
POST ID # = 996040
c-attack c-attack Reputation: 103
Default

Fan curve looks good. Remember to always make alterations for noise. That -100 rpm less will not make more than a fraction of a degree difference.

1.30 to 1.35 is probably too big a jump. Try moving at 0.015v at a time. So if 1.30v produced errors but no crashing, you are close. Try 1.315v or 1.32.

Is it worth it? Don't know. Not in terms of FPS at this point, so you could quite easily sit at 4.8GHz, 1.24v and enjoy the lower temps with no real cost in performance. The counterargument is OCCT temps only matter if you use that kind of load. My 5.0/1.30v setting for the 8700K will top out at 60C for my uses (delidded), so that is quite a large safety margin.
Reply With Quote


  #57  
Old 03-04-2019, 01:20 PM
elric75 elric75 is offline
Registered User
 
Join Date: Oct 2016
Posts: 59
POST ID # = 996364
elric75 Reputation: 10
Default

Thanks, will try some test when I have time.

I'm really happy with my new case and fans; i'm arond 27c when doing work or just being online on the coolant temp, it's crazy the difference with a better airflow.

I had to change some stuff on my fan set up too.


By putting the aio fan on intake, the temps have been lower. ON my case it was not possible to put the fan under the rad so i put it above but the issue was that the fans were not blowing through the rad, by putting it intake all seem better.

Maybe i should have put the aio on front and let the exhaut on top but it will be less esthetic sooooo

Reply With Quote


  #58  
Old 03-04-2019, 06:16 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 7,730
POST ID # = 996397
c-attack c-attack Reputation: 103
Default

2x200mm fans on the front is the signature look for that case, so it would be a shame to mess with that. As for intake vs exhaust, don't decide based on idle numbers. The intake air is typically cooler and there is a complex relationship between air temp in vs coolant exhaust out. Regardless, a 1-2C change in idle temp is irrelevant. Your room temp fluctuates more than that. What is likely more important is what happens during the course of normal use or load uses. If the top is intake and you are not able to exhaust the majority of the waste heat (CPU and MB components + GPU), then the entire case heats up and the coolant along with it. Even when top exhaust, the upper part of the case is going to be a few degrees warmer as heat comes of the VRM heatsink, RAM, etc. It's subtle differences normally. However, if you see really large changes in coolant temperature, then it may be time to experiment the other way.
Reply With Quote


  #59  
Old 03-05-2019, 04:45 AM
elric75 elric75 is offline
Registered User
 
Join Date: Oct 2016
Posts: 59
POST ID # = 996443
elric75 Reputation: 10
Default

Oh even on load or in game i saw a big difference

In game such as Assassin's Creed Odyssey it went fro 65 to 57 max with those changes
Reply With Quote


Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 03:04 AM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.