-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
want to lower temps on loop recommendation
- Thread starter Thekamikazepr
- Start date
- Joined
- Nov 10, 2012
- Location
- Ankeny Iowa
Rads are almost cheaper than new fans
Could you specify? Is that CPU temp, GPU temp, or coolant temp.
You're running some hot hardware so it might be difficult to get temps down with your case. Some great advice has been given in this thread by others (esp. the 120.6 rad recommendation).
I'm not sure, however, that your case will take 120.6 in rads so that would seem to be where you are limited. 120.5 should get you close but 120.6 would be better if you could manage it. Thicker rads will give you slightly better cooling, such as the Aquacool NexXxos UT60 360 versus the XT45 360...but 60mm thick radiators might not fit in your case. I have a pair of UT60 480mm radiators and they are huge, but I am using a Caselabs Magnum TH10 case and it can take anything you can throw at it.
It might also help you to do a little research to see what sort of water cooling performance others with your hardware are getting, paying careful attention to their setups. While it might be hard to find others with your exact setup, you should be able to get some additional insight with some carefully targeted Google searches.
Last edited:
OP
- Joined
- Jan 13, 2014
- Thread Starter
- #25
I forgot to answer your first question.
They have a whinning noise if using a fan controller so you either have low or high rpms, if you use the little resistance extension they have it helps regulate and eliminate the whining noise but RPM take a hit.
I heard cougars have similar performance but well I have not tested them
OP
- Joined
- Jan 13, 2014
- Thread Starter
- #26
Those are components
on low rpms:
54 first gpu (1st item in loop) - 57 Second GPU (main GPU second in line) -58-59 (CPU)
HIgh RPMS
51 first gpu - 54 second gpu -56
Ultra high RPMS (if i remove the resistance) changes the high rpm temps 1-2 degrees give or take.
NOW If i look VRM 1 temps my card with fujipoly thermal has 10 degree lower temps than my card with EK pads. (I bought it used and had fuji pads)
VRM 2 have 2 degree dif same as core
When I clean and replace my loop ill add fujipoly to my main card.
I've seen people with similar setup but with 360 + 240 or dual 360s (removing 5.25s bays and have 10-15 degree lower temps
BTW whats the difference between the Nexxoss and this performance wise other than reservoir and price:
http://www.jab-tech.com/120mm-dual/swiftech-mcr220-qp-res/
I haven't seen reviews of any 240 radiators, but the NexXxos XT45 360 and Swiftech QP 360 have been reviewed here...
http://martinsliquidlab.org/category/radiators/
My last build was dual loop with two Swiftech 480mm rads. I had three EVGA GTX 570 Superclocked cards in a serial water cooling configuration. 120.4 wasn't really enough rad for those cards. My GPU temps were in the mid 50s and I was running Cooler Max Sickleflow 1800 RPM fans and the time I was completely satisfied with it.
My new build has 120.8 of radiator in a single loop. I am now running three EVGA GTX 780 Classifieds in a parallel water cooling set up. My GPU temps get to about 44 under full load. I can knock another 6 to 8C off of that if I run my fans at max speed -- but I don't bother. I can get 44C GPU temps running my fans at about 1200RPM or less. This rig is much, much quieter than my last build.
If you are going to change out rads then I like the NexXxos all-copper units over pretty much anything else in its price class. It's easy to tell someone how to spend their money. But it's up to you to decide what's best for your system and your wallet.
http://martinsliquidlab.org/category/radiators/
My last build was dual loop with two Swiftech 480mm rads. I had three EVGA GTX 570 Superclocked cards in a serial water cooling configuration. 120.4 wasn't really enough rad for those cards. My GPU temps were in the mid 50s and I was running Cooler Max Sickleflow 1800 RPM fans and the time I was completely satisfied with it.
My new build has 120.8 of radiator in a single loop. I am now running three EVGA GTX 780 Classifieds in a parallel water cooling set up. My GPU temps get to about 44 under full load. I can knock another 6 to 8C off of that if I run my fans at max speed -- but I don't bother. I can get 44C GPU temps running my fans at about 1200RPM or less. This rig is much, much quieter than my last build.
If you are going to change out rads then I like the NexXxos all-copper units over pretty much anything else in its price class. It's easy to tell someone how to spend their money. But it's up to you to decide what's best for your system and your wallet.
- Joined
- Jun 17, 2010
- Location
- Marietta, GA
I see a lot of recommendations for 120.6 flying around based on the notion of 120.2 per chip, but have any of you run the math on the heatload for the loop vs dissipation numbers on the radiators? It seems like a lot of people are just shooting into the dark.
OP, you're looking at about 300W per card on those 290s (as per the first review in google for the card). The CPU will pull about 120w. That's about 720W total, at stock.
GPUs don't push power all that hard when you clock them, we'll add a generous 75W per card. CPUs, on the other hand, suck a lot of power when OC'd. Let's call it another 100W. That puts you at approximately 1000W of heat that you need to remove, worst case scenario.
If you want between a 5 and 10C dT, that puts you at 2500+ RPM on a 120.3 radiator.
However, this is not quite your case. You have 120.4 of radiator (though that GT stealth is going to be underperforming, most likely, as it's an older design). To digest these numbers down, 120.1 of rad at 2000RPM (which is, I believe, about where your fans can run) will dissipate about 200W at 10C dT.
Hopefully this helps you understand what all is going on inside your system and make some educated decisions, rather than having to go off of "everyone told me to get 120.6 of rad so I did". Personally, I would be looking in the 120.3-120.4 region, but I run shrouded AFB1212VHE's on my rads because I don't mind the noise when I'm doing something that pushes temps, so I have SERIOUS legs up to 4k rpm.
OP, you're looking at about 300W per card on those 290s (as per the first review in google for the card). The CPU will pull about 120w. That's about 720W total, at stock.
GPUs don't push power all that hard when you clock them, we'll add a generous 75W per card. CPUs, on the other hand, suck a lot of power when OC'd. Let's call it another 100W. That puts you at approximately 1000W of heat that you need to remove, worst case scenario.
If you want between a 5 and 10C dT, that puts you at 2500+ RPM on a 120.3 radiator.
However, this is not quite your case. You have 120.4 of radiator (though that GT stealth is going to be underperforming, most likely, as it's an older design). To digest these numbers down, 120.1 of rad at 2000RPM (which is, I believe, about where your fans can run) will dissipate about 200W at 10C dT.
Hopefully this helps you understand what all is going on inside your system and make some educated decisions, rather than having to go off of "everyone told me to get 120.6 of rad so I did". Personally, I would be looking in the 120.3-120.4 region, but I run shrouded AFB1212VHE's on my rads because I don't mind the noise when I'm doing something that pushes temps, so I have SERIOUS legs up to 4k rpm.
I believe that all of the 120.6 recs were qualified on the notion that the OP wanted a reasonably quiet running system while under load. If he doesn't mind his rig sounding like a jet engine then he could keep his radiators and run with high speed fans.
The end result would be the same as far as temps go. But one would be a hell of a lot noisier than the other.
The end result would be the same as far as temps go. But one would be a hell of a lot noisier than the other.
OP
- Joined
- Jan 13, 2014
- Thread Starter
- #30
OK thanks for the graphs. Also found this post helpful:
http://martinsliquidlab.org/2012/04/14/360-radiator-shootout-summary/
Now is the difference between the
ST30
XT45
and UT60 the thickness of the rad?
http://martinsliquidlab.org/2012/04/14/360-radiator-shootout-summary/
Now is the difference between the
ST30
XT45
and UT60 the thickness of the rad?
- Joined
- Nov 10, 2012
- Location
- Ankeny Iowa
Yes the UT60 is 60mm thick but the rad is really only like 50mm thick due to the screw protectors under the screw holes. 5mm each side
Also what is your loop flow now?
Are you going to your gpu's first?
Also what is your loop flow now?
Are you going to your gpu's first?
Radiator thickness isn't the only difference. The ST30 is more flow restrictive than the XT45 and UT60. The ST30 performs better at low fan RPM than the XT45 and UT60, but not as good as either of those two at higher fan RPM.
You can read reviews of all three at the site I linked you to above...
- Joined
- Jun 17, 2010
- Location
- Marietta, GA
FPI and flow restriction are both huge factors, in addition to thickness.
- Joined
- Jun 17, 2010
- Location
- Marietta, GA
All else being equal, yes. Often times higher restriction is as a result of a mechanism to induce turbulent flow which allows for better heat transfer (though I'm not sure if this is used in radiators).
Also, to circle back some, FPI dictates airflow (though rad thickness does factor in a little bit). Both airflow and...uh...waterflow (I just made that word up) are relevant characteristics.
