View Full Version : best 2x120mm radiator
aznkc730
09-20-08, 11:46 PM
hey guys,
i'm looking into a new wc setup. I've already decided on a MCP655-B pump with a dtek fuzion v2 block, so all I need now is a good radiator. All suggestions would be welcome. And also, how do black ice extreme radiators perform?
Blazing fire
09-21-08, 12:17 AM
Feaser Xchanger or thermochill. Both are great. Get the one which is cheeper.
Those 2 are still better than black ice.
You have to decide what kind of performance vs noise you are willing to deal with first. Then know how much you can afford to spend. In general, it is better to go with the largest rad you can fit as a larger rad gives you more surface area and room for the future. A triple rad is usually going to fit this bill.
When it comes to what rad to buy, there is no one single best. For low noise/low cfm fans, the Swiftech MCR rads offer the best performance/cost ratio. There is only about a <10% difference between them and the Thermochill and Feser rads. The Thermochill rad is the king of the low noise rads in terms of performance, and the Feser is practically a copy of the Thermochill and performs about the same, but both are over twice as much as the MCR rads.
For medium/higher airflow situations, the Black Ice rads like the Extreme will outperform the others since they are thicker with a higher fin per inch density and are optimized for higher airflow.
If you are looking for a good balance between cost/performance/noise, the MCR rads are really hard to beat, and I would recommend finding a way to mount a triple rad instead of a double if you can. A double may pull your heatload ok, but a triple will allow you to use low cfm fans and still have good performance in most situations like yours where you are using one GPU.
Blazing fire
09-21-08, 05:21 AM
I'm having a hard time proving that.
^ that's agreed if you're using low speed fans. Thermochill and xchanger from TFC have low fin density (folds) while black ice has higher fin density. A higher fin density a larger surface area if you're able to push enough air across it, hence the need for a powerful fan.
What about deltas? :)
I'm assuming you're referring to the Black Ice® GT Stealth 360 by hardware labs. There are many triple radiators produced by them. http://www.hwlabs.com/gen2/gtx/index.html
I said the stealth would win at high CFM fans. I got that information from http://www.effizienzgurus.de/main/index.php?content=article&action=view_spec_article&article_id=222.
And martin said the x-flow will win at higher CFM fans. http://www.xtremesystems.org/forums/showpost.php?p=2987437&postcount=3
I'm not 100% certain that the black ice stealth would win the MCR320, let alone the PA120.3. When I googled to compare these 3 radiators, they were all using low speed fans. Therefore it's not relevant.
Because of insufficiant information, I an unable to tell you at what CFM would the black ice beat the PA120.3
So, please give me a source. Thanks!
Conumdrum
09-21-08, 08:11 AM
You want low noise and good heat removal, the Feser and Thermochill. You don't mind a bit more noise for cheaper, I'd say the MCR series.
The BIX is great if you have some panaflows or deltas laying around.
All 3 types will pull out the heat. All 3 have a slightly different usage. Thats the way it works, none are better than the other, because they are used differently. Same with the stealth, I didn't mention that one.
For the money, the MCR series is best. Hands down the best. Works great, almost as good as the Thermochill/Fesser on low fans. With medium fans the MCR is very very close to equal, and at almost 1/2 the price.
Me? The best is Thermo/Feser. I want low noise above anything else and I'm willing to pay for it.
I've read enough charts and reviews and posts for quite a while, and voigts has been at it longer than I have by a long shot. You get to the point where you just know through experiance by being at it a long time and don't need to post links etc anymore.
the testing that Martin from XS has done recently is excellent. Take a look at his website martinsliquidlab.com . Conmndrum said it well in that if you want the absolute best at low noise and don't mind paying for it, then get the Feser or Thermochill. There is no noise difference per se between the MCR and Thermochill/Feser rads. There is just a slight performance difference between them at the same airflow. The Black Ice Extreme rads however at medium and higher airflows can outperform the Thermochill/Feser rads, because that is what they are optimized for.
I personally bought 2-MCR rads on sale at Sidewinder delivered for $100 for my current setup, which is about 2/3 of one Thermochill or Feser, and will outperform them all day long.
I try to read and keep up on this stuff as I enjoy it as a hobby, but I am riding on the backs of others who take the time and money to test this stuff for the rest of us.
aznkc730
09-21-08, 12:44 PM
thanks for the advice guys. just to recap, feser and thermochill are the best, with swiftech's being the best for the money, right? The reason why I am going for a 2x120mm size radiator is because I'm going to try to mod my exos2 to fit it. Basically, I'm going to use the exos2 shell, and put in the mcp pump and the radiator.
Also, I currently have 3/8 tubing, would there be a noticable difference moving to 1/2?
Clockwork_Apple
09-21-08, 01:06 PM
I made the same decision back when I was deciding between the Swiftech MCR series and Thermochill PAs, the MCR has a better performance/cost ratio, and is suited well to low airflow which suits me needs. I couldnt justify the extra cost of a Thermochill for a 10% performance increase.
Blazing fire
09-22-08, 02:42 AM
the testing that Martin from XS has done recently is excellent. Take a look at his website martinsliquidlab.com . Conmndrum said it well in that if you want the absolute best at low noise and don't mind paying for it, then get the Feser or Thermochill. There is no noise difference per se between the MCR and Thermochill/Feser rads. There is just a slight performance difference between them at the same airflow. The Black Ice Extreme rads however at medium and higher airflows can outperform the Thermochill/Feser rads, because that is what they are optimized for.
I personally bought 2-MCR rads on sale at Sidewinder delivered for $100 for my current setup, which is about 2/3 of one Thermochill or Feser, and will outperform them all day long.
I try to read and keep up on this stuff as I enjoy it as a hobby, but I am riding on the backs of others who take the time and money to test this stuff for the rest of us.
Isn't it similar to this?
^ that's agreed if you're using low speed fans. Thermochill and xchanger from TFC have low fin density (folds) while black ice has higher fin density. A higher fin density a larger surface area if you're able to push enough air across it, hence the need for a powerful fan.
I don't recall seeing him doing a rad from hwlabs, except for the 480gtx.
Also, I currently have 3/8 tubing, would there be a noticable difference moving to 1/2?
For 25 feet, 3/8 vs 7/16 = 0.52GPM.
7/16 vs 1/2 = 0.15GPM.
Also, I currently have 3/8 tubing, would there be a noticable difference moving to 1/2?
Not enough to make an appreciable difference. Take a look at the sticky on the impact of tubing sizes.
Blazing fire
09-23-08, 01:32 AM
Hey voigts! Could you give me a source regarding thermochill vs black ice xtreme on high CFM fan please?
Hey voigts! Could you give me a source regarding thermochill vs black ice xtreme on high CFM fan please?
Here is the google translated version from effizienzgurus.de that I've bookmarked that I can easily remember. Its not the only one, but its the only one that comes to mind at the moment.
http://translate.google.com/translate?u=http%3A%2F%2Fwww.effizienzgurus.de%2Fm ain%2Findex.php%3Fcontent%3Darticle%26action%3Dvie w_spec_article%26article_id%3D222&langpair=de|en&hl=en&ie=UTF-8
Blazing fire
09-24-08, 01:36 AM
Great! If you checked my link, it showed "frames"...
What's 400U/min?
Conumdrum
09-24-08, 06:49 AM
Using my general knowledge and such, and looking down the page, I think it means Airflow per min.
hochgeregelten fans means high-regular fans. So I can deduce that 400U/min means low speed fans, like Yate Loons on low.
Just takes a bit of Google. Let it rest Blaze.
Conumdrum
09-24-08, 06:57 AM
thanks for the advice guys. just to recap, feser and thermochill are the best, with swiftech's being the best for the money, right? The reason why I am going for a 2x120mm size radiator is because I'm going to try to mod my exos2 to fit it. Basically, I'm going to use the exos2 shell, and put in the mcp pump and the radiator.
Also, I currently have 3/8 tubing, would there be a noticable difference moving to 1/2?
Ya know? I saw a build log where a guy replaced the rad inside the same thing. Grr if I can remember where. make sure the rad you buy will fit. You can get measurements all over the place, peeps ask that question all the time....well sometimes.
You will get a better temp if you go to 1/2", but it depends.
If your flow rate limited, bigger tubing will help, otherwise it's a small increase, stay with 3/8" for now and see if it's fine. Otherwise, get more tubing....fittings/barbs...clamps.... It's a never ending addiction.
Have fun with it is what matters, and have a crappy second PC as backup in case it all goes to hell. Thats why my son/wife gets my hand me downs over the years... 14 years now. Heck my son gets first pick, he's a gamer, but not a leet as me. I can always go on their PC for help from you folks.
Great! If you checked my link, it showed "frames"...
What's 400U/min?
U/min is rpm
aznkc730
09-24-08, 09:58 PM
Ya know? I saw a build log where a guy replaced the rad inside the same thing. Grr if I can remember where. make sure the rad you buy will fit. You can get measurements all over the place, peeps ask that question all the time....well sometimes.
I've got all the measurements ready. Once I'm on break, I'll start the teardown
Blazing fire
09-25-08, 05:21 AM
Just takes a bit of Google. Let it rest Blaze.
Sorry, but I don't get you. What do you mean by "let it rest"?
Nebulous
09-25-08, 08:09 AM
Sorry, but I don't get you. What do you mean by "let it rest"?
Exactly what it means Blazing: "let it rest" = Stop beating a dead horse. Not the first time you tried to pick an argument and this isn't the first time I had to step in.
-Reported.
Exactly what it means Blazing: "let it rest" = Stop beating a dead horse. Not the first time you tried to pick an argument and this isn't the first time I had to step in.
-Reported.
Neb is correct. Blazing fire, move on with the thread topic and what the OP is asking for please. This thread is not for personal debate. Please start your own thread if you're interested in a debate.
g0dM@n
OCForums Moderator
P.S. aznkc730, if you put your system specs in your original post, it would greatly help. If yo'ure ont producing much heat, you can go with a budget rad. If say you're doing a crossfire/sli setup with monstrous cards and full waterblocks, you'll need something big.
aznkc730
09-25-08, 05:41 PM
aznkc730, if you put your system specs in your original post, it would greatly help. If yo'ure ont producing much heat, you can go with a budget rad. If say you're doing a crossfire/sli setup with monstrous cards and full waterblocks, you'll need something big.
appreciate the tip, but i thought people could see my rig from my signature. All i plan to watercool is my cpu, since gpu watercooling is too expensive for me. Not going to spend $100+ for a gpu block every time i get a new video card.
but back on topic, would a "budget" rad be an improvement over what i have now? On koolance's site, it says the exos2 is rated for 750 watts of heat dissipation, yet I'm gettin pretty high load temps (67C). I found a link (http://www.ocforums.com/showpost.php?p=5803036&postcount=7) showing a review using a thermochill pa120.3 rad, and getting the same load temp as me. But on the other hand, Spawn-Inc was kind enough to test his q6600 at my speed and voltage, and he told me he loaded at 42C with his swiftech mcr-320 rad, which is supposedly not as great as a thermochill.
Conumdrum
09-25-08, 06:12 PM
Ohh just noticed. Your running that Xeon pretty high with a lot of volts. Could you link to Spawns post where he duplicated your settings? He does have a different processor which could make a lot of difference.
Punch in your CPU etc here and Spawns here.
http://extreme.outervision.com/tools.jsp#cpuoc
If that link you provided is accurate and is at 67C under load, and you get the same with a lesser capable system, but someone with a different CPU is MUCH lower, something's odd here is all I can say.
Please don't take this the wrong way, but Koolance is lying to you.
I don't know how Koolance can even think about 750 watts dissipation. It's a 220 rad right? Here is martins testing on a MCR 220 rad, and we all know it's better than that weak Koolance system that has a weak pump/rad, small tubing, and a not-so efficient CPU block.
http://www.martinsliquidlab.com/SwiftechMCR220-Review.html
I quote martin here with the above mentioned rad in the link I provided.
"Just to run through some options/design. A single overclocked dual core processor can dissipate around 125 watts of heat between cpu and pump, so in that case an MCR220 would be capable of delivering a high performance 5C delta with fans at only 1200RPM, this would be a nice silent setup. However on a quad core, you would need more substantial fans for that level of performance where you would need something in the 2000RPM range". UNQUOTE
So how can a known weak cooling system with a 120x2 rad dissipate 750 watts?????
Don't trust Kooloance's word, please
Just get the MCR220 or even better the MCR320, you'll be fine.
aznkc730
09-25-08, 07:53 PM
Ohh just noticed. Your running that Xeon pretty high with a lot of volts. Could you link to Spawns post where he duplicated your settings? He does have a different processor which could make a lot of difference.
I pm'd Spawn-Inc to get him to test his setup for me, so no link there. Thanks for the two websites though. Really helpful to see how much heat my quadcore is dumping. Based on the calculations, my exos2 setup has a c/w of .167, while Spawn's c/w is .0812; that's more than twice as low as mine!
what's interesting is the results from the tech repository. I have no idea why they would get the same load temp as me when they were running a 3x120 thermochill rad on a cpu only loop.
I think your best chance of fitting something into your exos case is the mcr220, the thermochill and other top end radiators look to be much thicker then the current radiator, although I really don't know how much room you'll have to work in your exos once you crack it open
that said, you'd have some extra room on the sides as the exos rad is ~2 inches wider then your typical radiator
and, conumdrum is right, 750 watts is way more then that radiator could handle, numbers I've seen in other reputable sites suggest numbers closer to 200-400 watts on 220 radiators to get reasonable temperatures
QuietIce
09-26-08, 07:55 AM
Since the rad is external, I'd replace everything else in the loop first and see what kind of temps you get. Then you can decide what rad and fans to replace your existing unit with, if any.
You never know how good the thing will perform until you test it ... :)
Blazing fire
09-26-08, 09:35 AM
Neb is correct. Blazing fire, move on with the thread topic and what the OP is asking for please. This thread is not for personal debate. Please start your own thread if you're interested in a debate.
g0dM@n
OCForums Moderator
kk :p. I'm not interested in a debate. I just want to clarify my doubts. I will pm if I have something not pertaining the thread in future.
I'm just glad you gave me a good explaination. At least I know how to correct myself. :)
aznkc730
09-26-08, 02:32 PM
Since the rad is external, I'd replace everything else in the loop first and see what kind of temps you get. Then you can decide what rad and fans to replace your existing unit with, if any.
just to make sure I'm reading this right, you're suggesting that I buy the pump and waterblock first, before buying a new rad? I thought the biggest difference would be in changing the radiator, since i'm running a cpu only loop.
ocing newbie
09-26-08, 03:59 PM
I'm contemplating a little different setup and was wondering would using push pull config with 70cfm panaflos be considered high enough airflow to make the blackice perform well?
Blazing fire
09-26-08, 08:02 PM
According to the German site, http://www.effizienzgurus.de/main/index.php?content=article&action=view_spec_article&article_id=222, it states that at 1200u/min (rmp, thanks supra (http://www.ocforums.com/member.php?u=16439)), the HWL black ice gt xtreme 360 and 360 lite outerforms the PA120.3.
So, yes, 70cfm should be better on the Hardware lab ice gt xtreme than on thermochill PA120.3.
I'm not sure I'd go after the pump first, while I can't find specific numbers for the exos pump, most of the pumps they sell individually seem to have at least decent flow, but i certainly think a better waterblock would help your temps a bit, not that the koolance block is horrible, but its certainly the easiest part to replace at this point
past that, then I'd say go after your radiator.
In terms of pump performance, you might want to give it a test when your replacing gear through your loop, then you can better assess if its holding you back
QuietIce
09-27-08, 01:25 AM
just to make sure I'm reading this right, you're suggesting that I buy the pump and waterblock first, before buying a new rad? I thought the biggest difference would be in changing the radiator, since i'm running a cpu only loop. Yep.
Koolance has two pumps (a DDC & D5) they use for some newer systems but their other pumps are not so good. If I'm seeing this right the exos isn't one of those systems. The pump and res for the exos2 is $65 - the DDC (pump only) is $78 & the D5 is $89.
As for water blocks, Swiftech dropped temps 5°C (or more) in a year from the Apogee GT to the GTZ. I doubt Koolance has kept the pace. ;)
I'm not saying you won't need a rad - I simply don't know if you need one or not. I doubt anybody else does either unless they've been through the upgrade in pieces. If you don't really need one you've saved yourself some $$$ and if you do have to replace it it's easy enough to replace without taking everything else apart ...
aznkc730
09-27-08, 02:12 AM
ok i guess it comes down to this. I probably will end up swapping out everything, but if I could only change one thing, what would give me the biggest performance increase? a pump, waterblock, or a rad?
Conumdrum
09-27-08, 06:52 AM
Not the pump. You'll need to redo hose size down to 3/8.
Not the rad. LOTS of modding, probably will need a dremel etc.
Block is your only choice.
My suggestion is to save your pennies till you can get it all in one bang. You will save shipping costs by bundling it.
Your system needs replacing. Thats the key here. By piecing it it's more work/frustration etc. Drop your OC by a bit so temps are better till you got enuff to buy the setup.
My suggestion IMHO. Had fun working this out with everyone.
Rickpatbrown
09-28-08, 12:41 AM
Koolance isn't lying with the 750watt dissipation. They are just being misleading. If you put boiling water in the loop and ran it with a house fan, the loop could dissipate more than 750watts. The question is what kind a water to air temperature delta you have. The lower the better, 5-6 degrees C is excellent.
Hit Martins Liquid lab site hard. read up on his testing methods and you'll get a better idea of how radiators work.
It's hard to say what would benefit you the most to change. We need to know your flow rates and water/air temperatures. I bet that you would see a little bump for every component you replace. The most gratifying would be to do it all at once. The water block is the easiest to change.
QuietIce
09-28-08, 11:58 AM
ok i guess it comes down to this. I probably will end up swapping out everything, but if I could only change one thing, what would give me the biggest performance increase? a pump, waterblock, or a rad? That's a tough question. If the pump is only moving 2L/min that's very poor so any block/rad combination will run hotter than normal. On the other hand, the greatest change would probably be from the block since designs have changed so much - even if the pump is moving slow.
I guess the block would get my vote as well ...
aznkc730
09-28-08, 12:04 PM
well i guess that means I'm gonna get the fuzion v2 first unless anyone else has a better block suggestion. Then I'll probably finish out the set with a mcp 655 pump and a thermochill or mcr 220. thanks for all the advice guys
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.