Advice wanted on Radiator [New to P.C Modding]

ok excellent. I like links for my shopping list!

a week today, I get payed. which means for me, im going shopping with a big fat list. for you, ill stick up a build thread I guess on when parts arrive!

so for now, ill keep my eyes peeled on that page for a 360 copper deal! :D

when I get time, im going to make a bullet point list, with links, to stuff I need to buy. that way, I can keep track of what I need & how much, & tick them off as I go along
 
OP: I'm a bit "tardy to the party", but was there a specific reason for looking at those AMS rads? I ask because if you're only looking for a good 360, you have so many other options, cheaper options that perform better.

I started with a pretty inexpensive XSPC EX360 rad, and that thing performs very well, and has a low fin count. So you can slap any quiet fan on there without worrying about needing high static pressure fans. The only downside is it doesn't have many port options.

If port options appeal to you, I'd 2nd the Alphacool 360 rad recommendation. I just added a pair of 360 Monstas to my loop, and they are EXCELLENT. A fair bit more expensive, but worth every penny. Since you're new to water cooling, the port options will give you much more flexibility with how you run your tubing. Plan on making a royal mess of it on the first attempt. ;) With the port options on Alphacool rads, it's easier to clean up that mess IMO.

Just my $0.02CDN, which ain't worth much. :D
 
Please, do stick around, that's what the forum is for after all :)

I think the idea of going with AMS was to get a single rad solution. Now if something like an RX360 (RX is the 60mm, right?) or a G-Changer v2 will do the job for 2 graphics cards and a CPU - then that's even better, considering they're both a lot cheaper than the copper finned AMS. If you've looked through the earlier posts, you're aware that I can only vouch for what a G-Changer will do with a single 670 and an Ivy running at essentially stock volts, the rest is pretty much my best guess based on the heat dump I've seen into the system, so considering the heat dump of two 7950s and an SB-E quad, I'm interested to see how the radiators you've used measure up. The problem is that the cases Black Magic likes won't take a pair of Monstas, and so far, the AMS is the top candidate for a single rad solution.

Now I'm not sure if it will be any better than a 360 and 240 Monsta in the same loop, so if you've played with these rads, or you know someone who has - the input is welcome :D We don't want to end up with the equivalent of a very good case going into an AMS 360 if a 360+240 Monsta will do the job better, and I hate to be relying on a single web source's synthetic testing (granted, done in a scientific manner) when deciding or helping someone decide, I'd rather hear it from someone who's actually played with them and can compare to different combinations of other radiators.
 
My knowledge is very limited too, as I built my first loop this year. :) The first loop was only 1 7950 (stock) plus a mildly OCed 3820 (4.3Ghz), and I ran the EX360 slim rad (3 CM SickleFlow fans in pull) and an EX140 rad (Bitfenix fan in pull) was added later, just because it fit nicely in the rear exhaust area. :) That rad combo worked very well, but only 1 CPU and 1 GPU isn't much load. Temps were awesome.

The loop I *just* finished 2 days ago has added the second 7950, but also added a pair of 360 Monstas. So my knowledge is VERY skewed, since I've had rad overkill in both builds. :) With the 3 360s, the cooling power is ridiculous, but I never ran the Monstas by themselves to get a proper comparison point. Temps are now better than awesome. :)

But I would vigorously recommend any XSPC rad, slim or thick, because they are a proven performer, and I've seen it. Price is nice too.

OP: some RX360 info for you here https://www.dazmode.com/store/product/performance_triple_120mm_radiator_xspc_rx360/

The EX360 I use is here: https://www.dazmode.com/store/product/slim_high_density_triple_120mm_radiator_xspc_ex360/
 
I see where you're coming from :)

Well my idea was to essentially stick the biggest, baddest rad in there, keep things simple, as this is indeed a first time build, and by carefully selecting the parts, hopefully end up with something sufficient and at least "sort of" neat, that won't need to be redone say 3 months down the line for extra performance. As far as I'm aware, the RX360 follows the conventional radiator design, so it will be at essentially the same performance level as all the other 360x60 radiators, within a few degrees of course. I am at some point going to get a copper AMS myself, because a single 360 is the only thing that will fit in the HAF X without modding, and the AMS is quite a way past even the Monsta rads (just exactly how far though I'm not quite sure :(), so hopefully that'll give me the overhead to stick a second 670 in the loop without having to delid the 3770K just to keep that under 60*C, I'm having to run newer titles without AA, sure sign that there's need for upgrade :lol:
 
I didn't realize the AMS performed that well. I thought Martin did some testing on those and they weren't so great, but I could be wrong, and it was quite some time ago. I hope I'm right, otherwise I'll be kicking myself for getting the Monstas!
 
Well if you go back a page or two, I've linked an article comparing the AMS copper to standard radiators. Please do send a link to any testing, the only credibility I can assign to this website is that it's a data source, and so far - it's my only data source, so as I said - I welcome more testing. Don't forget, they also do an Aluminium version, which I estimate would be about as good as a Monsta 360 (within an error margin, could be better, could be worse) IF my data source is accurate, so if that's what Martin tested, I'm not surprised it didn't turn out so well. Copper has about twice the thermal conductivity of aluminium, so that's what would have held them back if the internal flow design is good. Without knowing a lot more about the internal design and fin design of the rad, and flow characteristics of the loop, there isn't a lot I can tell you I'm afraid :( I'm still working on some equations to estimate overall heat transfer rates of conventional radiators from geometric characteristics, but since I only work on them in my spare time, they're a long way away :( Still haven't even figured out how to work out the flow rate for the entire loop, so we can tell that's going well :rolleyes:
 
& please bear in mind, im only focusing on cooling two cards, not the CPU :)

That aquacomputer seems, from the results shown, looks perfect for me. if im just web browsing, ill just leave the fans on tick over. if im doing any hardcore gaming, even a little touch on fans, will no doubtly cool the cards to reasonable levels.

looking at them results, with game noises coming out my speakers, even if I turned the fans on full blast, I probably still wouldn't hear them. & they'll cool the cards so cold, itll probably start snowing in there LOL!!
 
Doesn't work like that I'm afraid, if you're familiar with the energy conservation principles you'll know exactly why :D

That heat has to go somewhere, and unless you mount that rad externally, that somewhere is your room ;)

If you stick to SP120QEs, even in push-pull, you won't be able to hear the fans over background noise. If your computer is the only thing making noise - then sure, they'll stand out without a fan resistor. Don't forget though that until you start pushing some air through the AMS rads, they don't stand out that much from standard 360x60, so you're probably better off getting some NF-F12s in push. I think they're pretty much built for that kind of rad anyway. I heard that they've released a black version as well, so if you hate the colour scheme, you're in luck :), though I think that stainless and copper on cream and beige would be better than just black on cream and beige.

The F12s are a bit louder than the QEs, however they push more air at the higher noise levels, and you've always got the option to run them at PWM from the motherboard using the included fan splitters, and they get quieter if you stick the included fan resistors on as well.

I find about 800 rpm to be a good compromise of silent running with decent airflow, the LNAs (resistors) take them down from 1500 to 1200 rpm, which makes little difference with a widely spaced rad such as my G-Changer, or even the Monstas, however you will see a greater benefit in terms of the static pressure it will generate to push through the dense AMS fins.

They run silent up to about 850 rpm, then they're audible up to about 1050 if there's nothing else making noise in the room, up to 1200 you start to hear them and past that to 1500 they're hard to avoid unless you're watching something, listening to music, gaming, etc. I find they do the job at 800 in winter times (up to 20*C room), but you'll probably need to have them at about 950-1000. I'd just run them off the CPU header on the motherboard, during testing, stick them on a constant speed to evaluate performance, and if they do the job in the silent range - might as well set them to a constant speed, if not - a fan profile might be needed.

Since you want to use them solely for GPU cooling, you might want to get a thermal probe for the fluid temperature if your board supports one, that might be a better indication than CPU load temps.
 
couldn't say if my board supports it or not. like Ive said, im far from knowledgeable on these kinda things.

& just so you know, fan noise really doesn't bother me much. yes, I know, WC is meant to be for quiet PC's etc, but its ok. im not one to complain that something is "noisy". small things like that don't phase me :)

so how many SP120QEs you reckon ill need? 3 in pull & NF-F12s in push correct?

PS: my last sentence in last post was meant to be a "joke" but I think, that you thought, I was being serious!
 
Black Magic said:
PS: my last sentence in last post was meant to be a "joke" but I think, that you thought, I was being serious!
Click to expand...

One of these :rolleyes: would have hinted :p Or maybe this one :rollinglaugh:, though it's a bit over the top in that case I think ^_^

I thought it was a tribute to Tom's 40 fan troll, but I figured I'd throw some wisdom in the mix anyway.

If I were you, I'd stick to just one type of fan. If you don't care about noise, you'll have good temps running NF-F12 push at full speed, though I'd stick resistors on them anyway, no need to waste power :)

The SP120QEs are more for systems intended to run quietly, not exactly a delta fan by any stretch of the imagination, so better have the option to throw some more airflow in there at the expense of noise. Of course, if you really want just low temps and don't care if your computer wakes up the neighbours, some performance edition fans will go a long way with an AMS rad.

I still think the F12s are a good compromise, considering it won't be difficult to silence them if in the future you decide you don't want noise, while still being able to maintain good performance on demand. Or, if you like the look of them, some NoiseBlocker Bionic eLoop fans will do the job nicely. Just don't use them in pull, the exoframe of the blades will catch on the radiator and make grinding sounds. It will also be detrimental to performance that way. Have a gander at those, should be some on specialtech, scan or aquatuning.

NeverBackDown said:
Stick to the same fans throught a build. Looks more uniform and thought out. I would rather have pull and a thicker rad.
Click to expand...

The only reason for push are the compressor vanes on the NF-F12s, that makes them more efficient in push, as they don't just send swirling air to the rad, they use some of that kinetic energy to compress it slightly, which is useful with densely packed radiators such as the AMS, though not necessary with more conventional designs.

Would an AlphaCool Monsta 360 perform better than the AMS Copper 360? I'm feeling too doubtful over this still.

<EDIT>
Just fount this techpowerup rad round up, they tested the AMS Aluminium, it seems to be falling short of the Monsta among other rads, but I still need to look up a comparison of the Al and Cu rads from an overclockers perspective, because it seems like you'll be better off with a Monsta 360 if the Cu one isn't much better, so I'll keep a lookout for a comparison.
<EDIT>
 
Last edited:
233Kosta, you're a helpful man! If I wore hats, my hat would be off to you. :) Too bad Martin hasn't done testing on the Monstas. He has the Alphacool UT60 consistently at the top of his performance charts, and I *think* the Monsta is supposed to perform better than the UT60 at higher fan RPMs, and equal it at mid/low RPMs. The AMS copper is right up there too. There's no really bad choices for Black Magic here is there. :p Especially only cooling GPUs.

Links to Martins testing:

http://martinsliquidlab.org/2012/04/14/360-radiator-shootout-summary/

http://martinsliquidlab.org/2012/05/12/aquacomputer-airplex-modularity-ams-copper-360-radiator/5/
 
I would never put a rad thats aluminium in a loop. Mixing metals is the last thing you want to do to prevent corrosive growth. As far as i know Alphacool UT/Monsta and XSPC RX are the best rads for low spinning fans.
 
Well I guess I know what rad I'm having next :p

Looks like the Monsta has come out on top, despite the undoubtedly unique design of the AMS. The two links provided by Bartacus (kudos!) put the UT60 on top, and considering the higher fin density and thickness of its bigger brother, the Monsta, it seems we have a clear winner :) Thanks for settling the doubts!

NeverBackDown said:
I would never put a rad thats aluminium in a loop. Mixing metals is the last thing you want to do to prevent corrosive growth.
Click to expand...

I'm curious about this actually, is there a knowledge base out there on why that is, or is it a modder's rule based on experience?

NeverBackDown said:
As far as i know Alphacool UT/Monsta and XSPC RX are the best rads for low spinning fans.
Click to expand...

To the best of my knowledge, Black Magic isn't particularly interested in quiet fans, however it'll soon be time to retire the ol' G-Changer, so I think a Monsta will take its place. How soon exactly is anybody's guess, but whenever it happens, I'll finally throw the motherboard in the loop as well :)
 
I only know all this from the 3-4 years i have been extensively around PCs and that was sort of a big thing in that time frame. People all over were stating don't mix metals.. I honestly can't remember why because it didn't apply to me but i just know you shouldn't..

Ali vs Copper for heat dissapation is quite difference with Ali being a a failr poor conductor compared to Copper(one of the highest conductive metals). So generally speaking if you took 2 of the exact same designed rads and one was 100% Ali and the other 100% Copper... difference would be probably around 5C i would think(same fans of course).
 
Bartacus said:
Hmm, not so fast. :p Bundymania says otherwise: http://www.techpowerup.com/forums/showthread.php?t=173028

So lord knows, I bought the Monstas just because I could fit them in my Case Labs cube. :) I'd definitely second the "no aluminum" thing too.
Click to expand...

Wait, I swear I saw this yesterday, I thought I linked it, though it might have been one of those uni computer fails, can't seem to find it in my post history. In any case, it shows the Monsta outperforming the Al version at higher fan speeds, though considering the heat conductivity of Cu, the Al one is expected to be significantly behind anyway, so that throws some ambiguity in there. It shows however that at very low fan speeds, the thicker radiators (as expected) prove inefficient. Now that's not because they're bad radiators, but because that's not how they're meant to be used. The simple explanation is that viscosity causes a layer of fluid to be essentially stuck to the fins. The slower the fluid moves - the thicker that layer (which has slip within it). It grows with distance, so at some point, the layer at one fin is interfering with said layer at the opposite fin, creating more back pressure, and reducing radiator efficiency. The detailed explanation involves quite a lot of maths, most of which I don't understand (yet), but suffice it to say that laminar flow isn't very good for the efficiency of water cooling radiators. But I'm straying too far off topic :mellow:

NeverBackDown said:
I only know all this from the 3-4 years i have been extensively around PCs and that was sort of a big thing in that time frame. People all over were stating don't mix metals.. I honestly can't remember why because it didn't apply to me but i just know you shouldn't.
Click to expand...

Hmm... interesting. I'll do some digging and poking around in about a week, maybe I'll catch one of the chaps who teach materials and ask about it, it really has me curious :headscratch:

I'm also aware of the thermal properties of copper, and like yourself, I wouldn't use aluminium instead of copper either, based solely on heat transfer numbers, so thanks for bringing up the metal mixing issue, it would have passed me by otherwise :)
 
I think what it does is start algae growth inside the loop and makes the coolant have a cloudy color over time. It gets more complicated but that's the best i can explain it.
 
You must log in or register to reply here.

Similar threads

C
Advise on my proposed build please
Replies
4
Views
647
AlienALX
AlienALX
AlienALX
The Fairchild Ghost subwoofer project
2 3
Replies
41
Views
980
AlienALX
AlienALX
P
Do some people never learn from the past?
Replies
2
Views
914
Pugsport
P
P
Sim Racing Rig Advice/Ideas
Replies
2
Views
12K
KingNosser
K
P
A little bit of advice needed please
Replies
4
Views
3K
Pugsport
P
Back
Top