Watercooling Newb Question

Associate
Joined
10 Aug 2009
Posts
690
Location
Malvern
Hello all, so I've been building pcs for many years but only recently got into watercooling.
My graphics card (rx5700) was the first part of the loop and now I've decided to cool my cpu(ryzen 3700x) I will be using a 240(top exhaust) , and a 280 rad(front intake) as that s all I can fit in my phanteks 350x!
Does it matter which waterblock I cool first?

I was thinking 280-Res-pump-cpu-240-gpu-

any advice would be good thanks

Kev
 
Associate
Joined
24 May 2015
Posts
261
Location
London
It doesn't matter, it's a closed loop, the water temp will be pretty much the same everywhere in the loop. The only things that matters is the IN/OUT that some CPU waterblock have and you'll have to follow
 
Soldato
Joined
18 Oct 2002
Posts
8,016
Location
New Eltham, London
other than res directly before pump, try to order your loop with the smoothest runs and overall shortness. If your aesthetic desires result in additional turns be aware that every turn adds friction. Every additional length of tube adds friction. Keep it smooth and keep it short. If you can fit a top crossflow 240 it will help. Also, look for a crossflow that has fitting holes on top and bottom. Having a hole at the very top of your case adds an easy fill/drain port (drain by inverting the case obvs).
 
Associate
OP
Joined
10 Aug 2009
Posts
690
Location
Malvern
I'm using soft tubing to start with, for this build anyway , I'll still like it to look nice, and will bare in mind about keeping my pipework short, stuff I didn't know, probably get it done by Xmas so I'll get a picture up if I'm happy with the results, so far I've got a monsoon kit Coming, along with a second hand ddc pump and top and a cpu waterblock, just gotta get a crossflow rad for the top like you suggested
 
Soldato
Joined
1 Apr 2010
Posts
3,036
I had 2 crossflow rads in my loop and performance wasnt as good as a regular rad from the same make/series, water temps was about 3c higher. I ditched them.
 
Soldato
Joined
18 Oct 2002
Posts
8,016
Location
New Eltham, London
I had 2 crossflow rads in my loop and performance wasnt as good as a regular rad from the same make/series, water temps was about 3c higher. I ditched them.

Interesting. The water goes through a crossflow in a different way of course where the first fan deals with the hottest water as opposed to a regular rad where half of the first fan gets the hottest and is then the last half fan with the coolest. In principle there should be no performance difference.

Jayztwocents did a test and found no difference but as always, each experience is a valid data point.
 
Soldato
Joined
1 Apr 2010
Posts
3,036
Interesting. The water goes through a crossflow in a different way of course where the first fan deals with the hottest water as opposed to a regular rad where half of the first fan gets the hottest and is then the last half fan with the coolest. In principle there should be no performance difference.

Jayztwocents did a test and found no difference but as always, each experience is a valid data point.


I believe its because the x-flow rads are less restrictive (only single pass) so the water ends up spending less time in the radiator and this effects the cooling. Or something to this effect, its been a while since read up on rads.

I was using the xflow versions of the Alphacool XT45 420 and 480 for about 5 months until a review came out of the Alphacool ST30 xflow (slimmer version of my rad). The review showed a fair bit of difference in performance between the xflow and standard rad especially at lower fans speeds (over 4c @ 600rpm). This bugged me so I bought the non xflow versions and tried it for myself. I ended up seeing better performance with the standard rads so I sold the xflows on and kept the standard rads in my loop.

http://thermalbench.com/2016/09/16/alphacool-nexxxos-st30-x-flow-480mm-radiator/5/

https://www.xtremerigs.net/2016/11/26/alphacool-st30-x-flow-360mm-radiator-review/5/


Same result with this xflow rad:

https://www.xtremerigs.net/2015/09/19/hwlabs-nemesis-360-gts-x-flow-360mm-radiator-review/5/
 
Last edited:
Soldato
Joined
18 Oct 2002
Posts
8,016
Location
New Eltham, London
The point about x vs U is that with an X the water crosses in half the time but with twice the flow therefore in practice there should be no difference. Jayz testing showed this exactly.

It's like the argument about fast flow loops and slow loops. In a fast flow, the water goes through the block faster picking up less heat with each pass but there are more passes. In a slow loop, the water picks up more heat on each pass but there are fewer passes.

As in all things watercooling, the myths outlast the facts. Back in the days of the forums on procooling, the great Bill Adams used to do detailed pressure/flow tests of waterblocks, radiators and pumps. Now we get Youtube "stars" doing much more basic tests and proclaiming their opinions to be facts and of course they're better than nothing but they're not detailed. Is a X=U for all flowrates and all heat loads? The same for high static pressure fans vs high airflow fans? Who knows?
 
Associate
Joined
13 Sep 2013
Posts
1,625
Location
Aberdeen
Just one thing, usually good practice to have the pump before the rad as it'll generate a wee bit of heat on its own. Ideally want the rad being the last part before the component being cooled. How much difference this makes, if any, I couldn't comment on as never actually tested myself but have always set up my loops this way
 
Soldato
Joined
1 Apr 2010
Posts
3,036
It's like the argument about fast flow loops and slow loops. In a fast flow, the water goes through the block faster picking up less heat with each pass but there are more passes. In a slow loop, the water picks up more heat on each pass but there are fewer passes.

As in all things watercooling, the myths outlast the facts. Back in the days of the forums on procooling, the great Bill Adams used to do detailed pressure/flow tests of waterblocks, radiators and pumps. Now we get Youtube "stars" doing much more basic tests and proclaiming their opinions to be facts and of course they're better than nothing but they're not detailed. Is a X=U for all flowrates and all heat loads? The same for high static pressure fans vs high airflow fans? Who knows?

Its a shame that custom watercooling is so niche and hence reviews are more thin on the ground compared to other pc related hardware.

When it comes to watercooling I don't really pay much attention to youtubers like Jay because as you mention the review is too vague, in fact these days I only ever watch a small selection of youtube channels as most seem to be glorified adverts. Case in point with the video you linked to, Jay mentions that Alphacool sent him the rads and when he asked if there was any performance differences they said they didn't know and for him to test it and find out. I mean come on Alphacool knows there is a difference in performance between the xflow rad and standard rad layout, if I can see a difference benching at home and the review sites I linked to can then they surely know there is a difference. There is a reason why the standard rad layout makes up most of the rad market.

Personally the only time I would now use a xflow rad is if the layout of my case made it a headache to use a standard rad. I guess 3-4c @ slow fans speeds isnt going to make or break a system but it did bug me when I saw that review hence why I had to try it out.
 
Soldato
Joined
11 Sep 2013
Posts
11,664
I believe its because the x-flow rads are less restrictive (only single pass) so the water ends up spending less time in the radiator and this effects the cooling.
That's part of it, yes... however, most people generally choose X-Flow rads for one of two reasons - Restriction and Tubing. Sometimes both, as with my case.
Restriction - I have a lot of 90º fittings in my loop, and both my GPU block and my CPU block are quite restrictive, the latter in particular. So to help alleviate the issue, I chose rads with low restriction.
Tubing - For both simplicity in routing layout and for aesthetic reasons, I chose X-Flow. Otherwise I'd have had to have tubes going all over the place and bleeding would have been a real pain. This way I tip the case just one way and all the high points are then uppermost, which makes bleeding so much easier.

U-Channel rads have more restrictions, due to the two 90ºs at the 'U-end', which slows the flow and (IIRC) creates more turbulence which can help with the heat transfer... but similarly can also impact an already-restricted loop.

But from reading up, the XT45s are supposedly better with push-pull though, even more so than the XT30s. In push-only they're kinda wasted.
 
Soldato
Joined
1 Apr 2010
Posts
3,036
I mainly chose the xflow rads as it was something different and I thought it would help with the tubing run in the case. Bleeding was definitely easier with the xflow rads that's for sure. However swapping over to the standard rad didn't really make much difference for me in terms of making the the tubing any easier but I am using a big case. Everyone's use case is different.
In terms of restriction I don't have a flow meter so couldn't really gauge the changes to my loop plus the XT45s aren't very restrictive to start with so I am not sure what effect changing rads had to my loop.

Every rad does better in push/pull, the thicker the rad the more the benefit it sees in this config. If I remember correct there are no XT30s but you probably mean ST30s? ST30s at 750rpm in push get outperformed by the XT45s and the gap widens at higher rpm.

https://www.xtremerigs.net/2016/02/27/alphacool-nexxxos-ut60-280mm-radiator-review/5/
 
Last edited:
Top Bottom