Doubling rads

Catsonar

[H]ard|Gawd
Joined
Oct 3, 2003
Messages
1,886
Hey guys,

Wanted to see and get feedback on this idea. I'll be testing it out but is this useless or will/should it work just fine? x2 420mm 30mm thick(each), 30fpi. This would be with fans in a push/pull config. What do you think?
20220205_140110.jpg
20220205_140213.jpg
received_398359678722742.jpeg
received_367489041442212.jpeg
received_248563620770369.jpeg
 
It will likely help a little bit, but it will be similar to just getting a single thicker radiator.

Conventional wisdom is that using more fan slots gets you more cooling than going thicker does, for a few different reasons.

Firstly, in your configuration the air exiting the first radiator is already going to be warmer than the air that entered it, so when it enters the second radiator it won't cool as effectively as it would if the radiator were standalone. The same effect happens if you just go with a single thicker radiator. There are diminishing returns.

A second factor to consider is getting airflow through the radiators. They are going to require more static pressure if you double them up like that.

You'll get some benefit compared to just running one of the radiators, but you won't get anywhere near the same amount of benefit as you would if you ran the two radiators in a different configuration where they alwerent doubled up.
 
Ideally pump the water through the second radiator first and use the first one to drop the temps the rest of the way.
 
Ideally pump the water through the second radiator first and use the first one to drop the temps the rest of the way.
This would work far better with the radiators in separate locations with a single row of fans each. It doesn't make sense to duct tape these together unless you are totally out of money and your case only holds one 360mm radiator which would be totally incapable of cooling your setup. If that's the issue, you need to get your priorities in order as this makes no sense.
 
This would work far better with the radiators in separate locations with a single row of fans each. It doesn't make sense to duct tape these together unless you are totally out of money and your case only holds one 360mm radiator which would be totally incapable of cooling your setup. If that's the issue, you need to get your priorities in order as this makes no sense.
Yup as stated earlier in this thread two separate radiators is better.

If the choice is made to double up pumping the water in that order will still allow some benefit to be realized compared to a single 360. With the proper airflow of course.
 
Yup as stated earlier in this thread two separate radiators is better.

If the choice is made to double up pumping the water in that order will still allow some benefit to be realized compared to a single 360. With the proper airflow of course.
Radiators aren't really all that expensive either. If you need a really thick radiator due to space constraints, buy one and set it up properly. It won't be as effective as mounting those separately in the chassis, but it will work.
 
Forgot to mention that this isn't going in a case. It's for a PC bench. This is my current rad setup(x2 280mm 45mm thick). Fans being used on the double rad setup are Noctua ippc NF A14 3k.

I do like the suggestion of routing water through the rear rad first then through the front. I'll try that.

Just wanted to try something with two extra rads that I have.
20211108_131433.jpg
20220122_225109.jpg
 
Last edited:
View attachment 441319View attachment 441320

View attachment 441321

Second rad pic was my plan and unfortunately he didn't test that on top of only using 800-1200rpms fans for testing. I will be using Noctuas NF- A14 ippc 3k. I do want to use the flow he illustrates of going from the rear rad to the front rad.
It sounds like the findings from the martinsliquidlab test back up your plan to try with different fans. The conclusion really lays it out there and you do have a plan for high-pressure fans, so I think you should at least try it and let us know what happens!
 
Second rad pic was my plan and unfortunately he didn't test that on top of only using 800-1200rpms fans for testing. I will be using Noctuas NF- A14 ippc 3k. I do want to use the flow he illustrates of going from the rear rad to the front rad.
I can't imagine your proposed configuration varying much at all from configuration #6.

However, if you're going to run those fans at their deafening max speed of 3,000 RPM, then you'll probably see much better results than any of these tests. Whether or not it ends up being better than simply running a single HWL GTR 420 in push/pull with those same fans at 3,000 RPM though... I wouldn't bet my money on it.
 
30fpi rads and 3k noctuas is going to be..., well you already know.
Personally I'd go the rad stand route, like your other setup. Add some qdcs and you could park them anywhere out of the way.
Give em hell!
 
It'd work great with a couple 6k rpm thikk SanAces. ;)

With good sp fans that push enough cfm, it doesn't matter how many you stack (within reason), you'll see an improvement. Obvious caveot being the noise may be unbearable. :D
 
This is the progress I have on it so far. Missing the 3 other "Pull" fans. Just waiting on hardware that I ordered.
20220210_230820.jpg
20220210_230804.jpg
 
If you could have a significant depth I wonder how much benifit there would be to having a small gap between the rads and the fans in a push/push config.
Someone tested that and it was "Not Good", iirc.
 
Link? Want to see what their finding were. Thanks
I can't remember where, was a while ago. IIRC, was on a dual tower cpu heatpipe cooler. They put a fan on the outside of each tower pointing in, and pointing out. Both configurations were worse than any other, and that was on a cooler that allows the air to escape through the edge of the finstack.

Anyway, generally, you want air flowing smoothly from one end of the case to the other so you always have fresh cool air in the case. That config is counter-productive in that respect, and also restricts airflow through both radiators (because the two columns of air collide and then change direction, and hopefully exit the side).
 
I can't remember where, was a while ago. IIRC, was on a dual tower cpu heatpipe cooler. They put a fan on the outside of each tower pointing in, and pointing out. Both configurations were worse than any other, and that was on a cooler that allows the air to escape through the edge of the finstack.

Anyway, generally, you want air flowing smoothly from one end of the case to the other so you always have fresh cool air in the case. That config is counter-productive in that respect, and also restricts airflow through both radiators (because the two columns of air collide and then change direction, and hopefully exit the side).
This isn't for a case.
 
This isn't for a case.
I know, but it would only be marginally better if the reduced flow from the colliding air columns is the primary cause of reduced performance, even outside a case. Not saying you shouldn't try, just saying don't get your hopes up.
 
You said this was going in a computer desk?
No, I'm making a stand for it and tubes will have quick disconnects to easily remove from loop. That way I can compare it against the current 560mm stand pictured above that I'm using. It's for a PC test bench.
 
Last edited:
Back
Top