water cooling pump help

alichowdry · Jun 18, 2015

Hey guys so basically I completed my customs cpu loop for my gaming pc, pretty decent temperatures, now I am looking to add in my dual gpus, I have a pump/res combo, and I was thinking of buying a second standalone pump, hiding it off somewhere in the case, because with just the cpu the flow rate is not brilliant and adding about twice the tubing to reach the gpus and back to the pump would probably reduce it even more especially having another radiator, would there be any problems using two pumps in one loop?

I'm not looking for any advice saying to just buy a new more powerful pump or anything similar, just whether two pumps would work in one loop

Thanks

drixkarasu · Jun 18, 2015

Yes there is no problem at all using 2 pumps in a single loop.

mcnumpty23 · Jun 18, 2015

it works fine i have 2 pumps in my loop for redundancy should one fail

rubix_1011 · Jun 18, 2015

The one issue to consider is that your current pump (what pump are you running now?) and the second pump might actually run at much different flow rates. This can cause some issues where you're actually competing for flow between them if you don't have enough flow restriction to account for the differences. If possible, can you list what you currently are using...that will help to give more insight into the best move to make.

alichowdry · Jun 18, 2015

rubix_1011 :

well currently I am using a magicool 450l/h pump, and the rest of the parts are all from china, but they all use a 10mm ID. I have one cpu block, 2 120mm rads and the magicool pump/res combo believe it or not I managed to build a fully custom loop for just my cpu for only £50, also believe it or not my idle temps are as cool as 9 degrees C! Let me know if you need any more info

thanks

rubix_1011 · Jun 18, 2015

Can you add a link to the Magicool pump? Curious as to specs/model.

alichowdry · Jun 18, 2015

Sure, here it is:
http://www.watercoolinguk.co.uk/p/12V-DCP450-Magicool-Quiet-PumpReservoir-450lph_50000.html

mcnumpty23 · Jun 18, 2015

looks like a budget pump

and you need to fix how you are reading your cpu temperatures

water cooling wont cool below ambient temperature

so unless its like 0 degrees where your pc is your temp of 9c cant be right

alichowdry · Jun 18, 2015

well it is pretty cold here, close to 5 degrees so yeah

rubix_1011 · Jun 18, 2015

Flow rate on that pump is right in line with a DDC/MCP355, although max head is about 50% (2.4m vs ~4.5m).

mcnumpty23 · Jun 18, 2015

alichowdry :

its 5c? indoors?

rubix_1011 · Jun 18, 2015

mcnumpty23 :

Was also thinking the same thing...

mcnumpty23 · Jun 18, 2015

rubix_1011 :

they didnt say which cpu but think they arent getting a proper reading

i would expect 10c to 15c over ambient on idle once the loop has heated up

and as you said not a great deal of head on that pump--probably why its so cheap

rubix_1011 · Jun 18, 2015

Flow is pretty good considering the head is lower than I would expect, however, we all know that flow rate isn't really notable without knowing head on a pump. I am thinking this is kind of on-par with the XSPC X20 pumps out there.

mcnumpty23 · Jun 18, 2015

rubix_1011 :

its once they start adding those 2 gpu blocks etc that head is going to matter

never tested it but i assume 2 pumps with say 2.5 head still only gives 2.5 head? not 5

rubix_1011 · Jun 18, 2015

It is my understanding that you could theoretically maintain closer to the maximum head pressure of 2.5m (assuming your loop's pumps only were of this model and performance spec) and show a curve that had less declination when using more than 1 pump, but I don't believe you can exceed the max head (2.5m) when it is the max of each pump in the loop. Meaning, you wouldn't double the potential head capability, but you'd be less likely to have flow and restriction degradation due to multiple pumps.

If someone has more info or data that correctly defines this question and answer, please add. I am an IT engineer/architect, not a fluid dynamics engineer.

mcnumpty23 · Jun 18, 2015

rubix_1011 :

yes thats what i would expect to happen also

though dont have a couple of pumps and 5m of tubing lying about to test it out

i bet martins liquid labs have tested it to answer the question before though

mcnumpty23 · Jun 18, 2015

hmmm maybe not looks like it depends if the pumps are parallel or in series

http://www.engineeringtoolbox.com/pumps-parallel-serial-d_636.html

rubix_1011 · Jun 18, 2015

That's pretty interesting. Considering the graphs are correct, I personally wouldn't have thought head would double; I figured it would simply result in a flatter curve of degredation from the max head of 2 like pumps.

water cooling pump help

Reputable

Reputable

Polypheme

Contributing Writer

Reputable

Contributing Writer

Reputable

Polypheme

Reputable

Contributing Writer

Polypheme

Contributing Writer

Polypheme

Contributing Writer

Polypheme

Contributing Writer

Polypheme

Polypheme

Contributing Writer

Share this page