Watercooling CPU Only, Does Loop Orentation Matter?

[AkaJay]

i bought the swifttech h20-120 kit and have the pump sucking from the res to the rad and from the rad to the cpu to the res.

i was wondering if this is the correct way to do it? or should i have the pump going to the cpu instead?

 

[madmat]

It really doesn't matter beyond making certain that the pump is feeding from the res. After that it's mainly about keeping the tubing as short as possible.

 

[qdemn7]

I would go Rad > Res > Pump > CPU > Rad That's how I set mine up.

 

[Erasmus354]

It really doesn't matter beyond making certain that the pump is feeding from the res. After that it's mainly about keeping the tubing as short as possible.

What he said.

 

[Eastcoasthandle]

Well that all depends, personally I would never use my pump to push hot water. It makes no sense to put your pump at risk like that. Although there is no real gain in flow or head pressure either way. Allowing all the heated liquid to pass through the rad before entering the pump or res should increase the life of your pump (pending how it was manufactured, allowed liquid heat rating, etc). I have read a lot of posts with people who had pump failures and most of them had a loop orientation of pump>rad>cpu>gpu or something like this. Where the hot liquid is passing through the pump before it's cooled. Now I am not saying that is the reason for the pump failure. What I am saying is that with most cases those with pump failures have this loop orientation. Can you add 1+1= 2 or do you just say 1+1= 1+1

 

[Sideroxylon]

Well that all depends, personally I would never use my pump to push hot water. It makes no sense to put your pump at risk like that. Although there is no real gain in flow or head pressure either way. Allowing all the heated liquid to pass through the rad before entering the pump or res should increase the life of your pump (pending how it was manufactured, allowed liquid heat rating, etc). I have read a lot of posts with people who had pump failures and most of them had a loop orientation of pump>rad>cpu>gpu or something like this. Where the hot liquid is passing through the pump before it's cooled. Now I am not saying that is the reason for the pump failure. What I am saying is that with most cases those with pump failures have this loop orientation. Can you add 1+1= 2 or do you just say 1+1= 1+1

That's a weird theory you have there. Especially considering there is only about a 2c water temp difference anywhere in your loop. That's at the inlet and outlet of your rad. just as an example, at my rad inlet my water is 22.5c right now. That's the hottest part of my loop. I don't think 22.5c is going to help contribute to the death of my pump or any pump.

Don't worry about your pump getting the coolest water temps. Just make sure your res or t-line feeds your pump. The rest of your loop should be the shortest routing that makes sense for your build.

 

[pissboy]

It really doesn't matter beyond making certain that the pump is feeding from the res. After that it's mainly about keeping the tubing as short as possible.

this is correct.

 

[Arcygenical]

Order doesn't matter in the slightest, as long as the right barb on the waterblock is being fed water.

 

[Erasmus354]

Well that all depends, personally I would never use my pump to push hot water. It makes no sense to put your pump at risk like that. Although there is no real gain in flow or head pressure either way. Allowing all the heated liquid to pass through the rad before entering the pump or res should increase the life of your pump (pending how it was manufactured, allowed liquid heat rating, etc). I have read a lot of posts with people who had pump failures and most of them had a loop orientation of pump>rad>cpu>gpu or something like this. Where the hot liquid is passing through the pump before it's cooled. Now I am not saying that is the reason for the pump failure. What I am saying is that with most cases those with pump failures have this loop orientation. Can you add 1+1= 2 or do you just say 1+1= 1+1

Completely and utterly unfounded. As mentioned earlier, loop water temperature hardly varies at all. Your water WILL NOT damage your pump IN ANY WAY. In fact, having your rad right before your pump could cause more damage than good. Most of the damage to a pump is caused by running it dry. Having the reservoir right before (and hopefully physically higher) than the pump means that you run no risk of running the pump dry. Having the rad right before greatly increases the risk of running the pump dry during the initial filling of the loop.

Please don't fall for baseless FUD like this. The order doesn't matter, but it is generally best practice to have the reservoir right before the pump.

 

[TDW]

I have some WC parts on the way. I have been reading every thing I can to make me a good system. It seems like the most bang for your buck would be a Very good rad and a 110v fan. I have been in computers for well, when 8 and 16 k computer were tops. Heat kills computers. Computers do not have fans that put out very many cfm.

 

[Arcygenical]

I have some WC parts on the way. I have been reading every thing I can to make me a good system. It seems like the most bang for your buck would be a Very good rad and a 110v fan. I have been in computers for well, when 8 and 16 k computer were tops. Heat kills computers. Computers do not have fans that put out very many cfm.

A double or triple radiator with 2-6 yate loon fans will be enough. Always.

No 110v fans... No need for CFM's in the 100s.

 

[Erasmus354]

yea, definitely no need for a 110V fan.

 

[friedchicken]

Res -> Pump -> Rad -> CPU/GPU Block -> Res.

Really, you want the pump sucking water from the res, cause the res should work to remove air bubbles from the system.