Strip down part 14 Degassing water cooling solution

The final part of the series that steps out a build that included water cooling for both the CPU and an XFX 5870 video card.

Category:

Howto & Style

Tags:

• Degassing
• 5870
• Water cooling
• PC Build
• Zalman LQ1000 PC building
• How to water cool
• PC
• Water cooling 101
• Part 15
• Stripped down
• Strip
• Stripped
• Computer
• I7 Water Cooling
• VGA Water Cooling
• Vapor Lock
• Quiet PC
• Liquid Cooling
• Radeon 5870
• 5970 Water Cooling
• Radeon Water Cooling
• Best Case
• Computer Case
• Quiet Case
• Coolest Case
• XFX

[OAbrey]

A second thought on that, I think you may have thought that way because of the way the air block removed slowly. Also known as a vapor lock, a vapor lock is elastic and resists clearing, however once removed, flow moves as it should.

[OAbrey]

Actually the flow was adequate. With the flow dialed all the way to the lowest setting, .75l/s the cpu temp is 25C.

I bought a big swiftech mcp655 in case, but I have never installed it... because what is there is good enough. Perhaps it is because I used extra large tubing... the net resistance to flow is less than you were thinking.

[OAbrey]

I can increase the flow and the fan speeds. At maximum, the flow is 1.5 l/min. *( about double the flow) out of 2.0 maximum.

So its flow is about 75% of the maximum measurable.

I have bought a larger replacement pump, but it stands on standby, because I am convinced I will only need it in the event of total failure.

[OAbrey]

Further update, Jan 1, 2011. The longer I have this compy the happier I am with it.

The Y-split to the video card and cpu has not been a problem. Zalman's cases come with a flow meter. It measures from 0-2.0 l/min. In manual mode, dialing the dial all the way down, I record a flow rate of .75 l/min.  The PC's nominal temp is 22C at that rate.

[OAbrey]

Yes it is a y splitter. Remember volume = Pressure x Resistance.

The pressure is constant, the resistance is less but the volume is more so the net cooling effect is pretty equal. What I was most worried about was equal resistance so equal cooling effect would occur. The hose length accounts for some resistance differences in the water jackets. I am confident the cooling effect is even. The Zalman case has a flow meter, and the flow-a bit reduced, but still within nominal range.

[sanahe2000]

1st of all i think your tube is too long and i dont really recommend that case.. i think its waste of money.. because of that case Zalman case isn't really the best of water cooling case... the pump isnt even strong enough to push through all the liquid

[OAbrey]

@sanahe2000

On the pump situation. Its been 1 year now since I started the project.

As I type this the water temp is 24 C. Working it hard brings it up to 27/28 C.

Working it hard on the hottest days of the year had a max time of 32 C.

[miestarSqils]

Thank you very much for this series

[bloodduty]

wow hope you fixed the flow to the right way just looking at that the flow is so week on the gpu side best way is

Reservoir - Pump - CPU - GPU - Radiator with no spitting

[getlow1985]

uhh is that a y splitter!?! if so your losing a lot of pressure try pump -> cpu -> gpu or gpu -> cpu it doesn't matter but you would get better temps by removing the y splitter.

[OAbrey]

@sanahe2000

I respect that observation. There are more suitable cases than the Zalman especially for water cooling--cases that wouldn't require the heavy alterations I had to do to make things fit.

However, I still thing the LQ-1000 is the baddest case out there. The engineering on the case is difficult to describe except that each case looks hand made, by a journeyman craftsman.