Setting up watercool with some Q's

[ForTiLLiaN]

About what speed do i want my pump running (GPH)? Also, i leave my window slightly open all night to keep my radiator cool (radiators mounted to window blowing outside air onto it), what keep from condesation from building up? Thanks

[Finx]

For the pump you probably want to go with a Mag 500 GPH at minimum if you have a full tower and if you have some extra money go with the model 7 that is 700 GPH. Good luck =] Here is the link for the pumps - http://www.petsmart.com/fish/shoppin...duct_770.shtml

- Finx

[mkosem]

that's a little rediculous. it's not necessary to have so much flow. all you need is a pump that can provide a decent amount of flow with your amount of head.

--Matt

[ForTiLLiaN]

WEll, there will be some heavy OC, but the thing is if the water is going so fast then it wont have enough time in the waterblock to collect heat, and not enough time in the radiator to cool it down.

[ForTiLLiaN]

Also i work at a petstore, so i can get a pump at cost, just need to know around the right flow rate im looking for.

[Butcher]

Faster water flow does NOT negatively affect temps because the "isn't enough time in the waterblock" or any other crap like that, that's complete fiction. The main thing is bigger pumps are noiser, take more power (and therefore run hotter), and cause more hose pressure (more change of a leak). IF the faster rate thing was true it;d apply for air as well yet people get better results with a delta black label than a cheap 10cfm fan.

[resago]

this arguement has already happened.

once you hit the "magic" flow rate for your system, addition flow yields no better or worse results.

500GPH is probably slightly more than you need, but it woun't hurt you to have that much.

[jastrckl]

http://www2.apex.net/users/timwhita/flowrates.html

The first study I've ever seen on optimal flow rates. YMMV, but he uses a very typical setup. Point of the matter is, buy a strongish pump, and throttle the output until you hit _your_ optimal flow rate.

[hielko]

Weird results. As far as I know the higher the flowrate the lower the DeltaT between the waterblock and the radiator. And lower is always better in this case...

[Finx]

I am having to agree with delta on this - the more gph the better its plain and simple - especially if you throw in some water wetter and it will also increase your fricion. a slower pump also could result in not even being strong enough to run your loop - that is the exact reason i got my new pump. And if you have multiple radiators , waterblocks and/or loops it would be a definate need.

- Finx

[ForTiLLiaN]

Any one ever heard of water cavitating? Its when you restrict the pump and it tries to pump out more than it can take it and thus u get air in the lines. And also with the 250 mag drive i tested, it heated the water up, would my radiator take care of that long with the CPU heat? thanks

[Darkhorse]

I've not heard of 'cavitating'. Can you point us to more info?

[ForTiLLiaN]

Ive havent found an article on it yet, but i work at a Pet store and its something that is know to happen when working with the low pressure water pumps. Water cavitation is formation of vapour bubbles within a liquid at low-pressure regions that occur in places where the liquid has been accelerated to high velocities, as in the operation of centrifugal pumps, water turbines, and marine propellers. Cavitation is undesirable because it produces extensive erosion of the rotating blades, additional noise from the resultant knocking and vibrations, and a significant reduction of efficiency because it distorts the flow pattern. The cavities form when the pressure of the liquid has been reduced to its vapour pressure; they expand as the pressure is further reduced along with the flow and suddenly collapse when they reach regions of higher pressure. The sudden growth and collapse of these vapour cavities cause the extreme pressures that pit the metal surfaces exposed to the cavitating liquid. In other terms: unlike liquid, bubbles can be compressed. When the water heats up it cause expantion and thus the bubbles begin to show. Also with water evaporation occuring the water evaporates causeing high pressured in the lines of cooling. This can cause high pressure and put a leak in you lines. Hope this help, anyways to keep this from hapenning to you use an open resavoir so the air can escape if needed, just a very small hole will do.

[ForTiLLiaN]

Also changing the Line sizes drasticly will cause higher pressure and sometimes can lead to water lock. I had problems when i went from 1/2" to 3/8".

[Darkhorse]

Thanks, good explaination. Might rethink my original idea of going closed loop in my new setup.

[jimsnyder]

An explanation for hielko: The delta drops at the radiator with increased flow, but the chip gets better cooling due to the increased flow. The whole system needs to be considered. Temperatures actually do start to go back up after the flow gets so high, but it is because your more powerful pump is adding more heat into the system. There is a peak efficiency (depending on your specific system) between about 350 and 500 GPM according to published tests.

[hielko]

Xperience: I understand you have to consider the whole system. But that doesn't explain the results of the test a few post above. There all done with the same pump and a control valve. And as long you aren't pumping the water with insane speeds (say 100km/h) I don't see why a slower flowrate can be better.