If flow rates matter that much...

[Xel]

Why not do this?

Have two pumps in a reservoir. One pump pumps water from the reservoir to the waterblock and back very fast. The other pump pumps water to the radiator and back, but instead very slow. The one going to the waterblock would put the output water (the hot water) next to the pump for the radiator, and the radiator would put the now cooled water next to the pump for the waterblock. The tubes wouldn't be connected to eachother and would just feed from the reservoir, so that the flow rates wouldn't conflict.

???

[Xel]

[JimS]

Due to the different flow rates, wouldnt one side run out of water after time? Great thinking though.

I would think that the pump going to the waterblock would need more water than the pump going to the radiator could supply.

[Xel]

How would one side of water run out after time? The reservoir would be filled enough so that the radiator would really only cool the reservoir water, not directly the water going to the waterblock.

[Brians256]

You know, that might work well if you tuned it. If the radiator pump was too fast, you would be adding too much heat from an oversized pump. The key would be to ensure that water coming back from the waterblock did not go right back. The hot water should be near the radiator pump intake.

Frankly, I think it is difficult to get this one right because of the increased complexity of the heat paths. I think you'd be better off using a larger pump with radiators in parallel if you have problems with cooling your current CPU.

[DigitalChaos]

i think he meant: put the output hose on each pointing to the intake of the opposite pump...

[Xel]

Yeah. That way, as long as the radiator could keep up with the amount of heat the waterblock was making, there shouldn't be any problems. In fact, if you -really- wanted it cooled, you could put another radiator and pump in.

[DarkEdge]

I never considered this option. Using seperate pumps to cool diffrent parts of the system. The downside is that you couldn't do a closed loop. With a res. system though you could do it. As of right now I can't try it. I gutted my system out and am back on air cooling. I'm looking for a way to cool more efficiently, also waiting on AxionTech to get stock in of the Danger Den products. 38 dollars for a maze 2!

[Xel]

I wish I could try it too but right now I'm waiting for my w/c stuff to get here =[

[Brad]

you'd be better to spend the extra money and room it'd take for a second pump on getting a bigger radiator I think

[DarkEdge]

Don't know. Hard to tell without testing. I did see something interesting in the HardOCP forum tonight. I saw to 802 power heads in a res. One pushing and one pulling.

Whats great about them is they are both 400gph pumps. They can have there flow reversed with a flick of a switch. I have 2 in my 150 gallon fish tank and they just simply rock. They have to pull water up almost 2 feet in my tank and there pushing alot of water. Atleast 100gph. Hard to tell with out a flow meter, but its not below 100.

I may go buy two more. Can't take them out of my fish tank or I would loose alot of money from my fish dying. Either way it may work well, but like anything in life its all about trial and error.

[Brad]

I thought the fish wouldn't mind that much if they had no pumps for an hour or two

[jaydee]

Well, my water temp is 2-3C over room temp so why spend all the money and do all that extra work to get it closer to room temp? You cannot get below room temp without some colder source like a pelt.

This idea would work much better for a pelt run rig I think though.

Anyway you do it though you will still need the output of the radiator going either as fast or faster than the output of the waterblock or the water would heat up because it would be recirculating the warmer CPU warmed water faster than the cooler rad water.

[mkosem]

this would work

--MAtt

[futRtrubL]

Sorry that won't work, if the flow rates are different you will get mixing of cool and hot water, this will send warm water to the CPU. Chech out the following thread, all this was discused there.
http://forums.procooling.com/vbb/sho...=&threadid=709


If you have different flow rates then there will be mixing.A large flow into the res. can't turn into a small flow out to the rad,some of that flow will have to go out to the cpu. If both flows are the same then why have them going through the same res. and risk mixing by turbulence?

Edward

[Volenti]

I thought the fish wouldn't mind that much if they had no pumps for an hour or two

If the pumps are hooked up to a well established undergravel filter stopping the flow will after a time kill most of the aerobic bacteria that have established themselves in the filter media.

Then they are replaced with an-aerobic bacteria, which can release toxins into the water when the flow starts up again.

May not happen in an our or so, but I wouldn't want to risk expensive fish in the expriment

[grep]

if the pump to the water block has 1/2" diameter hose and passes water at say, 100 gph and the pump to the radiator has 1" diameter hose and pumps at 100 gph the volumn of water in each loop will be the same. However the flowrate through the radiator will be at 1/2 the speed of the waterblock loop. You get the effect you want and the speed differential you want.

[futRtrubL]

Unfortunately hose diameter doesn't affect the bore of the rad itself so you still get the same water velocity in the rad as before (it'll be slower in the tubing but that doesn't help).

Edward