Difference betweem 1 or 2 pumps?

[wuzzapiman]

What is the differce between:

lets say all pumps have a 4' head
and 6' of tubing in the system

2x 150Gph vs. 1x 150Gph

2x 150Gph vs. 1x 300Gph

 

[Skulemate]

Well, that all depends on how the pumps are incorporated into the system. If they are connected in series, it's simply that for a given flow from the two pumps the pressure would be twice what one pump alone could produce (let's assume for simplicity that the pumps are identical). However, if the pumps are plumbed in parallel then for a given amount of pressure produced the flow rate is doubled. (Said another way, for a series system pressures are added for like flows, as opposed to flows added for like pressures in a parallel system.)

So, to clarify a little, consider your first situation. If the two pumps were plumbed in parallel, then you'd have a system with a maximum flow of 300 gph (double that of a single pump) and a shutoff pressure of 4' (same as for one pump). Alternatively, if the two pumps were in series, then you'd have a system with a maximum flow of 150 gph (same as one pump) or a shutoff pressure of 8' (double that of one pump).

There's one last thing I'll say. For the majority of watercooling systems, it's usually beneficial to use pumps set up in series (ask Volenti) because this increases the amount of pressure that can be developed, which will improve the amount of flow through the system. Increasing flow without a substantial increase in pressure is generally wasted due to flow restrictions (from tubing, pressure drop through the block, etc). Hope this answers your question.

 

[Paxmax]

What is the differce between:

lets say all pumps have a 4' head
and 6' of tubing in the system

2x 150Gph vs. 1x 150Gph

2x 150Gph vs. 1x 300Gph

First right off: I don't think that there is a substancial gain between running 1 pump and 2 identical pumps, except safety.
Maybe if you have a really restricted system a series connection would help somewhat.

Getting a NEW pump with higher head is the way to go, OR lowering the restrictions in your system.
I had a high restriction system due to my rads loooong series of tubes of only 1/4 inch inner diameter. I used a pond type centrifugal pump, like Ehiem with a max flow of 315 GPH, maxhead of 8 feet. I got like 24gph !! I talked to the distibutor of the pumps asking about a second pump. He claimed that it doesn't work that good, it isn't that simple. So I used 2 rads in parallell, and each rad was split in half, so the water had 4 paths to go. With that mod I got flow to 104 gph. All measurement via the "fill da bucket method" Just to be safe I bought another bigger pump aswell, with increased head.

Another reason for 2 pumps not working too good is all experiments with 2 fans stacked, that hasn't brought much good, I can't remember a single positive review. Maybe comparing air/water is a bad example.

On a more physical level:

You have 1 pump with a maxhead of 4' max flow 315 gph.
For ordinary centrifugal pumps the max flow and max head occures at different occasions in a nonlinear fashion.

You have 6' tubing, a block, a rad. It has a backpressure of x psi at a flow of 100 gph.

The pump max cap isn't reached because of the backpressure. It balances out at 100 gph through your system mentioned above.

Then you add a identical pump in parallell:
The flow will increase some due to the flow being halved through each pump. With halved flow the pumps head capability will increase. By how much? I don't know.

You add a pump in series:
Okay, theoretically this should mean that we have double max head at our disposal. As the pumps see only half of the resistance each, we see a increase of flowrate. As the flow through the pumps increases the head capability decreases. Since the head capability decreases, the ability to fight the higher backpressure(a result of forcing water quicker through the system) also decreases.
Again, a smaller result than expected.

I think this is nasty. But then again, this is just my theories. Maybe some ppl have have been successful in using 2 pumps.

 

[Diggrr]

I used pumps in series before, trying to lower my temps.
I have a Danner Supreme 250gph and a ViaAqua.
I set them up on opposite sides of the system so it went pump>radiator>pump>cpu>northbridge>gpu (mind you, my tubing is about 15-20 feet total) and adding the second pump gained me 30% flow rate over a single pump, no matter which pump I was using. I think the increase was due to pressure added by the second pump.

Anyway, I gained nothing in system temps, so I went back to a single pump. Your mileage may vary, if your system is starving for more flow you'll see a difference. It's all in tuning your waterflow to your particular waterblock/setup. Some need high flow, some don't. I use a low flow Gemini block.

Oh yeah, personally, I'd just get the single 300 gph pump. It's more versatile in the long run. You'll probably have the pump for a long time more than you'll have the computer. The larger pump will run better, and be more usefull in whatever type of system the future holds. I've had my Danner for 3 years now, and it's really seen the full spectrum of what watercooling can be like (bongs, internal, external, relayed shutdown, 24/7 operation etc...).

 

[Skulemate]

... personally, I'd just get the single 300 gph pump. It's more versatile in the long run. ...

It's not really that simple. wuzzapiman asked about pumps that all have the same head, namely 6'. When dealing with two similar pumps with the same shutoff pressure it is seldom a good idea to take the higher flowing pump, as the increase in flow will simply cause there to be more flow resistance... the pump will operate much less efficiently (of course, some common sense is needed to deal with this, as is an idea of the magnitude of restrictions in your system, which is probablly not something that most understand). There really are no simple rules concerning pump selection. One suggestion though if you are intent on using more than one pump: don't mix brands, as this will complicate things slightly (i.e. would like identical specs for simplicity's sake).

 

[Paxmax]

Anyway, I gained nothing in system temps, so I went back to a single pump.

Strange, you gain about 30% in flow but no better performance.
Maybe the added pumpheat ate all gain in flow!
Arrgh... again thoose nasty centrifugal pumpers messes us overclockers up.
Maybe I should visit mom, and have grab her well pump, pure piston power!! POSITIVE DISPLACEMENT !!

LOL! Diggrr, what do you do with all the numerous feets of hose?!?

 

[Diggrr]

I don't think I got better performance because I have a Gemini low flow block. Pushing more water through wasn't really expected to gain better performance, but I can hope.

And all that tubing goes to my heatercore. It's in the basement under my desk...absolutely noise free and cool all summer long in that chilly stone walled cellar.
Pump head wasn't really the issue, because though I have a reservior mounted on the wall, it's sealed tightly.

And the reason I'd recommend the 300 gph pump, is that a 150, with all the flow restrictions in a system, wouldn't power a bong should he decide to do so in the future. Usually, one can expect actual flow rates as low as 25% of rated flow or less (at least that's my experience). The larger flow rated pump would be more versatile in future setups that he may not have even thought of yet.
And though his "exercise" pointed to two pumps of differing flow rates and the same head (shutoff) height, that's not always the case within the same style of pump. Take two Danner's for example, the 250's head is 6' and the 350's head is 10' (their 150 version is a different pump body and impeller setup, so the comparrison is apples/oranges).

Flow restriction is not as understood as it should be, I'll agree. But that's not something that's easily taught on a web forum, hence is up to the user's self-teaching (and understanding what they've learned).

Have a great one, gents!

 

[Volenti]

I'm currently working on a major pump mod that will incorporate 3 pumps in series (to feed my pressure hungry direct die cooling)

I'll post in detail when I'm finished, meanwhile, here's a teaser pic