Do I need another pump?

[davidm71]

Hi,

About to setup my next water cooled build and confused why or if I'll need a second pump. I have three radiators I'm planning on using chained together to cool a 5820K and a single gpu (maybe two if I can afford it). My pump is a Koolance PMP-500 and using 7/16 Primoflex. Loop order is fill port -> res -> pump -> rad 1 (alphacool 240mmx60mm) -> rad 2 (Hardware Labs SR-2 280mm) -> rad 3 (Hardware labs Nemesis GTS X-Flow) -> CPU Block (Heatkiller IV) -> GPU 1 (980 ti hopefully -not purchased yet) -> Back to res.

Any ideas?

Thanks

 

[aigomorla]

who told you that u would need a second pump for a pmp-500?

The pmp-500 is a monster pump, it should have no problems with restriction or flow.
Infact, i would instead upgrade your tubing diameter to 1/2 instead of 7/16, as the inlet on the pmp-500 may cavitate if the line is too long from the inlet.
(i have seen this happen on my rd-30 with 7/16ID tubing at full blast. The pmp-500 is almost near spec to a RD-30. )
when it cavitates, it collapses the inlet tubing on the pump.... this is why a 1/2ID 3/4OD is preferred as it has a thicker wall to help with cavitation.

If you want to add a second pump, its best to pair them off, and at your point it would only be for redundancy.

 

[davidm71]

I just have seen certain people on the net use more than one pump. As far as cavitation goes I never heard of that. I have already spent a lot of money on compression fittings so I can't just go get 1/2 inch tubing. I wish I know before hand. Maybe I could lessen the volts to the pump? Is it really that much of a risk?

thanks.

Edit: Maybe if I run 1/2 Inch tube just to the inlet from the resevoir to the pump's inlet? Its about 6-8 inches of length. Still a concern?

 

[aigomorla]

just keep your inlet as short as possible...
Inlet is the tubing that goes directly into the in port on your pump.

https://en.wikipedia.org/wiki/Cavitation

keeping a shorter intake... means higher resistance to tube collapsing from lack of pressure..... This isnt a problem on a pmp-400 or a pmp-450..... a pmp-500 is a whole other beast tho....


2nd pumps is mostly for redundancy incase one dies.
The only time a second pump is really recommended is if you have a GANG of restrictions, which your setup does not, or your using a injector cpu block, which EATS flow, which rarely are they made now as the gains in performance on them are not worth the extra eq. to facilitate it.

So a second pump under most conditions is redundancy, or you just to go all out and have a multi pumps.... They do help you later on split your loop, however we dont have 1000W from the GPU anymore, unless ur running multi titans X in SLI. Then it should be a no brainer to split the loop up, as you have invested more then 2000 dollars in your GPU's. (however this is also debatable...)

 

[davidm71]

just keep your inlet as short as possible...
Inlet is the tubing that goes directly into the in port on your pump.

https://en.wikipedia.org/wiki/Cavitation

keeping a shorter intake... means higher resistance to tube collapsing from lack of pressure..... This isnt a problem on a pmp-400 or a pmp-450..... a pmp-500 is a whole other beast tho....


2nd pumps is mostly for redundancy incase one dies.
The only time a second pump is really recommended is if you have a GANG of restrictions, which your setup does not, or your using a injector cpu block, which EATS flow, which rarely are they made now as the gains in performance on them are not worth the extra eq. to facilitate it.

So a second pump under most conditions is redundancy, or you just to go all out and have a multi pumps.... They do help you later on split your loop, however we dont have 1000W from the GPU anymore, unless ur running multi titans X in SLI. Then it should be a no brainer to split the loop up, as you have invested more then 2000 dollars in your GPU's. (however this is also debatable...)

Well how about running 1/2" tube from res to pump and leave the 7/16 everwhere else? Shouldn't cost much to just buy two compression fittings and ten inches of 1/2 inch tube? Anyhow that pmp 500 should be able to handle two titans or 980 ti's in sli if it ever happened? For now sticking with just one gpu.

Thanks.

 

[BonzaiDuck]

just keep your inlet as short as possible...
Inlet is the tubing that goes directly into the in port on your pump.

https://en.wikipedia.org/wiki/Cavitation

keeping a shorter intake... means higher resistance to tube collapsing from lack of pressure..... This isnt a problem on a pmp-400 or a pmp-450..... a pmp-500 is a whole other beast tho....


2nd pumps is mostly for redundancy incase one dies.
The only time a second pump is really recommended is if you have a GANG of restrictions, which your setup does not, or your using a injector cpu block, which EATS flow, which rarely are they made now as the gains in performance on them are not worth the extra eq. to facilitate it.

So a second pump under most conditions is redundancy, or you just to go all out and have a multi pumps.... They do help you later on split your loop, however we dont have 1000W from the GPU anymore, unless ur running multi titans X in SLI. Then it should be a no brainer to split the loop up, as you have invested more then 2000 dollars in your GPU's. (however this is also debatable...)

Debatable, perhaps, but logical as an extension of that investment, knowing how much use and salvage-value you'd get from the Titans. And that would hold for the additional parts expense, as well.

 

[davidm71]

Debatable, perhaps, but logical as an extension of that investment, knowing how much use and salvage-value you'd get from the Titans. And that would hold for the additional parts expense, as well.

Well it's going to be a basic loop, three rads, cpu block, hopefully mosfets if there's a compatible block, then a single gpu. Love to go Sli but costs would be too high. Thanks

Ps: still good idea run just 1/2 inch to the pump from res and leave rest at 7/16??

 

[Rubycon]

Increasing the tubing internal diameter between the receiver and the pump's inlet isn't going to matter. Cavitation risk is highest when pump head is low and there is a restriction on the suction side. With proper plumbing this should never happen. At the temps (pc) coolant is running cavitation isn't going to happen even if the suction line is running 10" Hg vacuum! But don't allow this to happen; the pump will still be unhappy and you will be unhappy due to the noises.

Your scenario isn't a complex one and even though 1/2" ID tubing throughout would be better, 7/16" should not present any issues that would be alarming in nature.

Serial mounted pumps can provide redundancy and increased delivery at higher head pressures but are certainly not necessary with your setup.

 

[davidm71]

Increasing the tubing internal diameter between the receiver and the pump's inlet isn't going to matter. Cavitation risk is highest when pump head is low and there is a restriction on the suction side. With proper plumbing this should never happen. At the temps (pc) coolant is running cavitation isn't going to happen even if the suction line is running 10" Hg vacuum! But don't allow this to happen; the pump will still be unhappy and you will be unhappy due to the noises.

Your scenario isn't a complex one and even though 1/2" ID tubing throughout would be better, 7/16" should not present any issues that would be alarming in nature.

Serial mounted pumps can provide redundancy and increased delivery at higher head pressures but are certainly not necessary with your setup.

I'm sorry I dont understand. Your saying risk is low and running 1/2 inch from res to pumps intake wont matter? Whats proper plumbing? No kinks you mean?

 

[aigomorla]

I'm sorry I dont understand. Your saying risk is low and running 1/2 inch from res to pumps intake wont matter? Whats proper plumbing? No kinks you mean?

she means flow is uniform... so it goes based on the greatest restriction.

So having a 7/16ID outlet is still going to cause back log pressure when the pump is pushing the water from a starved inlet.

Thicker wall tubing is really the only way to combat a collapsed tube.


And yes u'll be OK with the inlet being 1/2ID, but u probably wont get any net benifit from going larger diameter.

 

[davidm71]

she means flow is uniform... so it goes based on the greatest restriction.

So having a 7/16ID outlet is still going to cause back log pressure when the pump is pushing the water from a starved inlet.

Thicker wall tubing is really the only way to combat a collapsed tube.


And yes u'll be OK with the inlet being 1/2ID, but u probably wont get any net benifit from going larger diameter.

So maybe I should turn down the voltage?

 

[aigomorla]

So maybe I should turn down the voltage?

You can...

because im fairly sure at the point ur at, you way over 1.5gpm....
Where net gains are VERY VERY TINY.

 

[davidm71]

Hi,
Its been a year since I started this thread and got to tell you I never got it off the ground. Too afraid of that cavitation issue. Anyhow since then I picked up a second gpu and ready to get this off of the back burner. I found out that Koolance makes a pump speed controller which I'm going to get and possibly a new resevoir that isnt so tall. Not sure about using 1/2 inch tubing though but will use your advice and keep tube length short at the inlet if I can.

Thanks.

 

[Rubycon]

Thicker wall tubing is really the only way to combat a collapsed tube.

Lower radiator hoses on liquid cooled internal combustion engines use a steel spring inside the hose to prevent collapse. It may be awkward to find something in the size needed for this application and have it of suitable material (316 SS) but nonetheless it will keep the hose round even in an impeller eroding 27" Hg vacuum!

 

[davidm71]

That's interesting


Sent from my iPhone using Tapatalk