Doubling D5 for flow?

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208
If they are placed in series you increase pressure - not flow rate. The volumetric flow rate must be constant - if its not you will starve the latter pump - and could wear out the bearing.
Ideally both pumps need to run at same speed, and have same flow rate.

He can only benefit from 2 pumps with variable speeds in parallel config. i.e. 2 pumps separately connected to rez, then pumps connected to the loop with a single T junction connection to the rest of the system.

Thanks for repeating what I have been writing?

And yer wrong, you can run either D5 or DDC variable or not in series or parallel. And stop with different pumps and crap may cause damage BS. These are not geared pumps.
 

cyklondx

Limp Gawd
Joined
Mar 19, 2018
Messages
419
Thanks for repeating what I have been writing?

And yer wrong, you can run either D5 or DDC variable or not in series or parallel. And stop with different pumps and crap may cause damage BS. These are not geared pumps.

go ahead and put one slower pump before the faster one in series, enjoy the sound. Just do it, make your dreams come tru.
 

cyklondx

Limp Gawd
Joined
Mar 19, 2018
Messages
419
Noise could be created from use two exact pumps as well. Again wtf does that have to do with your misinformation?
the noise i'm ref. to would be the bearing wearing out.

I stated "The volumetric flow rate must be constant" with variable speeds on pumps where one pump is slower than the other it pumps fluid into.

But hey reading whole thread - conversation might help.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208
the noise i'm ref. to would be the bearing wearing out.

I stated "The volumetric flow rate must be constant" with variable speeds on pumps where one pump is slower than the other it pumps fluid into.

Dude, you really have no freaking clue.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208

Stop googling shit that isn't specific to these loops.

https://martinsliquidlab.wordpress.com/2011/04/26/pump-setup-series-vs-parallel/
The result was somewhat as expected (Very Poor), but there are some interesting oddities. When you run two pumps in parallel, the curves somewhat get stacked in the X or flow rate direction with an averaging of pressure. Unfortunately this means the real gains of parallel don’t happen until you get beyond the useful restriction range of water cooling loops. On a very high restriction loop you actually don’t gain anything at all. The parallel curve crosses the single DDC curve at about the same point. On a low restriction setup you would see some gains, roughly a 22% increase in pressure, but not at all what you might have hoped for. Parallel simply doesn’t show benefits for the higher restriction levels that we typically see in water cooling and that holds true for pretty much any typical water cooling pump.


As you can see, it really doesn’t matter much which pump, the result simply doesn’t favor parallel pumps.


Beside the expected poor results, it was interesting that the parallel result is not simply adding the X direction as I thought. Max head pressure for example is not the 6.5PSI of the DDC, instead it is an average of the two pumps(6.1PSI). In the end it’s more complex than just adding the X direction, but the bottom line is that it’s not good and definitely not recommended. Parallel pump performance is poor for water cooling.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208
It is specific to all pump setups... please stop posting you have no clue.

This article you posted, is referred to something completely else.
It compares single pump vs parallel ~ they refer to loop in series not pumps in series...

Omfg, what is your issue? Too much idiocy for me...

Just get bigger rad or reservoir, or faster pump...

2nd pump won't do anything, and won't speed things up - may even cause problems, or damage them.
 
Joined
Apr 5, 2016
Messages
2,195
I can see that being the case with a straight thru pump setup, like in most of the industrial scenarios that your link is referring to. But in our scenario, I don't think that's the case.

Pump flow increases as resistance decreases. Your first pump in the series is going to see essentially zero resistance - the second pump is consuming water as fast as the first one can produce it. So even if your first pump is rated at 3gpm and your second is rated at 5, when put up against the restriction of a cooling loop, the second pump will only manage 1 due to the restriction, while the first, seeing no restriction, will run much closer to it's rated 3.

These pumps are tiny, our pressures are tiny, our velocities are tiny. We aren't going to see cavitation, bearing thrust or any of the other things that plague industrial size centrifugal pumps.
 

cyklondx

Limp Gawd
Joined
Mar 19, 2018
Messages
419
I can see that being the case with a straight thru pump setup, like in most of the industrial scenarios that your link is referring to. But in our scenario, I don't think that's the case.

Pump flow increases as resistance decreases. Your first pump in the series is going to see essentially zero resistance - the second pump is consuming water as fast as the first one can produce it. These pumps are tiny, our pressures are tiny, our velocities are tiny. We aren't going to see cavitation, bearing thrust or any of the other things that plague industrial size centrifugal pumps.

and I agreed with that, to a point.

The cavitation will happen, its just a question of time. Short tests won't show it - but after running it for couple weeks 24/7 you'll see it.

so i hope sir thesmokingman tries it, so that i could enjoy myself.
 
Last edited:

Tsumi

[H]F Junkie
Joined
Mar 18, 2010
Messages
13,380
and I agreed with that, to a point.

The cavitation will happen, its just a question of time. Short tests won't show it - but after running it for couple weeks 24/7 you'll see it.

so i hope sir thesmokingman tries it, so that i could enjoy myself.

Multiple people have been running dual pumps in series for years now. Watercooling and dual pump tops aren't new, they've been around for over 10 years. If it was as big of a problem as you are claiming, we would know it by now. You are the one that is completely clueless and bringing in completely irrelevant scenarios.
 

cyklondx

Limp Gawd
Joined
Mar 19, 2018
Messages
419
its fine that we don't agree. We don't have to.

from what i know
Most people turn off their systems at least once a week which definitly is going to help; I'm more close of running datacenter on water with no downtime for months - and seen the issue.
 

Tsumi

[H]F Junkie
Joined
Mar 18, 2010
Messages
13,380
its fine that we don't agree. We don't have to.

from what i know
Most people turn off their systems at least once a week which definitly is going to help; I'm more close of running datacenter on water with no downtime for months - and seen the issue.

Proof or it didn't happen. And make it relevant to our scenarios, nothing with industrial pumps running at 20+ gpm.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208
and I agreed with that, to a point.

The cavitation will happen, its just a question of time. Short tests won't show it - but after running it for couple weeks 24/7 you'll see it.

so i hope sir thesmokingman tries it, so that i could enjoy myself.

I've ran multi pumps for years in the past w/o issue. You're seriously borderline trolling now.
 

cyklondx

Limp Gawd
Joined
Mar 19, 2018
Messages
419
Proof or it didn't happen. And make it relevant to our scenarios, nothing with industrial pumps running at 20+ gpm.


i'll try to record something once i'm home ~ i should have some old pumps laying around, not sure about piping tho ~, we'll see when we get there.
 

Zarathustra[H]

Fully [H]
Joined
Oct 29, 2000
Messages
30,802
I've ran multi pumps for years in the past w/o issue. You're seriously borderline trolling now.

How would you know if you have cavitation?

Slightly more noise?

I'd imagine we have cavitation in our pumps all the time, but that it is at a small and mostly irrelevant level.
 

Zarathustra[H]

Fully [H]
Joined
Oct 29, 2000
Messages
30,802
If you have witnessed cavitation you'd know it. It's hard to miss.


I think I've seen stuff like that right after fill, but once the air works its way out of the loop, I e never seen it.

I'd imagine our comparably slow, quiet and low power home pumps (originally designed for small home acquariums) probably are not fast enough to cause that kind of cavitation on their own.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208
I think I've seen stuff like that right after fill, but once the air works its way out of the loop, I e never seen it.

I'd imagine our comparably slow, quiet and low power home pumps (originally designed for small home acquariums) probably are not fast enough to cause that kind of cavitation on their own.

That's not cavitation though. On fill up that's just air bubbles trapped that are loose in the loop. Those loose bubbles get chopped up by the pump. Cavitation is completely different as explained in the video.
 

Zarathustra[H]

Fully [H]
Joined
Oct 29, 2000
Messages
30,802
That's not cavitation though. On fill up that's just air bubbles trapped that are loose in the loop. Those loose bubbles get chopped up by the pump. Cavitation is completely different as explained in the video.

Yep, understood, symptoms are similar though.
 
Joined
Apr 5, 2016
Messages
2,195
I don't believe our pumps are strong enough to cause cavitation, and I'm positive they aren't strong enough to withstand it. The instantaneous pressure spikes caused by collapsing cavitation bubbles would immediately destroy the plastic and ceramic parts in our pumps. It tears pits and holes in carbon-steel impellers and volutes in industrial pumps.
 

Zarathustra[H]

Fully [H]
Joined
Oct 29, 2000
Messages
30,802
Alright,

So I am finally at the point where I can speak from experience on this topic.

I am in the process of building a system with an EK Dual D5 pump top (pumps are in series)

It's not completed yet, but I just did some testing with the loop running and my aquaero (and my optional Aquaero flow meter) hooked up to my laptop.

Build is here.

With one pump running at 100%, I get a flow rate of 1.1 GPM.

With both pumps running I get a flow rate of 1.6 GPM.

It doesnt matter if the first or the second pump is the one running, the 1.1 stays the same.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208
Out of curiosity, what happens if they run in parallel?

Flow would go up but head pressure would stay the same. Flow is tricky with D5 pumps in that it is kind of a misnomer. D5 has crazy flow when there is no restriction, like on paper. However soon as you add restriction, they have rather weak head pressure, ie. the ability to push fluid thru a restriction.

With one pump running at 100%, I get a flow rate of 1.1 GPM.

With both pumps running I get a flow rate of 1.6 GPM.

The above is a measure of head pressure not flow. The flow is indicative of the head pressure with the current flow restriction.
 
Joined
Apr 5, 2016
Messages
2,195
The above is a measure of head pressure not flow. The flow is indicative of the head pressure with the current flow restriction.
No, it is not. The units are in gal/m, it is a measure of flow.

Now - obviously - those flow rates are partially a product of the pressure drop through Zarathustra[H] 's loop, but since the entire premise of this thread is to ask whether doubling D5s appreciably increases flow rate, I'd say his answer was about as concise and relevant as an answer could be.
 

thesmokingman

Supreme [H]ardness
Joined
Nov 22, 2008
Messages
6,208
No, it is not. The units are in gal/m, it is a measure of flow.

Now - obviously - those flow rates are partially a product of the pressure drop through Zarathustra[H] 's loop, but since the entire premise of this thread is to ask whether doubling D5s appreciably increases flow rate, I'd say his answer was about as concise and relevant as an answer could be.

That's what I tried to point out. It's flow but it's misleading because its really about head pressure.
 

Zarathustra[H]

Fully [H]
Joined
Oct 29, 2000
Messages
30,802
The above is a measure of head pressure not flow. The flow is indicative of the head pressure with the current flow restriction.

Well, of course.

Flow and head pressure are intrinsically related. You can't have one without the other.

It's exactly the same principle as airflow and static pressure. Neither can exist without the other. Or electrical voltage and current.

Or even a force on a physical object and that objects velocity.

The head pressure is what creates any flow in any loop. The head pressure is the force that is pushing the water, the flow is the speed it winds up traveling at because of that pressure.

It's kind of silly to speak about either head pressure or flow in isolation of eachother. They are intrinsically interrelated.
 
Joined
Apr 5, 2016
Messages
2,195
That's what I tried to point out. It's flow but it's misleading because its really about head pressure.
Ah. I see. So even though there is an identical level of restriction in both cases, because it's literally the same loop with either one pump or both, the 0.5 gal/m delta is a useless figure because "it's about head pressure."
 

newls1

Supreme [H]ardness
Joined
Sep 8, 2003
Messages
4,338

Zarathustra[H]

Fully [H]
Joined
Oct 29, 2000
Messages
30,802
this is about the biggest BS reply yet on this forum.. wow! OP, I have this in my CPU only loop for the 3950x, and I dropped 2-3c in temps idle, and 3-4c on full load. Loop was CPU only with 1 420mm rad and a 240mm rad ( https://www.ekwb.com/shop/ek-xtop-revo-dual-d5-pwm-serial-incl-pump )

Damage is highly unlikely. You'd need some spectacularly high head pressure to do damage.

That said going with dual pumps won't always help with temps

At some point you run into limiting returns on how much higher flow will help temps.

At the same time, adding more D5 pumps means more heat dumped into the loop from those pumps, and higher pressure results in more friction in the loop also resulting in heat.

At some point if you keep adding pumps, you will reach the point where the next pump adds more heat to the loop than it helps lower temps.

Where this point is will depend on how restrictive the loop is. With a less restrictive loop you'll likely get there more quickly than with a more restrictive loop.

So, multiple pumps is not a silver bullet, but if your block+radiator count is greater than 4-5 somewhere, you'll probably benefit from an additional pump, but the benefit will likely be smallish until you get more restrictive.
 
Joined
Apr 5, 2016
Messages
2,195
Finally finished my upgrade, and can report on this.

Old loop: MO-RA3 420, Heatkiller Tube D5 reservoir. Koolance QD3s into case. Heatkiller IV Pro Acetal. (No GPU block at the moment as I'm waiting on the 3090.)

New loop: the same, except instead of the old Tube D5, I'm using a shorter Tube res with Watercool's newly released Dual D5 top.

Flows are recorded at 100% pump speed with an Aquacomputer MPS Flow 400 inline flow meter.

OLD FLOW: 1 GALLON PER MINUTE

NEW FLOW: 1.35 GALLONS PER MINUTE

Honestly I'm a bit disappointed. I had to do a redo a lot of custom wiring to make this happen, and I'm happy with the end result, but I think my loop may be too restrictive with the giant rad and QDCs to see much benefit to my flow rate.

One thing that remains to be seen is if that number changes once I parallel a GPU block back into the mix. I'll report back then. For now though, it will be nice to be able to maintain 1GPM at lower speeds than I was able to before, and of course the redundancy is nice.
 

NightReaver

Limp Gawd
Joined
Apr 20, 2017
Messages
188
Finally finished my upgrade, and can report on this.

Old loop: MO-RA3 420, Heatkiller Tube D5 reservoir. Koolance QD3s into case. Heatkiller IV Pro Acetal. (No GPU block at the moment as I'm waiting on the 3090.)

New loop: the same, except instead of the old Tube D5, I'm using a shorter Tube res with Watercool's newly released Dual D5 top.

Flows are recorded at 100% pump speed with an Aquacomputer MPS Flow 400 inline flow meter.

OLD FLOW: 1 GALLON PER MINUTE

NEW FLOW: 1.35 GALLONS PER MINUTE

Honestly I'm a bit disappointed. I had to do a redo a lot of custom wiring to make this happen, and I'm happy with the end result, but I think my loop may be too restrictive with the giant rad and QDCs to see much benefit to my flow rate.

One thing that remains to be seen is if that number changes once I parallel a GPU block back into the mix. I'll report back then. For now though, it will be nice to be able to maintain 1GPM at lower speeds than I was able to before, and of course the redundancy is nice.
This reminds me that I should grab a flow meter at some point. I have an aquasuite unit, but it's only controlling a fan curve atm. It would be nice to add pump control into the mix.
 
Joined
Apr 5, 2016
Messages
2,195
Interestingly, going from a CPU-only loop to a CPU/GPU loop in parallel increased my flow rate by ~20%. Pardon the unit change from previous posts, but I went from 300 l/h to 360 l/h with both pumps at 100%.
 

NightReaver

Limp Gawd
Joined
Apr 20, 2017
Messages
188
Interestingly, going from a CPU-only loop to a CPU/GPU loop in parallel increased my flow rate by ~20%. Pardon the unit change from previous posts, but I went from 300 l/h to 360 l/h with both pumps at 100%.
Could it be that the gpu block is less restrictive and more flow is going through that instead of the cpu block?
 
Joined
Apr 5, 2016
Messages
2,195
Could it be that the gpu block is less restrictive and more flow is going through that instead of the cpu block?
Any two blocks in parallel will have less restriction than either single block. The flow splits between them at a ratio defined by their relative restriction.

I expected an increase, but certainly not that big a one. Would be curious to see the flow if I temporarily cut the CPU block out and do GPU only, but that's likely more work than I feel like doing. XD
 
Top