Cavitation/Vibration central - RP-452X2 V2 /w 2 Pumps in Series

[D749]

Has anyone else with a similar configuration had noticed this issue? Basically with both pumps on at their lowest setting I notice an oscillating vibration. Please see my signature for system hardware details.

I finally said screw it and killed the power to the second pump and while my flow drop 20 gal/h the cavitation went away. Even with the drop in flow rate I haven't noticed any impact on cooling. Granted, I could always increase the single pump speed using my controller and I imagine that if I remove the second non-powered pump flow would increase.

[Stewie007]

Any time you use pumps in series, it is not advisable to put them so close in series. Any time you use such a system, you are doing so to increase the amount of head that the system can handle. I can't see any reason for you to use two pumps, unless the head pressure is so high that the first pump is struggling. It appears that it is not, based on your observation.

I understand the need to have "more power," but have myself learned that there are limits. If you have a pump that can handle... say.... 20 feet of head, and then add a second pump of the same design... your system will then handle 40 feet of head. However, if your system only provides a few feet of head, then you aren't benefiting. You're essentially trying to enhance performance at the top of the pump curve, which really has a small net effect.

If you want to continue using this pumping method, move the second pump down stream after the first restriction, perhaps. Don't feed it with a U-bend, or right in series. The pump needs to have effective suction.

[D749]

Any time you use pumps in series, it is not advisable to put them so close in series. Any time you use such a system, you are doing so to increase the amount of head that the system can handle. I can't see any reason for you to use two pumps, unless the head pressure is so high that the first pump is struggling. It appears that it is not, based on your observation.

I understand the need to have "more power," but have myself learned that there are limits. If you have a pump that can handle... say.... 20 feet of head, and then add a second pump of the same design... your system will then handle 40 feet of head. However, if your system only provides a few feet of head, then you aren't benefiting. You're essentially trying to enhance performance at the top of the pump curve, which really has a small net effect.

If you want to continue using this pumping method, move the second pump down stream after the first restriction, perhaps. Don't feed it with a U-bend, or right in series. The pump needs to have effective suction.

Good feedback - thanks. I tend to go overboard when I build a system and I see now I did so in terms of number of pumps. Turning off the second pump dropped the flow down to ~50 gal/h from ~70 gal/h. So I wasn't receiving a huge benefit from the second pump in terms of flow. Having redundancy was nice but I also monitor the pump RPM via my AQ5 Pro.

[Stewie007]

That setup seems to be giving you a lot of restriction, huh.

A 28% drop in flow seems pretty significant. Obviously more head pressure than I anticipated. Does that drop in flow equate to much of a performance drop in your blocks?

Edit: Also, how are the outlets fed? i.e. Where is the reservoir in this system? Above, below, or level with the pumps? I can't tell in the image.

[NKrader]

That setup seems to be giving you a lot of restriction, huh.

A 28% drop in flow seems pretty significant.

also you forgetting the restriction that the second pump not running but being in the way of the flow is adding..

[D749]

That setup seems to be giving you a lot of restriction, huh.

A 28% drop in flow seems pretty significant. Obviously more head pressure than I anticipated. Does that drop in flow equate to much of a performance drop in your blocks?

Edit: Also, how are the outlets fed? i.e. Where is the reservoir in this system? Above, below, or level with the pumps? I can't tell in the image.

The Koolance RP-452X2 V2 serves as both a pump housing and reservoir. Water enters the port with the 90 degree fitting (in the picture) and then exits through the fitting attached to the flow meter (also in the picture). I've done some casual testing with benchmarks and games and I haven't noticed any issues with this flow drop.

Overall my loop is probably more restrictive then most as I like to use QDC fittings. However, I use the Koolance VL4N (least restrictive) to minimize the restriction as much as possible.

also you forgetting the restriction that the second pump not running but being in the way of the flow is adding..

I plan to remove the 2nd pump once I decide on my next CF/SLI setup (tri-290Xs or tri-780 TIs). It will be interesting to see how much the flow increases with the 2nd pump removed.

[Stewie007]

also you forgetting the restriction that the second pump not running but being in the way of the flow is adding..

Not really. The impeller's inside those pumps aren't going to restrict flow as much as a water block. You can't attribute the drop in flow completely to the pump because the system as a whole is restrictive and the second pump gave a boost to the head pressure that the system could handle. That is simply how it works. I would expect the pump restriction to be minimal compared to a 4 restrictive water blocks.

[NKrader]

Not really. The impeller's inside those pumps aren't going to restrict flow as much as a water block.

It will be interesting to see how much the flow increases with the 2nd pump removed.

bet you more than you would think. or stewie thinks

[Stewie007]

I think I have a handle on your problem D.

However, which pump are you activating/deactivating? Is it the first or second pump in your sequence?

[Stewie007]

bet you more than you would think. or stewie thinks

Well, part of my problem is I didn't have all of the details on his equipment. And I think you and I have already agreed that he needs to remove the second pump if he isn't using it. So no need to start a debate when the recommendation is clear. I'm just trying to understand a few details regarding what he actually wants to do before I make a recommendation to perhaps help him increase his flow overall.

[D749]

I think I have a handle on your problem D.

However, which pump are you activating/deactivating? Is it the first or second pump in your sequence?

Well, part of my problem is I didn't have all of the details on his equipment. And I think you and I have already agreed that he needs to remove the second pump if he isn't using it. So no need to start a debate when the recommendation is clear. I'm just trying to understand a few details regarding what he actually wants to do before I make a recommendation to perhaps help him increase his flow overall.

Please see my signature for system hardware details. The 1st pump is at the inlet and the 2nd pump is at the outlet. I deactivated the 2nd pump (outlet) because Koolance recommends that if you run a single pump with the RP-452X2 that you run it at the inlet. I created this thread more as an FYI for people considering the RP-452X2 and to get general input. Thanks.

[Stewie007]

This explains a bit why you are loosing some flow. That goes right with what Nkrader is saying.

By doing that you're pumping into a large chamber then forcing the liquid out. So yeah, it is contributing more than I thought. I was considering it originally as a direct pump to pump exchange rather than a pump to reservoir and THEN through the second pump. Gotta have my eyes tested.

In this particular case, your second pump is adding no net benefit. The very reason it is showing less flow is because of that pumping sequence. Were it an inline pump that was feeding in series it would be different. However, when you are feeding pumps in series they should be matched for flow.

This is what I would do in your scenario. You have two pumps, and you paid money for them. This system would probably benefit from a parallel distribution system. That way you might see a good bit of increase in flow from the 70GPH that you're seeing. When you pump into a reservoir, there is no exchange between the pumps. IF you use them in parallel, the exchange will be realized when you combine the discharges.

So you're basically splitting before the reservoirs, and then you're recombining after the pumps. I would use some y splitters. That way you are not flooding any reservoirs, and you're getting a benefit from two pumps through combining the discharges.

Try experimenting with that and see if it gives you a benefit in flow.

[stren]

Martin wrote a great article on series vs parallel pumps:

http://martinsliquidlab.org/2011/04/...s-vs-parallel/

[Stewie007]

He might as well experiment with it given that he has a second pump already.

The writer of that article is dispensing advise on parallel loops by using two different types of pumps? I question that set up as it has no balancing what-so-ever. The test is not effective if both pumps deliver different suction pressures. That will help rob the overpowered pump of its ability to be effective. It should come as no surprise that performance didn't increase much.

The pumps here have dedicated reservoirs, so it would be worth experimenting to see if any differences are achieved, as 70GPH is not high flow (unless Koolance was joking, hehe). It may just be too small of a system to notice, however he doesn't have to buy an extra pump.... so why not test?

[Jakusonfire]

I have the same unit and it works fine for me. Vibration only gets noticeable when both pumps are running up towards top speed, but in my build the unit is secured from the sides and not just the front.
Two pumps is not needed in my loop, I just had an original PMP450 that I then replaced with a USB D5 and later converted the pmp450 to USB as well.
With two pumps running I can get the same flow rate at very low speed as I got from a single pump at higher and more easily noticeable speed. From minimum to normal loaded speed both pumps at 3000RPM I can't hear them

If it is set up properly for serial dual pumps the water moves straight from pump 1 to pump 2 and does not go via reservoir 2

I saw only a very slight difference in flow rate for the same loop with just a single pump vs 2 pumps with 1 switched off. Its not nothing but its not all that much either.
Martin also tested the earlier version of this res using the short serial kit like the OP has and using a very long piece of tube to connect the pumps and there was no difference.

What I have noticed with this unit is that especially in serial operation, because the pumps are upside down the unit really needs to be tipped right back to get all traces of air out of the pumps themselves. Trying to bleed it while level is an exercise in futility. Leaned back at least 45 or preferably 90 degrees and it is easy.

Edit: Even at minimum 10V D5 strongs are still spinning at relatively high speed and pressure.

[Stewie007]

You know, I didn't think of it the first time, but cavitation is likely to occur when both pumps turn on at the same time. You may consider running the second pump on a time delay of several seconds.