can't keep pump primed with Pentair rebel pool cleaner

[Dirk]

Cavitating? I don't know much about that, for sure. Except to keep it from happening when there's enemy submarines about!!

How would caveating contribute to losing prime? Or air in the pump basket? Based on my time aboard the Red October (OK, well, my time watching the movie!!), I would think the air would form after the impeller. Can an oversized pump pull so hard that the bubbles form before the impeller?

And if so, is that one of the symptoms that could happen when the flow is restricted before the pump, like from a vacuum? That would fit Keelie's symptoms, wouldn't it?

 

[Wheelie]

Okay, The pump has been working correctly for 24 hours so that's fine.

And yes I was referring to the flow gauge (thank you for the correct terminology) which I used at the end of the hose (as per the instructions ) - and it indicated the right amount of flow although I'm not sure what the pump was doing at this point. I have to work out how to keep the flow gauge under water and check the pump at the same time. If the flow is okay at the end of the hose, I wouldn't need to check piece by piece would I?

We're out today but I will come back to this....

 

[Dirk]

If you want to be thorough, and not have to repeat the whole suite of tests, you need to check each component, one at a time. If you want to take shortcuts, and test groups of the components all at once, that is a different kind of troubleshooting technique, which involves testing half of the things at a time, then halves of those, etc. It can save overall time, but it is harder to explain and the one at a time testing is simpler to execute, though potentially takes longer. If any of that makes sense, then go ahead and use the half at a time method. I didn't want to confuse things by listing all kinds of different ways to go about this, to add to confusion, because the problem we've been having, is the multiple methods and mixed results flying around that has extended this process. See what I mean? Make a plan (or use mine or someone else's). But pick one, and stick to it. Be methodical. Don't jump around, and instead work the problem from one end to the other.

Anywho... we're not testing for flow at this point. I asked you to record flow results just to be thorough, so we could go back and look at that if we needed to. The primary test we're doing is on the pump. We're starting from a point where the pump is keeping prime (and now you've got that), and adding things one at a time to see if we can get the pump to lose prime. Then you'll know which thing is causing the problem. So you have to be able to observe the pump, and you have to be able to run the test for long enough to rule out each part you add. I don't think just sticking the end of the hose in the water for a minute and checking the gauge is going to tell us anything. The use of the flow gauge is merely simulating the resistance caused by the vacuum, while the vacuum is out of the testing loop. So, rig something up that will keep the end of the hose underwater while you let the pump run for an extended period of time. See if you can get the pump to stall. You could tie it to a weight, or even the vacuum (as that is obviously heavy enough to keep the hose submerged). Just don't connect the vacuum to the hose. Or, if you want you can alter the order of things: you can connect the vacuum to the hose sooner than later, to test it in the loop sooner than later, then add the remaining sections of hose after that, one at a time. And there you go, now you've got me giving you all kinds of different ways to do this, and that'll put us right back where we were, running around in circles.

Sorry, this is all crystal clear in my head, how to run this test suite in any number of ways. It's harder to type out an explanation than it is to do the tests. The gist of it: you have to be able to test each component of the vacuum system against the performance of the pump. For each sub test, you have to either have the vacuum in the loop, or simulate its resistance. I suggested the flow gauge because that does two things: it simulates the vacuum, and it verifies that you're running the test with the pump providing the desired flow.

I can help you best if you follow my original plan exactly. It seems you started in the middle of it (did you test with just the gauge, and then with just the gauge and the valve, and then with gauge, valve and one section of hose? or did you just take the cleaner off and test everything else all at once?) I'd need you to start at the beginning of it, please. And I'm happy to re-explain all this as I'm sure it's all clearer in my head than in how I typed it out. If you want to use another plan, that's fine, have at it...

 

[Wheelie]

Okay will report back after the weekend. Need husband home to look after kids while I work this out!

 

[Dirk]

The cleaner itself, the Regultor Valve and the Flow Gauge will all be underwater when you test them. When you test the hose sections, try to keep as much of the sections out of the water as you can. Again, you're trying to locate how the air is getting into the pump, or if the pump is caveating and "creating" it's own air. Anything under water is not going to pull in any air through a leak.

If it's a leak, it's either in the vacuum system (and likely only the hose), the PVC pipe or elbows or unions or glue joints between the skimmer and the pump, any valves between the skimmer and the pump, any other component glued in in between the skimmer and the pump, or the pump itself. You'll need to test each of those things. We're starting with the components of the vacuum system.

If it's cavitation (which I don't know much about, but can guess) it'll be caused by some combination of pump size, pump speed and the amount of resistance being caused by everything in front of the pump (any or all of the things listed above). In order to test for cavitation, you need to maintain the resistance that was present when the problem occurred. For the most part, that means all the valves set as they were, and the vacuum in place (or the gauge standing in for the vacuum).

As you test each component, you're trying to get the pump to stall. You're starting from a point where everything possible has been removed from in front of the pump (which means every component of the vacuum system) and the pump is working fine (you've completed that step, right?), and then you're adding things back in, trying to get the pump to lose prime again. Each test of the vacuum system needs to simulate the suction resistance of the vacuum head. So either the flow gauge or the vacuum head should be included in each test. Since you cannot connect the vacuum head directly to the skimmer, or the regulator valve, then at least those two tests must be with the flow gauge standing in. If you want to test the entire hose all at once, you can, then if the pump loses prime, you'll need to go back and test each section by itself to find out which is leaking air (I know that contradicts my "one at a time" instructions, but this is just a variation of them, based on the fact that you're on your own here and probably need to get through this faster than I originally proposed. Up to you.).

As much as possible, for each of the tests, the flow should be maintained at the proper flow rate for the vacuum. So, again, that's why I suggested keeping the flow gauge in the circuit as much as you can. There's no point in testing when the flow rate is too high or too low.

If I was testing the hose sections for lengths of time like this, and had to observe the pump, I'd rig up something to hold the end of the hose (with either the gauge or vacuum attached), just under the surface. A length of pipe or 2x4, laying on the deck, with some sort of weight laying on it over the deck (a big rock, or some bricks, a sack of potatoes, a piece of furniture, whatever) and then I'd tie a piece of rope to the other end of the pipe/board and hang from that another weight to which the end of the hose is tied, the length of rope would be just long enough to keep the end of the hose just under the surface of the water. That way I could leave it like that for hours if need be, waiting for the pump to lose prime.

The wild card here is the filter, because it seems you've gotten mixed results depending on the state of the filter. I'm assuming the filter is still clean. Clean it if not before you test anything.