Periodic Cavitation???...
 

Or strange pump noises revisited...

every 4~5minuettes I get a sort of squiissshhh noise, I found out it's cavitation,(by squeezing the pump inlet closed), why would the pump(1250Ehiem) cavitate periodicaly?...
It stop[ed doing this for a while when I was leak testing out of the case, now it's doing it again, there's no air in the system that I know of, I've banged and tilted the rad and no bubbles are to be seen. even when it cavitates, though you can hear it going through the block if you're close enough...

I live in a house that creaks, bangs and groans(in time with my stomache :p) but surely my pump can't be haunted already!! :drool: ...

Hey MadDog,

As you probably know, a centrifugal pump (like the E-1250) creates a high pressure area at the discharge and a low pressure area at the suction. The diff pressure causes water to flow thru the system. Cavitation can occur in the low pressure area where water enters the pump impeller.

Are you running a closed loop system without a reservoir?
Are you using Silicone tubing?

If Yes... one situation I have seen is the system will gradually loose water over time due to the relatively high permiability of Silicone to water vapor. If you don't have a reservoir or fill tube to replace the lost water the system will actually start operating under a low vacuum. Air can be sucked into the pump inlet fitting and cause cavitation like noises. If the system is sealed really well so no air can leak in, it may be possible for the pressure to drop enough to actually cause the pump to cavitate.

I hope you get to the bottom of this and solve your problem.

RoboTech

I don't have silicone tube I'm afraid, I have PVC. I do have a reservoir, it's 'sealed/airtight' with airspace in the top 1/10, though I had the same noise/affect when the res' never had the lid on.
I have a head of no more than 8inches...

The head shouldn't be a factor, in a closed loop.

If you are indeed encountering cavitation, then it would be because your rig is too restrictive for the pump, causing a huge pressure drop (a requirement for cavitation).

What size tubing do you have? If you have PVC tubing (piping actually), then you probably have lots of elbows, right? That might be the problem.

To obtain cavitation pressure water must be under 2300Pa (water at 20°) locally, I doubt with your EHEIM you obtain a such pressure in a Watercooling system without a high pressure machine (axial machine) and a hard vacuum in a reservoir (conditions are hard to obtain) :D . It's not your case with your reservoir with no lid on.

It's probably air in your system which do some noise in the pump but why it is aperiodic :confused: . As bigben2k said the pump is perhaps doing some noise because system has a big pressure drop and the pump must force to fight against this one (this happened to another people who take big pump with a restrictive maze, the result-> noise in the pump..)

Is your pump suction line connected to the reservoir? If so, that should keep the closed loop from starving flow to the pump. What components does your cooling loop consist of besides the Eheim?

I agree with Rascal, I don't think the Eheim (or any pond pump) can generate the required low pressure for true cavitation (at normal room temps). A lot of what we call cavitation, I beleive is the impeller rattling/vibrating on the shaft. I have heard of people glueing their impellers onto the shaft to stop this noise. I have not experienced it with any of my pumps.

Maze3 with 1/2inch PVC piping:D with no elbows(out of case mess with everything place 'just so' to keep the tube'straight' and pressure off the block), 'BigMomma'(European Version ;)) heatercore with 1/2inch (inner) barbs. the reservoir is a tall(10inch x 5 diametre about 2litre I think) sweet jar with a sealed lid, there's quite abit of air in it, this would cancel out the 'closed loop' push/pull affect would'nt it?, or 'dampen' it cause the air is compressable?...

The reason I thought it was cavitation was cause it's exactly the same noise as if you pinch off the inlet(which I read was cavitation). if it was air in the system it would have worked it's way to the reservoir air pocket/trap (pump inlet is 4inch under water, in and outlet are to far apart to draw air from one to the other).

Help?:cry:

Maybe there's a funnel forming at the pump inlet, causing it to suck air.

The rig is a closed loop if, and only if, the coolant can flow in either direction, and a return of coolant is forced. If any air can get in, then it's an open loop.

I always thought of 'closed' as a reservoirless system:shrug: .

No air is funneling down, the pump suction tube is 4inches under water, the surface is very calm...

I think its very strange this!. could it be powerspikes or something to do with the mains?. I only ask cause it's the same sortof sound you get sometimes when you first turn the pump on...

Did you do the impeller fix?

No not yet, I want to be sure of it before I do, it could create more pump~heat if the pump cant 'slip' could'nt it?, the motor would be forced to do the work it would'nt normaly?...

There's definatly air forming with the noise, it gets to a stage where there's a trickling noise now!, if I turn the pump off then, and tap it one or two tiny(match head sized) bubbles rise into the tube :( . air is'nt getting into the system anywhere, so it has to be cavitation forming it does'nt it?, but why?!...

You might not be able to see it, but air can get sucked in at the pump inlet. Tighten up your fittings a bit!

Pump inlte is 4inches under water as well :cry: , It's getting noisy now, before the bubbles were getting sucked away(the cavitation noise?), now they are hanging around to have a gurgle.
I'd take the sucker apart now and temparary stick it ti see, but it's all temparary at the moment anyways, until I can afford some silicone and a new rad and move in~case...

** you mean the barb anyway dont you? :) ...

In some way your water is being oxigenated, that tini bubbles are being trapped at a low pressure spot in your pump impeller. A larger bubble starts to grow and make noise until is too large and is released, then the cycle starts again.

May be your water inlet to the reservoir is over the water level, cascading into it? that could be your air source.

I agree.

From the sound of it, you have an air leak somewhere.

I tested my pump with dirt cheap clear vinyl tubing, and 1/2 (actually 5/8 OD) CPVC barbs ($3 and $1). You might consider doing the same, just to check out that the pump is OK.

Work your way up from there, and find the leak!

How is the inlet/outlet set up on your res? I know with my res if the coolant level was too low, a tiny vortex would form and allow air to get sucked into the outlet. Didn't happen all the time, but it got my attention when it did. I know you said yours was calm, but I'm just trying to picture your setup.

The reservoir is an old 10inch tall sweetjar[Roses, ahhh! you can have the [H]ard centres :p] it's about 2ltr capacity (lots) it's full to a depth of 7.75inches, both the inlet and outlet are 4-5(slanted cut end) inches under water :( ...

There IS air forming somewhere though, and the cycle nyczeg mentions sounds spot on...

Tested it how BB2K?...

Just plug the inlet to the outlet, with a tube. I'm doing exactly that today, but with my new airtrap.

Come to think of it, the impeller is probably the source of the noise. Many people here have reported what you have.

You could try fixing the impeller with simple white (paper) glue, just to see if that's the fix you need. Pull it out of the pump, and let it dry overnight.

It won't run very long, but it should be long enough for you to confirm that the impeller is the problem.

If you're ready for a permanent impeller fix, then go for it.


The problem of this rattle appears only when a rig is very restrictive: some people have this noise, and some don't. You must be right on the edge, but it seems odd that the noise would come and go. I would point to tubing that collapses/kinks, but you're using piping, so n/a. I dunno... I'd crack open the pump to check it out, but I often do more than is necessary.

I will check it soon, it's just I hav'nt had it up and running long, the stiff PVC is a bitch to get in the case!, so I got it out of case with extremely straight, 'no kink' lines :D .
The M3 is restrictive, so is the Cu heater core I have, I want to replace it with a purpose built PC rad purely for fitting it incase easier(and low pressure drop!). I'm torn between an OCPC 'Abyss' or two 'The Deeps' in parallel, but I don't think I have a high enough flow to make parallel worthwhile. the other thing was I was going to get the 'Deeps' one at a time as I could afford them, well!, more 1 now and another if(when!!) I go TEC, as the Abyss may be overkill(well, deminshing returns for cost) to a H20 only rig...

The average best flow for a rad was 2gpm was'nt it?. can you give me a link to where BillA posts his work?...

2 gpm may not be the best overall, but there is a point of diminishing return, that may fall within that range.

BillA's article is at overclockers.com, and is titled "radiator roundup", I think... Keep in mind that the graphs have since been revised.

I thought it was revised from 0.5gallon?... then again it could be 5gallon from 2!:D ...

Sheesh, I just hooked up my system tonight and did some leak testing for a few hours and it was nice and quiet.

Well, as I made my way down to put my machine on its table to hum along, I carefully examined everything and made sure that everything looked ok before firing up the Athlons, and wham, as soon as I turned on the juice to the pump I hear nothing but a load rattling/buzzing sound. It's pumping water because I can see the swishing of water in the res, but it's noisy as all hell.

Well, needless to say it annoyed the heck out of me, so I turned the pump back off and reached into the res and pushed on the impeller and fired it back up. It was a lot quieter, but 20 seconds later it was back to the same old routine.

Is this almost always the "I need to glue my impeller to the shaft" scenario here? Or do you think it's a load issue?

I'm running a Profin 250 gph pump submerged in a 4x4x4 custom res that is filled about 80 to 90%. I have 2 cpus I'm cooling w/1 maze 3 and 1 K4.1 block, and an original black ice rad.

I'm thinking it'd definitely the impeller though. After pushing it in a little and firing it back up, it was quiet briefly before it went back on the rampage. I'm just wondering what you guys think.

God, I would hate to have to unhook that darn res and yank the pump off of the rubber pads I glued it to in order to glue the stupid impeller.....


-Ice-

Sounds like it!, I'm sure mine is because of to much backpressure/restriction in the system. I'm just too lazy to strip it all down just for the pump, I'll test it when I do the incase~rebuild...

argh, the thought of having to tear this thing back apart is ..... killing me.

So help me understand this.... The impeller has to be attached securely to the shaft otherwise there is no way it could produce water flow under all the pressure. How can it be somewhat loosely mounted that requires glue to fix?

Smack me with a stupid label, but it looked like it was sitting on there solidly to me and not loosely connected (only glanced at it). Are there little teeth or something that have to meet on the impeller and shaft before the impeller actually connects solidly to the shaft and spins? If so, then how the heck can it keep making the annoying noise if it is already in motion and spinning? It is only turning in one direction.

Also, this thing worked silently on every other startup before, and now it seems like something just... out of whack no matter what I do (short of a pump surgery procedure).

I just hate it when your hard work starts to get ugly on ya.


-Ice-

Nevermind :) I found an article on it. It looks like it's probably as simple as gluing the impeller to the shaft with some epoxy and I just shouldn't get the epoxy on the top/cap.

I guess it's worth a try, so I'll give it a whirl tonight.

-Ice-