How many bubbles...

miniyazz

miniyazz

miniyazz

Soldato
Joined
25 Dec 2008
Posts
5,976
Location
Sheffield/Norwich
... can there be hiding in a loop :eek:

My pump appears to be churning them out at a rate of knots (microbubbles, not big ones) although there's none going into the loop, I've got the intake covered!
I'm aiming to make this one a completely sealed loop - the quick disconnects have been put to rest, and I'm currently leaktesting (with the mobo+cpu on, 'cause I tightened those fittings real good, and I can't be bothered to fiddle with the power connector :o)

Anyway, stupid bubbles :( this may take a while.

On a side note, imbibing mouthfuls of month-old water contaminated with biocide while refilling the loop is decidedly not fun :p
 
Righty, I'm off to bed now. Any tips on how to get air out of waterblocks/pump? Hopefully it'll have sorted itself out by morning (yes I am about to leave my computer on and unattended for hours with an open water-cooling loop while technically still leak-testing :eek: :p)
 
Just musing...

Say you have a reservoir, there's a space at the top where there's a bit of air.

There is also air trapped somewhere along the loop.

Say you remove the air at the top of the res. You use a vacuum pump to suck out the gas which is the air.

There's now negative pressure in there.

The water won't go anywhere, it can't, it's not compressible, it will do what mavity told it to and sit at the bottom of the res.

However the bubbles along the loop are compressible because they are gas so when the pressure in the loop dropped due to the vacuum pump they would expand and as they expanded they would be more likely to be flushed out by the flow of water due to the greater surface area.
 
Hotwired said:
Just musing...

Say you have a reservoir, there's a space at the top where there's a bit of air.

There is also air trapped somewhere along the loop.

Say you remove the air at the top of the res. You use a vacuum pump to suck out the gas which is the air.

There's now negative pressure in there.

The water won't go anywhere, it can't, it's not compressible, it will do what mavity told it to and sit at the bottom of the res.

However the bubbles along the loop are compressible because they are gas so when the pressure in the loop dropped due to the vacuum pump they would expand and as they expanded they would be more likely to be flushed out by the flow of water due to the greater surface area.
Click to expand...

Interesting thought, but I don't have a reservoir. It could work, I suppose - JonJ678 would probably be able to confirm, but I presume he's away with exams or some such thing?

Biffa said:
I find they go away after a few hours, unless your system is sucking air in somewhere :)
Click to expand...

The intake is definitely clear, I've just double-checked it. Nothing going in there. A couple of microbubbles emerge from the CPU block every second or so. A few more from the NB block. And lots from the pump. It all seems to be leak free - if no water's escaping, is it safe to assume no air is entering any of the seals?
I'm inclined to think there's some air in each block tbh, though at a couple a second I'd also have thought they'd have sorted themselves out overnight :confused:
 
The rad is the last thing in the loop before it drains into my tupperware bowl I'm temporarily using as as res, so I'm not fussed about any air bubbles in there, if bubbles come out they just leave the loop. Besides, there's a fairly large number of bubbles entering it from the pump, so it wouldn't do much good :)
I've manoeuvred the case (with blocks & pump attached) as best I can, but it's really heavy :eek: the case empty is about 14kg, and even without the graphics cards it weighs a ton, when you're trying to move it around without disrupting the tubes entering and leaving it. However I shall now attempt to re-tip the whole system a few times in several directions.
 
I remember doing mine with a t line, took about 2 days to clear properly and bubbles kept on collecting by the CPU block, I had to guide them round like some giant incredably heavy 3D puzzle!
 
**** :p
Well got almost all of the air out of the blocks I think. Also got a ridiculous amount of air out of the pump, turning it off then turning it around and turning back on helped a lot. I could tell immediately when the air was out, the pump turned practically silent. 30s later, there's a bit more air in it and it's noised up again :rolleyes:
 
I just realised why I have bubble issues :o
Water can hold more 'dissolved' air when cold. So, I've got my sealed loop with 0 air bubbles. Then, I plug it in and crank up the heat. Boom, air bubbles. Looks like I gotta reseal it with hot water!
 
Hehe, that's one option. The other thing I was considering is using an iron on one of the waterblocks to heat it up, I feel this would be better for me as the loop can start with cold water. I don't fancy trying to fill it with near boiling water, it's hard enough as it is! :eek:
 
Well, another method you could use to degass your water is if you have a T-valve or 3 way valve, and a container which you can put water in that you can bung up with a cork or something. (I'm basically trying to recreate in your home the quick and dirty way we do in the lab)....
Attach T- to tap via short hose, so that it runs .... wait a sec, this might be easier with ms paint or something.....
brb.....
 
Right ok here goes, forgive my bad drawing....
2unu2pe.jpg
 
basically the fast tap flow creates a pressure drop (small vaccum), which then the gasses from the liquid come out to equilibrate.
 
I had a scare last night, one of the full cover blocks on a 295 wasn't seating properly and was lifted off one end of the pcb by about a mm. :eek:

So it was like one core was hitting 80 after a second of folding.

So I managed to drain half the loop (double pump res) without losing much fluid. Then it was the nightmare of pulling three cards all connected with SLI WC attchs off the mobo in one piece without breaking anything or pouring liquid all over the place.

Managed to sort it though after 3 hours, got it back together and man the amount of air that was in the system, left it running overnight with the filler plug out of the res to breathe the air out, talk about Bubble central!

So to stay on topic, I feel your pain :)
 
You must log in or register to reply here.
Back
Top Bottom