Air bubbles in a large case

[krojew]

Hi,

I have a quite peculiar problem. I wish to install water cooling in a Caselabs STH10 case (radiator in the upper chamber). For those who don't know this case - it's absurdly large and that's the problem. The case is so large, that it's a non-trivial task to wiggle it around, almost impossible. So the question is - how to get the air bubbles out of the system with minimal movement of the case? I honestly have no idea.

[RollinLower]

1 minute ago, krojew said:

Hi,

I have a quite peculiar problem. I wish to install water cooling in a Caselabs STH10 case (radiator in the upper chamber). For those who don't know this case - it's absurdly large and that's the problem. The case is so large, that it's a non-trivial task to wiggle it around, almost impossible. So the question is - how to get the air bubbles out of the system with minimal movement of the case? I honestly have no idea.

if you are going hardline tubing then there pretty much is no other way. if you are going soft tubing then you could always pinch the tubing somewhat.

 

really though, you are going to need to wiggle the case around a bit to get 100% of the bubbles out.

[Naeaes]

Leave the pump running for 24h. So long as the reservoir is above the pump, it'll gather all the air in there. If you tilt the case to both it's sides every once in a while you'll speed up the process. Generally speaking, radiators aren't much of an issue in terms of holding on to bubbles. If it's any kind of decent build quality, the pipes will be very smooth inside. Some wonky GPU blocks can be a huge pain in the ass.

[krojew]

Just now, Naeaes said:

Leave the pump running for 24h. So long as the reservoir is above the pump, it'll gather all the air in there. If you tilt the case to both it's sides every once in a while you'll speed up the process. Generally speaking, radiators aren't much of an issue in terms of holding on to bubbles. If it's any kind of decent build quality, the pipes will be very smooth inside. Some wonky GPU blocks can be a huge pain in the ass.

The reservoir is above the pump, but the radiator is above the reservoir. Won't air get trapped in the radiator instead of traveling to the reservoir?

[Naeaes]

13 minutes ago, krojew said:

The reservoir is above the pump, but the radiator is above the reservoir. Won't air get trapped in the radiator instead of traveling to the reservoir?

You'd think so, but it really is less of an issue. Usually the water pressure carries the bubbles with it just fine. And since the inside of the radiator is really smooth and there are no junctions, there are really no sites where the bubbles can kind of cling on. Most problematic are setups that have many forks in the loop. Either inside blocks (for instance the ones for dual-GPUs like GTX690) or just T-junctions in the tubes if there's like a parallel fork for RAM/Chipset. One side will have less pressure and hold on to bubbles better. And then there are situations there big bubbles get to the pump which crushes them into tiny bubbles or even foam. I mean, every loop will do that in the beginning, but after it's rotating fully and topped off, it should stop happening. If you have both ports of the reservoir below the liquid level, there's no route for the air to backtrack to the reservoir while the rig is powered off.

 

The easiest way to get rid of the bubble is really just running the loop for a long time and giving it a wiggle every now and then. Use a jumper wire to turn on the PSU and only plug in the the pump for the bleeding so you don't have to worry about burning your CPU and GPU. Any time I had the chance I left the loop bleeding over the weekend. It was a great way to weed out bad pumps too. It wasn't just once or twice when on Monday I got walked to a burned Laing DDC. They have a tendency to overheat since they're aircooled.

[Mick Naughty]

Turn the pump to max. If it doesn't work you need a stronger pump. Had no issues the my pump was the lowest part of the loop.