Air sucked into water loop on startup? No leaks.

[troyerjs]

Air bubbles are coming from my VRM waterblock only on unit startup. I see them coming out of the tubing from the VRM to my CPU block and will typically get stuck in my CPU block until I tilt my tower. Been happening for 6+ months, when I first commissioned the system I fully tilted and rotated the tower with the pump running to get as much air out as possible to my reservoir to vent. So for air to still be trapped and coming out of the VRM block after 6+ months I'm pretty doubtful. No leaks have been noticed. I thought it was the discharge fitting of the VRM block and recently drained the system and replaced the top fitting as well as sanded down the end of the pipe to ensure it was flush and I'm still having the same issue. I didn't think to change the inlet fitting of the VRM block but I inspected all the o-rings on the VRM block and CPU block and didn't notice any pinched/cut o-rings. My thoughts are it's either a fitting, or a design defect in the VRM block that came with the motherboard. Unless, there is some weird vapor pressure phenomenon going on when the PC shuts down almost like cavitation due to high pressure drop (wouldn't think so on this low pressure/low velocity system).

I actually came across the exact same issue on this thread from 2018, but it doesn't appear a resolution was ever posted.
Any thoughts?

 

[EarthDog]

Neat-looking build!!! Nice!!!!!

Well, If air is going to get stuck anywhere in that system, that's the spot (highest point). Ideally, you want your rad ports on the bottom too (same point as above, highest points in the loop are where air goes and your rad ports are there). How do you know that air is coming from the VRM area exactly?

I suggest draining and leak test with just pressure. Put some soapy water around your VRM ports and see if it bubbles.

 

[troyerjs]

Neat-looking build!!! Nice!!!!!

Well, If air is going to get stuck anywhere in that system, that's the spot (highest point). Ideally, you want your rad ports on the bottom too (same point as above, highest points in the loop are where air goes and your rad ports are there). How do you know that air is coming from the VRM area exactly?

I suggest draining and leak test with just pressure. Put some soapy water around your VRM ports and see if it bubbles.

On startup the first place I can see air bubbles is through that reversed 'C' section of piping in-between my VRM and CPU block, (I have frosted tubing, but it's translucent enough you can see through it). Then, the air typically gets stuck in the CPU block until I tilt the tower or throttle the pump speed to push them through. I think the VRM is probably the highest point in the loop, the rad might be just about even, but I was pretty aggressive when I initially filled the system with tipping and turning the tower on its side (as much as I can without running the pump dry). If there is an air pocket in my rad, I don't think it can flow backwards back to the VRM just by the piping layout. When I shut down, I don't see any air bubbles going into the VRM block from the pump. So, there is something about the VRM or it being the highest point, which air shows up there on every start up. Interesting enough, if I'm shut down for just a couple minutes it will be very small bubbles, versus if I'm shut down for a day there is little bit more air.

But yeah, I might just have to drain it and do a pneumatic pressure test. I didn't buy the equipment for it, so I was saving that for a last resort.

PS - Thanks for the suggestion on flipping the Rad, I was actually unable to do that because I put the rad in the 'fan cavity' of this case and I would have had to cut out holes for the ports (already had to cut a small section for one of the ports at the top but it was much more straightforward than the bottom would have been).

 

[gregmacknass]

Buy yourself a cheap 1/2" I'd 3/4" od soft tube compression and about 3 or 4 feet of matching hose. Attach it to your fill port on your res. Fill the majority of the hose with your cooling fluid. Have an assistant turn the case in every single extreme angle possible while you hold that hose straight up. That should solve your problems. Just drain the excess in the tube into a pan, then shut it down, put a towel around, and remove the hose and fitting. Plug then done.

 

[troyerjs]

Buy yourself a cheap 1/2" I'd 3/4" od soft tube compression and about 3 or 4 feet of matching hose. Attach it to your fill port on your res. Fill the majority of the hose with your cooling fluid. Have an assistant turn the case in every single extreme angle possible while you hold that hose straight up. That should solve your problems. Just drain the excess in the tube into a pan, then shut it down, put a towel around, and remove the hose and fitting. Plug then done.

Thanks, I can try this. Neat trick to really rotate the tower without cavitating the pump. Maybe there is a weird air pocket in the VRM that just hasn't come out after months of operating.