Bleeding air from a new custom loop

[brighttail]

Hey all I'm trying something new with my tubing and altered my original run.  

 

I now have the run going from my pump/reservoir, to my GPUS, up to my CPU and into the top radiator.   From there I have one long tube that goes from the top radiator to the bottom, then back up to the pump.

 

My problem is that big long tube from the top radiator to the bottom.  I'm having a problem bleeding it properly.  It is too much tube or do I just need to let it run and eventually it will bleed out on its own.  I've been tilting the whole case like crazy but there is just a lot of air in that long tube  it seems.

 

Suggestions?  Thanks!

[Enderman]

what pump

[DXMember]

maybe you could post some pictures?

[QBtech]

Sounds like you dont have enough headpressure to bleed the loop. Any chance you can run it with some extra power? Is the top of the reservior? 

 

Regardless a tube filled with air means your loop is not circulating and that you are killing the pump in the process. 

 

Without some additional info or pictures it is hard to help with the specifics

[brighttail]

I believe i have enough pump pressure, in the last 30 minutes that long portion of the tube has filled up about 3/4 of the way.  Every 5 minutes i'll get a big surge of air being purged.  Still I'm wondering if it is a waiting game or if I'll have to go back to my old setup.  This is much cleaner.

 

That being said, here is some pix.  The first pic shows the big tube coming from the top rad to the lower one.  Right now as I said it is 3/4 full from the bottom.

[brighttail]

Yeah it seems like the whole loop is good except for about 5-6 inches at the top of that tube coming from the top reservoir.   Maybe my pump isn't strong enough to get that last bit down into the reservoir.   Any other suggestions or do I gotta remap my loop?

[QBtech]

Believe me, that is a lack of headpresure, no doubt about it. Your loop is not circulating (enough) and running it like that is slowly killing ur pump. Your pump is strong enough to run the loop once that bubble has passed but for now it is struggling. 

 

Ive had the same thing a couple of weeks ago

 

If that is a d5 vario check if ur pump is all the way turned up to 5 or if it is a pwm controlled pump look in the manual/google how to run it at full speed if it isnt connected to a pwm source. 

 

[brighttail]

3 minutes ago, QBtech said:

Believe me, that is a lack of headpresure, no doubt about it. Your loop is not circulating (enough) and running it like that is slowly killing ur pump. Your pump is strong enough to run the loop once that bubble has passed but for now it is struggling. 

 

Ive had the same thing a couple of weeks ago

 

If that is a d5 vario check if ur pump is all the way turned up to 5 or if it is a pwm controlled pump look in the manual/google how to run it at full speed if it isnt connected to a pwm source. 

 

Well it is attached to a PWM source and right now I'm just doing a leak test.  I did a little bleeding with my drain, which is the lowest part of it. Right now it is about 1/2 from being perfect.  So I think what i'll do is after i'm doing doing this System clean from Mayhem, when I start flushing it with DI water, I may start up bios and ramp the pump to full to see if it will fix the issue. If not then I can redo my loop and test again before I start putting in my pastels.

[QBtech]

2 minutes ago, brighttail said:

Well it is attached to a PWM source and right now I'm just doing a leak test.  I did a little bleeding with my drain, which is the lowest part of it. Right now it is about 1/2 from being perfect.  So I think what i'll do is after i'm doing doing this System clean from Mayhem, when I start flushing it with DI water, I may start up bios and ramp the pump to full to see if it will fix the issue. If not then I can redo my loop and test again before I start putting in my pastels.

Because I still dont know exactly which pump you have,  im gonna assume it is a d5 pwm.  If the pwm wire of a d5 isnt connected to a running source it will only run at 40% pressure. Here is a way to side step that problem:

See image

My go to retailer has a little tool that just this but i lind seem to find it in it is probably just some generic wiring anyway

 

All pwm pumps have similar shortcomings, mostly for those people to lazy to plug them in. 

 

Than again I might be wrong, but your pump simple seems not to be running powerfull enough to push an air bubble through an unrestricted tube. To be it screams 'pressure problem' 

 

Nice build btw

[brighttail]

Yeah i Saw this.  Pushing an air bubble about 2 feet against gravity i think is going to be tough for any pump not running at full speed.

 

As I said I may just wait until I can boot the system up and get a proper PWM signal.  If i need to remap the loop I may have the pump start by pushing water up through that long length of tube then it is all down hill from there.   We'll see.

 

Thanks for the advice.

[QBtech]

You might want to try pulling that pwm wire of thr board though, im no expert but I heard a rumor somewhere that newer models will actually run a little faster when they are not attached

 

On a side note: flushing your system if your loop runs like this is really hard  if both radiators are orientated with the ports towards the frontpanel you might try running it on its back (so i/o on the table) that way all water entering the radiator will push the air up and through. But dont starve the pump

[brighttail]

Well in the end I just said screw it and went back pretty much my original loop with a minor change.  Seems to be working fine.

 

Thanks for the help.

[brighttail]

Grrr.

 

Redid the loop exactly like i did last time and there is about 1/4 to 1/2 inch that of air that won't clear at the highest part of the loop.  Turning the pump up, it actually makes it worse, turning it down to under 2k rpm fills it.  So very weird.  I can't tear this apart and redo it all over again.  I hope that the 1/4 of an inch isn't going to matter.

 

I reduced the speed to about 60% which I don't mind and it filled.  I'll let it run this way for a while.

[GhostHitWall]

ah If you still care, you can somehow add an air valve or loose point (preferably at top of res imo) for the trap air to go out.
You absolutely do have enough head pressure.

 

It could literally take days to let all the air in the loop aggregate into the reservoir.  Just be patient, even if you can't complete the loop for the first few hours.  Let the pump do its job.

 

And I would suggest that you jump another PSU that only powers the pump for a process like that.

[brighttail]

5 hours ago, GhostHitWall said:

ah If you still care, you can somehow add an air valve or loose point (preferably at top of res imo) for the trap air to go out.
You absolutely do have enough head pressure.

 

It could literally take days to let all the air in the loop aggregate into the reservoir.  Just be patient, even if you can't complete the loop for the first few hours.  Let the pump do its job.

 

And I would suggest that you jump another PSU that only powers the pump for a process like that.

I have the fill port at the top  of the reservoir that i'm using as a bleed valve. Sadly the radiator and the area with the area that needs to bleed is above that.  The upper rad only has two ports on it.    I suppose I could carefully turn the computer on its back. That would make the reservoir the highest point, but I don't like the idea of tipping it that far as I'd almost have to go  a full 90 degrees. 

 

I think i'll wait a couple of days to see if it works out.:)