Quote:
Originally posted by Cova
Uhh..., having air in a system isn't what determines if it's a closed loop or not. Whether the system is sealed or not is what determines if it is closed-loop. So long as his res is sealed shut it is a closed-loop, just potentially one that still has air in it.
As for where to put an air trap - it doesn't always have to be the highest point in the system. (hmm..., I'm gonna get flamed for this I think...) Air will not flow down in order to get around an obstacle in the system to find the highest point. If there is a bulge at the top of your res that is higher than the ends of any of the inlets/outlets then it will collect air just as well as anywhere else. And pump going or not the air will stay in there unless the system is tipped on it's side or something, such that the air rises to a different point. I wouldn't worry about putting a second air-trap before the rad - just make sure to bleed the air out of the system as much as you can (by moving the rad/cpu block around while the pump is running before they are mounted - keep getting air bubbles into the res and topping up the res) Then close the fill-hole so the system is sealed shut with very little, if any, air in it. The tiny bubbles that are left will flow around the system with the water, and will rise up out of the flow in the spot where the flow-rate is low enough - this should be in your res where the water comes to almost a complete stop - the tiny bubbles will get a chance to float to the top.
As shown in bad pic below - if all of your hoses come in to the rad far enough down to be submerged, then even with the pump off no air will enter any of the hoses - no re-priming the system, no air in the rad, etc. And because the flow-speed in the rad when the water moves from the inlets to the pump inlet will be VERY slow, it will allow even the smallest of air-bubbles in the water time to float to the surface, and get trapped along with the rest of the air. Also in the pic there is that large empty thing on the right - this is your "fill cap". As I mentioned in an earlier post if you attach a hose here and run it up, and do NOT have any length of barb sticking down into the res (the stuff in red is what you do NOT want - it will trap an air bubble in the top of the res) then you can fill the res right up with water, and fill that tube up a ways as well. If the top of that tube is high enough and has more mass in water than the rest of the system that is above the res (so the block and rad and their tubing all combined) then the water will not flow up that tube if the top is removed and the pump is not running. Whether you have that tube or not though - you should try and keep as little air as possible in the top of the res, and for the first month after you put it together watch the water level closely - as the smaller air bubbles slowly accumulate the water-level will drop. If it drops below the ends of the inlet tubes and the pump is shut off, air will go up the inlet tubes as water comes down through the pump and then you will have problems (need to re-bleed the air out again). So make the inlet tubes go quite far down, and keep the water-level topped up.
|
Let the flaming begin!!!
You're right, as long as that res is sealed, it would be considered a closed loop. However, if the inlets extend in the water, and the res is open, it is still a closed loop (I think you got that!)
You're also right about the airtrap not necessarily working at its best on top: as I explained, the Swiftech block (as an example) will hold air, if any is left in it: I used the term "prime" improperly before, I meant that you'd have to do the whole tilting thing.
The inlets don't need to extend down very far, in the situation where the res is open, because the system becomes a closed loop again, once the water line goes up (as a result of air being sucked up the tubes) and reaches the inlets again. It would cycle over and over, sucking air then water. The noise would be a warning on its own, to refill.
it seems to me that that bottom res, if it's going to be sealed, might as well be full of water.
I hate cutting holes unneccesarily. I think that if you could drop the rad an inch, you'd have room for a res without cutting anything.
Alternatively, Cova's idea of extending the bottom res opening all the way up, would work very well. You could just leave a tube that runs up the case. No holes, just take the cover off to reach the tube for a refill. No extra airtrap, and leave that tube half filled. You'd have to tilt the res to let the air go up that tube though...