Where does the water go, in a closed loop system??

[dabhpr]

I'm fairly new to the watercooling scene, having converted my PAL8045 system to watercooling (specs below in sig) back in January of this year.

I have an inverted "T" fill tube for my closed loops system. It's about 18" long (hey, this Supermicro 760A case is HUGE).

The funny thing is, I'm loosing about 1 1/2" of level in the fill tube every four days or so. I have NO leaks that I can find, none whatsoever. Every single connection is secured with a screw clamp.

I'm using the 1/2" ID Silicone tubing from Dangerden.. is it actually somewhat porous??

I'm very positive I don't have latent air pockets in my system, as I was very thorough to fill it underwater, and invert/twist/rotate the entire system components to evacuate all air bubbles.

And my temps are great: 32C max at full load.

So - where is the water going???

[Avatar28]

In all honesty, my guess would be that either A) you've got a small leak. Not much, but a little bit. B) it's not airtight and some of the water is evaporating. C) there is still a bit of air in the system and it's working it's way out. Sometimes tiny air bubbles will stick to the metal, even with decent flow. These air bubbles could eventually be washing out to get caught in the airtrap.

[Brad]

avatar is right, all 3 of those could be possibilities, and together they add up.

do you have any pics of the alpha?

[futRtrubL]

Do you have any dyelight or coloring agent in the water? If not you may wan't to put some in so that you can see if there is a leak that's so slow it's drying up almost emidiately.

Edward

[dabhpr]

Yes, I have florescent yellow dye in the system

But I still don't find ANY drips or leaks in the system.


I'm convinced that it MUST have a leak - or the pump is leaking air IN.

Guess some sleuthing is in order, eh?

[DigitalChaos]

uhh if the pump was leaking air in.... how would that make your water level go down?

[dabhpr]

Well, perhaps my suggestion of Leaking Air IN is in error. I was just thinking that if the pump is introducing air into the system, it would eventually end up in the fill tube... I wasn't thinking of the fact that it would actually be increasing the pressure of the system. Which is NOT happening, since I have one of Dtek's pressure gauges inline, too. It's still showing about 4 psi, no increase since it was built.

Still, I wonder where the heck the water is going, or exactly what is happening...

[MeltMan]

Air in the radiator. I had the same thing happen... all those little fins kinda hold the bubbles of air, even after swooshing it around at all angles. Just keep filling her up and eventually it will stabilize.

[redleader]

Effusion. You have a tiny leak somewhere and its letting air in a little at a time combined with all the air in the rad to give you the 1.5inches every few days.

I had the same problem. After monthly fills for 6 months, I went and replaced those plastic DD clamps with much better spring clamps. Worked nicely.

[Haddy]

might try putting some paper towels or something in the bottom of ur case....so if any water is leaking out ull be able to c where the dye hit the paper

[max007007]

OT:
I have some questions to ask, following are 2 ways to connect up a watercooling rig.

Method 1
Pump w/res --> radiator --> waterblock

Method 2
Pump w/res --> waterblock --> radiator

Which way is better and why?

Is a "T" airtrap required for the above setup?

[Brad]

25% of people will tell you one way, 25% will say the other, 50% will say it doesn't matter a damn

[Jim]

Originally posted by redleader
Effusion. You have a tiny leak somewhere and its letting air in a little at a time combined with all the air in the rad to give you the 1.5inches every few days.

I had the same problem. After monthly fills for 6 months, I went and replaced those plastic DD clamps with much better spring clamps. Worked nicely.

I am losing (still) about 1/ 1/2" of fluid for a two week period. I have 1/2" ID silicon tubing. The system has been up for over two months now, I doubt it is air. This 1 1/2" of fluid BTW accounts for about one (1) teaspoon of liquid. Still and all I can't see any leaks anywhere either.

Did you use the "band" type hose clamps?
Jim

[EMC2]

Jim - with that small of a volume of loss, I would place bets that most of it is simply your tubing

Most silicon materials can be considered as porous (it isn't really porosity per say, but permeability). This is why 'common' silicon tubing gets discolored from additives. It is also one of the big reasons that we don't use silicon products at work for sealing electronics from moisture and contaminants (along with issues of out-gassing).

Silicon materials are actually also used as permeable membranes because its makeup. It allows diffusion to occur as a result of it's physical structure at the molecular level.

But 1 teaspoon every other week isn't a biggy

[dabhpr]

Alright, I'm trying an experiment: I'm going to replace the 5 remaining plastic "teethed clamps" supplied by Dangerden with metal screw clamps, just in case they're not actually sealing as well.

I'll check back in a week to report if it has made any difference...

Wish me luck.

[redleader]

I found the gearbox metal clamps worked worst of all because they weren't quite round thanks to the gear.

These (the red things) are perfectly round and much tighter fitting.

[bmg]

I find the cheap plastic cable ties work great. I usually put on 2 just to be safe.

[mkosem]

I can see a drop of water underneath the hose to the right

--Matt

[Jim]

Originally posted by EMC2
Jim - with that small of a volume of loss, I would place bets that most of it is simply your tubing

Most silicon materials can be considered as porous (it isn't really porosity per say, but permeability). This is why 'common' silicon tubing gets discolored from additives. It is also one of the big reasons that we don't use silicon products at work for sealing electronics from moisture and contaminants (along with issues of out-gassing).

Silicon materials are actually also used as permeable membranes because its makeup. It allows diffusion to occur as a result of it's physical structure at the molecular level.

But 1 teaspoon every other week isn't a biggy

Emc2
I read this elsewhere. I thought this could be part of the problem. I wonder if this is why the stuff smells too?

I am going to go over to Tygon I believe it will be less permeable.

THanks
Jim

[Jim]

Redleader-
Where did you get those red clamps? Thanks for the snap.

Jim

[dabhpr]

Originally posted by mkosem
I can see a drop of water underneath the hose to the right

--Matt

Not to mention a big air bubble in that same section of hose. I think the system is off for this photograph...

...and not completely purged of air yet!

[EMC2]

Jim - those red clamps in the pic are from the world of car engines I personally always disliked them for the following reasons -

Without the proper tool (yes, there is a special kind of plyers made for them) they can be a bear to get off (at least in a cramped engine compartment). You have to compress the 'tangs' with pliers, keep them from slipping out, and slide the clamp all at the same time.

The other aspect I always disliked about them is that they didn't seem to hold up well to repeated removal/installation (may partially have been caused by not having the aforementioned special tool). The 'tangs' would tend to get bent and the clamp itself would tend to get distorted and no longer apply even pressure around the circumference of the hose.

I still prefer good ole fashion "rachet clamps" (the automotive hose clamps with slots evenly spaced around the circumference and a screw that keys into the slots and tightens/loosens the clamp.

[Brad]

I like those clips too, I find with all clips though that over time they get less round, especially the screw type, around the screw part itself they distort really quickly

[redleader]

Good pliers are essential. I've tried to take one off with my bare hands. Not fun. Ditto for needle nose pliers.

Mine seemed to wear a little with use, but even after 4 or 5 uses they seemed pretty much the same (though scratched up from my pliers). If that bothers you, they cost about 25 cents locally. Thats 1/4 as much as gear type clamps, so replace them.

can see a drop of water underneath the hose to the right

The fill on my res is just above that and I'm messy

[DodgeViper]

I have the same problem. System keeps using water and can not find any water anywhere. I even ran the pump over night with no fans on and still can not understand where the water is going. I have had the rad out a number of times and everything is dry. The only thing I have not tried is replacing my screw on cap at the end of my T line. Maybe I am losing water due to evaporation. I have taken a tissue and checked every fitting.