Close call...

I just returned home from dining out, to find my monitor on StandBy and my PC's power led and fans still whirring. Bad sign.

My first thought was that water had probably seeped out of the T-Line, over the vidcard, thus shorting it out. Upon opening the case, it was clear that this indeed happened.

I had re-tooled my setup a bit last week and it was running just fine. So today I wanted to seal the system off at the T-line. When I did this all seemed just fine. However, after about 30 minutes I noticed that there was some foam starting to develop in my tubing. Surely an indication there is air seeping in somewhere. Correspondingly, my T-Line water level started to rise.

I've been working on the system the whole day, tightening all clamps. I can't hear any air flowing and there is no water leaking either. However, I remained unable to seal the system off. Everytime I tried this, it would end up foaming and overflowing again. Often not immediately, sometimes only after a full hour of problemless clear-water running, but still.

So I decided to forget the seal and stick with the loose cap on the T-Line. This had worked fine all week after all. And it did now. Until I went out -hours later, apparently.

I have not had the foaming problem without having closed off the T-Line completely before, but now it did happen. Luckily, my videocard seemed unharmed and it runs fine now after having carefully dried it (hairdryer). There was only a tiny amount of water on one corner, so that may have limited the impact.

The system is running again now, and the foamy tubes are clearing up. I've got the T-Line fully open now and I tightened one clamp some more.
I'm not fully confident it will run ok now though, not after this sneaky turn of events. The problem is that I don't know where air could be getting in. Everything is tightened down pretty well. It may even be somewhere inside the radiator... Wherever it is, it's very small (since the foaming thing is so slow) and next to undetectable. You can imagine I'm kind of frustrated about this state of affairs.

On the bright side, the 7v mod I just did on my Sunon howler was successful and painless. And it distracted me from the watercooling dilemma for a few minutes.

Wow, close call! Glad to hear you didnt lose anything.

Would having the end of the T-line outside your case a good idea? would that have prevented any water from spilling inside of your case at all?

Yeah, I had a similar experience with my 9500. I had a lot of water spill all over the back of the card, around the core area, and some also dripped through the heastink mounting holes and onto the other side of the card. I flushed it all out with rubbing alcohol, and did some gentle scrubbing, and luckily it's ok now...phew.

I'm just now setting up my revised system for leak testing...so we'll see how it goes. I'll keep leaks even more firmly in mind.

johan851 said:

Yeah, I had a similar experience with my 9500. I had a lot of water spill all over the back of the card, around the core area, and some also dripped through the heastink mounting holes and onto the other side of the card. I flushed it all out with rubbing alcohol, and did some gentle scrubbing, and luckily it's ok now...phew.

I'm just now setting up my revised system for leak testing...so we'll see how it goes. I'll keep leaks even more firmly in mind.

Click to expand...

In my case it wasn't a leak though. I'm sorry if I was not clear; it was the T-line overflowing.

I've tested for leaks pretty well before final system assembly, but then again that goes also for this foaming + overflow behavior... it seems to be a very tricky issue. My current plan is to babysit the sytem for a few days at least. If this works out (still looking fine now), I'll close the case up again without touching the watercooling at all and then I'll hopefully be able to forget about it.

Since osmosis doesn't apply here, it's safe to say that air getting into your system is occuring at a low pressure (relative vacuum) point. This is typically at the pump housing, inlet barb, and reservoir (though you don't have a rez). Check your pump fittings, and re-grease the o-ring.

Anytime there is air in a system it's before the pump as having a leaky spot AFTER the pump would be a leak with water coming out due to pressure. What's your loop setup like?

What kind of blocks you using? HAve you watched them to see if you may be seeing a few bubles at the inlet or outlet? Could have a small leak that's actually "sucking" air. Yeah, you can get leaks in these types of sytems that actually don't lose water but suck in air into the sytem.

I had a Maze4 that had this problem. Constantly had a small stream of bubbles coming from the inlet side under the barb close to the edge of the plexi top. Thought I might not have had my system bled well (I use a closed t-line) so I let it run for a few days to see if I gained air bubbles in the lines. After 3 days I had considerable bubbles in the lines and removed the Maze4. The bubles disappeared and the system never gained air bubbles after that. Seems I had a VERY tiny leak in the Maze4 that was allowing the system to pull in air bubbles. It DOES happen.
After the pressure builds in closed system, why you had to leave off your t-line, then leaks usually start at the point of air entry from the pressure that builds from the air being in the closed system

The thing is, there is absolutely nothing visibly happening. The water is totally clear, there is no foam anywhere in the system. There isn't even the slightest amount of water loss, nor increase.

Then, magically, the lines slowly -very slowly- become clouded and the water level in the T starts to rise. There is no telling where the air enters the system I'm afraid. You can't see any separate bubbles inside the system, just a haze.

As I said, I've checked all connections pretty well, including the pump inlet. That one is actually probably one of the tightest connections of all- it's 1/2" ID Tygon stretched over the 3/4" barb and sealed off with a properly sized steel clamp.
There was one issue earlier, when I was assembling the system. I was able to hear air seeping in. This was happening somewhere near the waterblock, which is still about 8 connections though. I proceeded to tighten the clamps there and the sound vanished and so did the foaming. In any case, this previous event leads me to believe that there is a good chance that the current problem is somewhere in the same area. I can't hear it anymore though.

BTW My system specs are in my sig, and the loop is as follows: pump on my case bottom pumping to the radiator in the lower front, from there to the waterblock with the T-line in between. Finally, from the block back to the pump's inlet.

Try this, take the whole loop out of the case, pump and all, and with it running, submerge the whole loop in a bucket or large bowl of water then add a drop of food coloring to the water near the questionable areas (pump inlet, both water block fittings, T line fittings. If there is a leak somewhere and you have the whole loop underwater it will be pulling water in instead of air and the food coloring will be drawn to it. This should work, in theory.

Fast420A said:

Try this, take the whole loop out of the case, pump and all, and with it running, submerge the whole loop in a bucket or large bowl of water then add a drop of food coloring to the water near the questionable areas (pump inlet, both water block fittings, T line fittings. If there is a leak somewhere and you have the whole loop underwater it will be pulling water in instead of air and the food coloring will be drawn to it. This should work, in theory.

Click to expand...

Hehe right. I'd imagine having trouble with seeing the color of the water on the inside through a bucket of colored water though. And if the process is the same as it is with air it would just turn colored altogether, without indication of the source... but thanks for the inventive idea.

Anyway, I haven't been sitting on my hands and I've replaced some of the clamps I was using. I had 22mm ID clamps on some spots, which turned out to be slightly problematic. I replaced these by 15mm ID clamps (both steel types), which seem to fit much better on the smaller barbs (like those on the Y-piece). Refilling the system was completely problem-free this time, so I have good hope that the foaming thing will also stay away now.

gl hunting that down... thats one reason, of many, i put the t-line outside the case.

very frustrating im sure... ahhh... the thrill of watercooling!

poMONKey said:

gl hunting that down... thats one reason, of many, i put the t-line outside the case.

very frustrating im sure... ahhh... the thrill of watercooling!

Click to expand...

Haha yeah. I'm doing fine so far though. The new clamps seem to have done the trick. I've been foam-free for 2 days now.