why would a hose collapse?

I'm a long time overclocker, but a newbie water cooler. I bought a pre-assembled Swiftech Q-Power water cooled case. Yeah, I know, I probably should of built my own system, but laziness got the better of me.

This is a new setup and when I initially started running it it seem to work ok for about 10 minutes. Yes, I jumpered the relay at first and ran the pump. It worked ok and I saw no air bubbles. Yes, I put the relay wires back in the right connectors afterwards.

Now, something ain't right, I have the mobo alarm set at 60 degrees for safety purposes and it'll trigger the alarm shortly after starting windows. I've checked the waterblock contact with the CPU and it's ok. I found an inlet hose going into the pump partially collapsed. I'm assuming this is restricting flow.

I'm a newbie with water cooling, so before I replace the hose, I need to find out if there is some reason that pressure or suction is causing the collapse or maybe the hose is too thin and just partially collapsed at that bend where it goes through a case channel.

Anyway, the waterblock is a MCW5000 and the Swiftech website says it uses 1/2" OD tubing. I have no idea how thick the tubing is or what type was used.

Look at this pic, the place the tube is collaped is at the far left past the valve where it makes the bend going into the case on the way to the pump.



Any ideas?

It came like that, straight out of the box? Dissassemble everything and have it run in a straight loop... as if there was a CPU in the mix. Treat it like you are leak testing it. If the pump's input is strangeled by lack of flow it could create that kind of collapse... although I'd think it would equalize eventually. You need to make sure everything runs like normal before you put your system in the case... make sure there is plenty of flow and you can operate without any kinks. Then, build the components in the 'verified' working system to make sure you aren't bending the tubes at any junction.

It sounds like the pump is starving. Look for any indication of air pockets and flow impedance.

Thanks for the input. But, keep in mind I'm a total WC novice and some of this is sort of over my head. Like, this straight loop?

Don't think the collapse was in the hose before I started it. I looked things over, but since this area is half hidden, I could of overlooked it.

It sat overnight and it's still half collapsed. Would adding more water equalize the pressure? I have checked for air and have not seen any bubbles at all.

If I have to replace this section of tube, how do I drain the system? Sorry in advance for being an idiot.

Is it just pinched back there, like if you hold the house out so it isn't as sharp of a bend does it still do it?

Adding more water wouldn't solve it I don't think.

If that kink is severe enough, it could indeed be the cause of the overheating problem. A few things are assumed though; you had your pump running, the fan on the radiator too, no air inside (also no white haze) and the block made good contact (which you said you checked, so ok).

It's hard to see from the pic how bad the kink is. A hose collapses when it is forced into a bend that is too much for it. The inner side will get compressed too much and the outer side stretched out, resulting in a kink.

I'd focus on alleviating the kink thing first, so I suggest you rig up some zip ties over the troubling area on the tube. A zip tie can help the tube to maintain it's form. So to be perfectly clear: just encircle the kinked spot with a zip tie (or a few if the tube gives in at another place too), while holding it so that it can reach it's natural round shape.
This should work well enough to see whether the temperature problem goes away or not.

Well from my experience if your running the new MCW5000-P and MCP600 pump and the pump is sucking the hot water from the block I also had a hose collapse but not to your degree. The only thing that fixed it was to put a reservoir after the water block and let the mcp600 suck out of that and no collapse so far. Maybe the water just can't get out of the block fast enough?

Forgot to mention I'm using 1/2 id Tygon tubing.

Yup put a zip-tie on your kink and tighten it until the tube becomes round again. This is one way to fix it... I don't know how thick the tubing is but my 1/16" wall kinked pretty damn well and my 1/8" did not kink at all so I would also check on the thickness and change the tubing if necessary. The quickfix is the zip-tie.

so hows your 300FSB journey going so far? cant wait to get that WC up and running eh?

/edit/ another thing you can do is shake your system around (turn it upside down if you can, turn it every way possible) and see if any bubbles will appear. bubbles to have the tendency to get stuck in the pump's impeller chamber soo check for bubbles. I suggest you do this with WC system taken out of the case...

Thanks all, I'll try some of your suggestions. I've measured the fill tubing and it's 1/2" OD (3/8" ID) and it's only 1/16" thick in the wall. The more I look at the tube, the more it looks kinked rather than collapsed. The dang tube is a short section... almost too short I think... and there is no room to reach in the tight space to work on it. I doubt I can get in there to zip tie it because of the tight quarters. I might be able to use a small automotive metal worm gear fuel line hose clamp and slide it down into place. That might bring the hose back into a rounded shape as long as I'm careful not to over tighten it. You guys rock, thanks.

Yes metal clamps actually do work better than zip ties cuz they are bigger. as 1/16 is quiet thin, just be careful as you said. I did this before when I had 1/16 and it worked.

BTW, how bad is the kink? It would be nice if we can take a closer look at it.

Well, I did the clamp thing and I was able to restore most of the kink back to being sort of round. It did help a little. I went from about 60 to 48 degrees idle CPU temp. However, full load temp was 58 degrees. I know Abit mobos read a little warmer than most, but that still seemed too hot and was unacceptable to me.

I puzzled and thought and poked around. I checked again and there was no air bubbles in the lines. Water seemed to be circulating good now. I put my hand in front of the fans to check air flow. The side cover intake 120mm fan was moving air like crazy. The front 120mm intake and the two rear 120mm exhaust fans that blow through the radiator seemed wimpy. Hardly any air at all moving. Now that's strange!

I checked my power connections. I have an Antec 550 watt True Power PS and there are two "fan only" molex connectors. Well I had the rear fans (which are wired together) connected to one and the front fan connected to the other. The side fan that was blowing like a hurricane was connected to a regular harddrive 4 pin molex connector. Ah ha!

I pulled out my PS manual and sure enough, the "fan only" connectors are thermister controlled power. Well, since the power supply was staying cool way up at the top of the case with the top of the case off (so I could work on stuff), it was automatically reducing the fan RPM to cut down on noise. Holy guano!

I connected the fans up to regular harddrive power connectors and fired it up. Big difference. Now all four of those 120mm bad boy fans were moving plenty of air. My full load CPU temps are sitting right at 40 degrees now. Wow! That's an 18 degree difference. Sheesh!

aaaa glad you finally figured it out. Hope you reach 300 soon. If you do ill be joining you soon when I get my SMD grabbers hehe

Heheh air flow through the radiator is a very important factor in the overall performance of the system. A huge difference for sure in your little experiment.