First DIY H20 Job & Questions

[eva2000]

Finally put all the pieces together to leak test but think I done somethings wrong or could of done better :o

Silverprop Cyclone SE P478 waterblock
Clearflex 60 1/2" ID tubing and Norma hose clamps
Bravo Heatcore + 2x 120x38mm Panaflo 109cfm fans sucking off the shroud
MCP600 rev1 with fixed pump head
PCcasegear cyclinder res with 1/2" barbs vertically standing
Nulon Ultracool 1/10th of 2L mixture

520W Topower Gold Triple fan PSU with paper clip to short green ATX and black to start psu by itself.

multimeter reads the PSU's 12V as 10.51v

1. i think i used too much tubing length as i planned to setup the DIY stuff externally on the floor and the bench setup is up high on desk - see below for extact setup
2. i can't see if the water is moving. I did see some flow in the res barbs and bubbles in the little gap/empty top area of my res but now there's no bubbles to be seen. How do i check to see if water is flowing ?

setup is like this


pump inlet -> 15cm tubing -> res's lower barb - res's upper barb -> 35cm tubing -> radiator -> 200cm tubing -> waterblock -> 200cm tubing -> pump

- could it be the under volted 12V rail on psu is lowering the pump's power to pump the water around as it took ages filling the res and getting the water to fill all the tubing ?
- too long a tubing ?

I'm leak testing now, but with my above setup listed, if there were leaks where would I most likely look for them ? I've screw in tightly the Norma hose clamps and i doubt they'd leak.

oh and how the hell will i change waterblocks or add the flow meter to the setup ??? or change the water ???

help is very much needed and appreciated :)





.

[RotorHead]

Eva2000

Looks like you have some good hardware there.

1.Try and get your pump up on the same level as your system. The pump you have probably doesn’t have a enough “head height” to give you adequate flow rates.

2.Try rearranging your flow pattern as follows:

Pump -> Block -> Rad -> Reservoir -> Pump

This will give your radiator it’s highest temperature delta allowing it to work more efficiently.

To change blocks, just leave enough tubing so you can remove the block from the motherboard without disconnecting the hoses. That way you can pull it out away from the case to change the block.

Hope that helped a little.

[eva2000]

thanks

somes pics of the leak testing and tube length -- too long ?

http://www.fileshosts.com/watercooling/DIY/1/

http://www.fileshosts.com/watercooling/DIY/1/DIY_06_thumb.jpg (http://www.fileshosts.com/watercooling/DIY/1/DIY_06.html) http://www.fileshosts.com/watercooling/DIY/1/DIY_09_thumb.jpg (http://www.fileshosts.com/watercooling/DIY/1/DIY_09html) http://www.fileshosts.com/watercooling/DIY/1/DIY_10_thumb.jpg (http://www.fileshosts.com/watercooling/DIY/1/DIY_10.html) http://www.fileshosts.com/watercooling/DIY/1/DIY_11_thumb.jpg (http://www.fileshosts.com/watercooling/DIY/1/DIY_11.html)

notice the empty space in the res and that's with pump running/on

[RotorHead]

Looks like you got the flow pattern setup right. :thumb:

With the 1/2 ID tubing, you should be getting great flow. :yup:

I just noticed your voltage from the first post. 10.5v does seem kind of low. I would try and get that up around 11.5-12 if you can. :scratch:

[Moonwizard]

No, never do the pump -> block as the pump will give the water some heat. Maybe not much but it can depend after awhole day on. So keep it the way you got.

It's probably like Rotor said. You got to much tubing. If you got space on your desk have it at the same height and next to the computer. If it ain't possible then maybe you are able to put it inside your case. Otherwise I've seen ppl putting it in a separate case. :D Hopefully that helps. Otherwise there should be more at abxzone.com that knows about water. Like Skully, Gasper and some more. :D Try looking for the *back to water* there. It's like 65 pages just about water cooling. Good luck mate. :D

[Flamer]

Eva,

I think swifty recomends having the pump at the lowest part of the system. I have gotten the the swifty pump to prime and flow well having it above the water block but it it took a long time and my pump did fail after only a couple months of use.

[LostInSpace278]

As far as checking the flow, I personally know of 2 options. They make anInline flow indicator (http://www.watercooling.de/catalog/product_info.php?products_id=103&osCsid=2a3c65a77e7e57def7e4fe3848833913)
Option 2 is a radical one, but I have heard of people doing this. Before you do the actual leak testing, start pump, let it run for a minute, and pull a hose. If everything is running fine, then it should look like you severed an artery. A bit extreme in my opinion, but hey, to each his own.

[Moonwizard]

I feel on the pump and check my temps to see if the pump is vibrating a bit and the temps are stable then I know it works. I also see when I need to refill when I get airbubbles in the hosings. :D

[Hawk]

Looks good Eva2000. . . . :thumb:

[Moonwizard]

Awesome work by the way. :D

[eva2000]

noooooooooooooo..... bad news.. i install the whole setup onto my P4C800-E/2.8C test bench, turn on the system and i can hear the rad's fans spin up.. but i feel the Silverprop waterblock and it starts to heat up!!!!!!!!!! I go into bios and stock 2.8C@200fsb and 1.525v and the cpu temp is rising to 58C! :eek:

As people know P4C800-E already under reports cpu temps by 5-10C so cpu was really at 60+ C

A few bubbles in the tubing near the waterblock barb don't even move.

Sooooooo.. that means the MCP600 pump isn't even working right ??? or is the tubing too long ???

Arrrrrrrgh...

[Snafu]

Could be either (but I am not a water-cooling pro). Can you add a small amount of food colouring to the reservoir and see if it moves through the system when the pump is on? This is the only easiest suggestion I have to see if the coolant is moving through the system without taking it out again. Might help if you could test the system without powering it up?

It is possible that the amount of water (and its mass) is too large for the pump to move and/or there are some restrictions in the system that the pump cannot overcome. I think this is fisshy because the pump looks like it has 1/2" fittings so it should be designed to use 1/2" tubing and handle that kind of volume/mass.

The only other suggestion (and I hate to offer this cause you have to pull the whole thing out) is to test it in the sink as an open system. Open only one section first and if nothing then remove a component and try again until it is flowing.

[Moonwizard]

Yeah, do what snafu suggest. Hm, do you hear any sounds from the pump at all?