Loop finished, only one leak so far, no disaster...

I have d5 and you can deffo tell when it's been increased in speed. You could try running a little loop out of the system with just the pump and a small res etc and see how much it's actually pumping, at least it'll rule out a faulty pump.
 
I've requested an RMA now, in all fairness they've been very helpful. Had to strip the whole kit out, and run off the IGP. No games, just internet and basic multimedia. No cards in the rig, as they both had blocks on and one of them is going for RMA somwhere else. Looks very weird. First time I have ever used the chips IGP, and to be fair, it's a godsend.

I think I have been the unluckiest first time watercooler in the history of the forums. Here's hoping the rest of the process goes relatively smoothly.
 
I got in touch with XSPC, the lad there reckons I need two seperate loops, running two D5's. One 360 for the CPU, and then another 360 and a 140 for the two cards.

I went for a 2nd opinion

XSPC told you that you require two loops and a 360+360+140 setup with dual D5's???

No disrespect to XSPC they make great products but what they have told you is complete ****. If anything it sounds like they are trying to sway you into buying more product then you actually need.

I think I have been the unluckiest first time watercooler in the history of the forums. Here's hoping the rest of the process goes relatively smoothly.

Agree with you on that buddy, you may want to grab some extra tubing and check each component before fitting them, so you can check for flow restrictions etc...
 
This is the email I had from XSPC:

Hi, Mark

I will suggest you can separate them to 2 loops. Especially when you overclock the CPU.
1xD5 + CPU + 1x 360 rad.
1xD5 + 2 xGPU 580+ 1x 360 rad +140rad.

1xD5 for 2 rads 3 blocks. The performance will difficultly be every good. And 580s is hot.

Or, you need to the the rads outside of the case. It could improve a little bit.


I've got some spare tube. When I get all the kit back, I'm gonna try just that. Even though I jumped in at the deepend with no arm bands going for the 360 kit first, in the very least, you expect the stuff to work.
 
580's are hot, but not that hot :p

It doesn't sound like sound advise to me, it sounds like overkill with a side order of "buy some more of our products".

Especially as myself and many others have ran big big loops from a single D5 and not had any issues. I had a 4 block loop with 3 radiators totalling 1440 of radiator space with around 4 meters of tubing and a large 2L res all from a single D5, I didn't hit heat issues.

My personal feeling is the issue lies either within the block or within the pump/flow restriction. Reminds me a little of when I forgot to plug my pump in, first sign of stress everything near instantly overheated and shut down.
 
Yep :p

If it helps I'll be pressure testing all of my individual loop components before I rebuild into a new case so I'll be doing a similar thing to you. So I can always give you some pointers on what to look for.

What is the exact model of GTX580's you have and what is the exact model number of waterblock you were using on them? As I'm sure EK made a few variants for one reason or another.
 
I think I've found your problem....

According to EK they made specific blocks for the Gainward 3GB cards (Phantom).

The blocks you have 'EK-FC5X0' are for the standard reference version GTX580's and a select number of GTX570's.

The blocks you require are called 'EK-FC5X0 GW'

The odds are the cards you have, have a capacitor placed somewhere different to reference, which is preventing the blocks making good contact with the card, it has to be a minor difference as you've managed to partially fit them. Could also explain how the IHS managed to get dislodged earlier.
 
ahhhh :p

I thought 'BINGO!!' then as well :(

My hunch is something to do with the pump and or flow restriction in that case.
 
Yesterday, when I changed the L bend fitting to a red curved bits power fitting, I got an air lock between the cpu and the 360 rad. It hadn't happened before, and no amount of tipping the rig left or right seemed to move it at all. I don't know how it happened. There was obviously some kind of restriction, but as to where it came from, I have no idea. At one point I had the rig 90 degress on it side, and coolant seemed to move then, but before that, nothing.

Right now I'm torn between asking the lads at OcUK for a refund on the kit, and the 140 rad, and just buying a dual GPU card. Selling the 580's to who ever wants to put blocks on them, and just going back to air. Honestly. I don't even know if when I get the kit back, and the new block, if that will work properly. And then there's the hassle of stripping it all down again, then putting it all back together. It is really worth it?
 
It depends what you wanted to benefit from watercooling in the first place, some do it for silence, some for performance, some just for looks!

Unmovable airlock says to me that the pump isn't working hard enough, like its set on speed '1' regardless, the dials in the back of the pump become ineffective if you turn them the wrong way as they dislodge, I know as i've done it my self :p Luckily enough a firm turn in the opposite direction re-mounted the dial.
 
Well yes, it did seem to trickle down into the pipe that was going into the CPU, like a drip drip drip from a tap.

I went with W/C for the quiet rig aspect. Now I'm wondering if the blocks were at fault, or not, or it was the pump that was causing the whole problem from start to finish.
 
My money is on the pump but the best thing you can do now is hook the pump to a single block, no rads, outside of the case and run it to see what the flow is like is a best possible situation.

Unfortunately when something go's wrong with WC its often a complete nightmare to diagnose.
 
Given that you assume the blocks were fitted correctly, then poor flow from the pump, if any at all, would leave the coolant/fluid sitting around in one area for longer thus giving much higher temps, which also goes with how you say the fluid looks as it if is just dripping through the pipe.

At this point as said in previous post, i'll go with faulty pump. if you can test it on it's own out of the system then you can see how it's performing.
 
You need to keep the pump fed though, so you can see it constantly working, or not as it may be. Don't let it run out of fluid.
 
Back
Top Bottom