Upgrade itch - loop order questions

Cenedd · 29 Oct 2014 at 16:47

I'm feeling the urge to give OCUK some of my hard earned...or more accurately, I want new toys

I've got an existing Laing DDC 18W with a slightly damaged XSPC res top. Also have a broken Scythe fan controller - screen dead - which means I can't see my temps. I want:

1. to be able to read and control fans and the pump so I'm thinking of the Aquaero 6 XT which should handle 30W on a channel. The pro doesn't quite hit sate that new toy syndrome itch

2. Get flow indication and alert or failure - so I'm thinking the Flow Sensor MPS attached to the Aquero.
3.Read fill level from the res so I don't run dry and can detect leakage - The Aqualis XT 100ml ought to do that. Would also need a new DDC top to replace res/top combo.

Looking at laying things out, it might help if I go: GFX - Res - Flow - Pump - Rad - Rad - CPU
Does it matter much if the flow meter is before the pump? Either for starving the pump or for the reading? Depending on juggling it round, I might be able to do: GFX - Flow - Res - Pump- Would that be better?

I've got an MSI 580 Hydrogen GFX card. Does anyone know if it matters which way the flow goes through the block on it? Just thinking it may be easier to pipe if I can reverse it.

Many thanks,

Gareth

cahonis · 30 Oct 2014 at 08:31

To the best of my knowledge flow is the same as temps in your loop, i.e. it doesn't change by very much at all throughout the loop.

Not familiar with that card but if the block is like all the others then flow direction shouldn't matter at all.

Cenedd · 30 Oct 2014 at 11:14

Thanks Cahonis. I'd seen people saying similar things in articles from about 2007 but there's so many shouting the opposite that it's difficult to know who to listen to. Just read an benchmark too showing that tubing diameter has very little effect on temps too....which saves me some upgrade itch money

Had a look and cannot find any info from MSI other than how to stuff card into a PCI-E slot - thanks for that MSI - but it does seem to be a re-branded Watercool Heatkiller GPU-X3 GTX 580. Pics here: http://hw-lab.com/watercool-heat-killer-gpu-x3-gtx-580.html
Looks like the only effect would be that it cools the RAM before the GPU. I'm figuring that because of the path the flow takes that it would have some effect but probably so minute that I'd be lucky to spot the difference with lab-grade equipment. Does that sound about right?

cahonis · 30 Oct 2014 at 14:17

Water temps don't vary by more than 1 or 2 degrees throughout the whole loop so I don't think you'd notice the difference between gpu->memory or memory-gpu on that card.