Vapouriser, My Very First Watercooled PC........

[Syd]

Looking at your second picture. The blue impeller looks like it is touching the XSPC top.
(above the XS and between the delivery thread)
There is very little clearance in this area and there has to be a very small gap.
If there is no gap then the top will rub on the impeller, causing a lot of noise and heat.
Undo the screws and push the top forward to increase the clearance.
There should be only enough to see a very small amount of light between the two parts.
Hope this helps.

Thanks for that piggin I think I get what you mean.
Have moved it slightly.

[Syd]

Late night last night so haven't had much chance to get much done today - but while we caught up on some sky+ stuff and relaxed I dabbled with a little braiding - think it looks OK.







Slightly concerned about fans though - planning to run 4 x Xigmatek 120mm XLF-F1254 Blackline 1500RPM on the 120.4 and want to control them off of 1 channel on my Aquero. I think it can handle that sort of voltage
If anyone can confirm though that'd be great

Also 4 fan wires into 1 Female fan pin - will that work OK?

[Syd]

accidental double post ooppss - best get some sleep

[piggin23]

http://www.chilledpc.co.uk/forum/showthread.php?t=5312

http://www.watercoolinguk.co.uk/foru...ead.php?t=2410

As for running four fans of one channel, there will be no problem.

[Pob255]

There may be issues

If this is the fan controller Syd is planning to use
http://www.aqua-computer.de/e_index.htm

Fan-Connections: 4 outputs with rpm-signal, max. 10W/channel, 1A

4 fans may be too much for it and seeing how expensive these are killing it would be a costly mistake.

If the fans are 0.25a or less you should be ok if they are more then I'd contact aqua first.

You could safely run 4 fans off a psu molex but then you'd lose the speed control.

[Syd]

There may be issues

If this is the fan controller Syd is planning to use
http://www.aqua-computer.de/e_index.htm

4 fans may be too much for it and seeing how expensive these are killing it would be a costly mistake.

If the fans are 0.25a or less you should be ok if they are more then I'd contact aqua first.

You could safely run 4 fans off a psu molex but then you'd lose the speed control.

Thanks for the input - I have buzzed them an email. Looking at the manual for the controller it states that:

"connection possible for four fans with RPM signal. ATTENTION max load 5 watts per channel"

I also have a few Noctua NFP12's which are rated at 1.08W iput power/0.09A input current - these sound OK in four?

Would other 120mm fans be so different? I was planning to use the 4 channels like this:

1) 2 x 120mm Enermax UCCL12 Cluster TWISTER Fan (M.B.T) Veri Speed, White LED (on/off)
2) 4 x 120mm Xigmatek XLF-F1253 White 4 LED Quiet Fan 1500rpm Transparent
3) 2 x 80mm Enermax UCCL8 Cluster TWISTER Fan (M.B.T.) Veri Speed, White LED (on/off) inc Rubber A/V Plugs
4) 1 x 120mm Xigmatek XLF-F1253 White 4 LED Quiet Fan 1500rpm Transparent

[redders6600]

Keep us updated with the fan situation

[Syd]

Ok tonight have now taken case apart



Not half as hard as i'd expected actually!!!

Now looking to fit the bottom rad - do I take it the order is rad>bracket>gasket>fans, and i'm going to run them pulling - is that ok?

The next big question is which fans to use, i need to try and get the stats on the ones i have:

4 x Noctua NP12
4 x Enermax UCCL12 Cluster TWISTER
4x Xigmatek XLF-F1253 White 4 LED

I wanted to use the xigs as they have the Leds which are going to work with the case side i've chosen - but i'm paranoid about killing the aquaero with too much draw - so must try and get the amps for these fans (i think they are all 12v) so I can do the 'math' - thanks to Lauben on watercoolingUK for the Physics brush up!!!

Bed time beckons

BTW gotta love you guys that fab your own parts - hats off to you!!

[Pob255]

Ooo let me know what you think of thoes enermax twisters please, I'm currently considering getting one of the 92mm versions for a cpu cooler.

oh and the Xigmatek XLF-F1253 is rated at 0.2a so it should be safe to chain 4 together
The power rateing can normally be found of the back of the fan
http://www.hwreviewlabs.com/reviewph...3/HPIM1649.JPG

[Syd]

Ooh that's reassuring I have just done some maths so I make it

0.2 x 12 x 4 = 9.6w Xigmatek XLF-F1253 White 4 LED
0.4 x 12 x 2 = 9.6w Enermax UCCL12 Cluster TWISTER Fan (M.B.T)
0.2 x 12 x 1 = 2.4w Xigmatek XLF-F1253 White 4 LED
0.25 x 12 x 2 = 6w Enermax UCCL92 Cluster TWISTER Fan (M.B.T.)

Total 27.6x - is that correct?

[Bugbait]

Regarding your loop, are you going for the i7?

If you are, then

Res > Pump > 120.4 > CPU Block > GTS240 > 285Block > Temp Sensor> Res

If you are using a socket 775, then

Res > Pump > 120.4 > 285Block > GTS240 > CPU Block > Temp Sensor> Res

The idea being, you put the largest rad before the hottest item in loop, so you get the biggest heat transfer.

Shouldn't it be:

Res > Pump > CPU Block > Primary GFX > Secondary GFX > Rad > Temp > Res ?

Rad's are generally not very restrictive (ie. Fine for end of loop) and should be removing heat before hitting the reservoir. I always link the CPU block as the first component from the pump to maximize pressure and flow (usually being the most restrictive). GFX cards are generally engineered to run at higher temperatures (around 100C) compared to CPU's (about 70-80C) so the coolest water should be flowing over the CPU.

[piggin23]

Bugbait
If you don't pass it through the rad first then the heat generated by the pump will be passed onto the CPU block.
Also Syd has two rads in his system, your setup only shows one.

[Bugbait]

Bugbait
If you don't pass it through the rad first then the heat generated by the pump will be passed onto the CPU block.
Also Syd has two rads in his system, your setup only shows one.

Is the heat from the pump significant enough to warrant this? Isn't passing hot water over a hot pump worse for it's longevity than introducing a bit of extra heat into the CPU (but cooling the pump)?

I usually chain my rad's at the end of my loops, so whether it's one or five they are all inclusive under "Rad" in my flow. Smaller rad's can be fitted throughout the loop if needed depending on where things are actually mounted. If the rad's are mounted high then it's a long trip back to the CPU.

EDIT: Looking at Syd's case, you're probably right with the rad being mounted low and likely near the pump. It would be less distance to run Pump > Rad > CPU block.

[scaryjim]

I usually chain my rad's at the end of my loops, so whether it's one or five they are all inclusive under "Rad" in my flow.

Surely that's massively reducing the cooling potential for any component looped after the CPU, then? Given that heat dissippation (sp?) rate is proportionate to dT you'll get far more efficient cooling of your GFX / later components if you pass the hot water through a rad immediately after it leaves the CPU block...

[Syd]

Right day off today Struggled to sleep much last night thinking of what oder to do stuff in - i'm a little overwhelmed TBH

However now the case is stripped down I figure I'll fit the 2 rads and cut the mobo tray (first time dremelist today!!!)

First job - while waiting for some more buts including the dremel, fit the mobo stays......................Mmmmm that was fun. I guess the powdercoating makes the holes a little narrow so ended up putting a screw in a stay and then screwing that in to 'pilot' throught the powder coat - killed about 5 screws doing that - anyway that is now done. Now I have marked up and masked the Mobo tray for CPU slot - if I need to do any more cable management that will become apparent later im sure!



So now just to wait for the courier wih my Dremel and last bits............

Got these this morning special delivery well done Tom :








Not quite sure what it is - but i find myself fascinated by metal fittings

[Syd]

The dremel has made its first cut....................





Yipee I think as i feel generally pleased with m first ever mod - oh the hole is too small for a 775 backplate - rats............back to the drawing board/garden table