Some suggestions for Newbie water cooling build

[edhahaha]

Hi guys/girls, I am fairly new into the forum, and hopefully getting some really basic answers here, lets start with my specs, I currently have
2700k with silver arrow
p8p67 rev 3.0
gtx 580 amp! with artic cooling accelero extreme II
8 gb of corsair vengenace ram
ax 750
and 600T with side window panel

The comp runs fairly quiet already, but I am really looking into some upgrade for some water cooling action to prepare some quietness for my school in sept. So my potential upg list is as follow

600T ----> NZXT swith 810 matte black case
p8p67 -----> Asus maximus v formula (not yet release, with built in water cooling compatibility, main reason for going wc for the full loop)
gtx 580 -----> debating whether I should go gtx 680 or 670... 680 are running low in the water blocks from ek, and 670 is still in precess of making the blocks

water cooling parts:
res: Swiftech MCRES-MICRO REV2 Clear HI-FLOW Water Cooling Reservoir
pump: Swiftech MCP655
Tubing and fitting will be depending on the built in size of the mother when it comes out, and compression fittings will be chosen.
cpu block : EK-Supreme LTX or xspc Raystorm, or Bitspower Summit Ef Cpu Liquid cooling block.

and now my questions for the Rad is that I will be doing triple 120mm up top in the switch 810, if I choose a think triple 120 rad like RX 360 it would not fit a push pull, so would it be better performance if I go with RX 360 ( think rad) with only push OR　pull, or should I go with a normal triple 120 rad like swiftech MCR 320 with push and pull? fan wise I am thinking of cooler master R4 red led.

And would it be a lot more beneficial if I add another 240 rad ? but with four extra fans I would assume that defeats my purpose of silence though.

**And last questions would be if i go with 360 rad, and i put six 120mm fan for push/pull plus one additional case fan 120 mm, thats seven 120 mm fans in total, so how is that going to propose a silent pc operation? or do I just add resistors on all the fans to make them run at 5v while maintaining a fair temp?**


So basically this is my project for the summer, lots of ordering and most research are done, just need a bit of suggestions and advise before I actually get started!, thanks a lot for any help

 

[Conumdrum]

Okay. How much do you know about your heatload? If the RX 360 is enough that's great. Goyya size your raddage/fannage to the heatload first. The RX rads do very well with quality fans like the GT AP-15s, no need to go P/P on fans. Adding a RX 240 with AP-15's and cranking them down to 800 RPM or so is quieter than 3 AP15 at 1850 or if you undersize your rad, even louder fans.

Kinda a concensus here. More raddage, quieter. But if you have no clue, then it don't matter. You gotta understand the basics first. Have you read our stickies? ALL of them? I always ask.

We are an assebly line here. We have people break, rush, ruin parts. We have some with minimal input from us build and do well.

Which do you want to be?

This hellhole? A really fun read, really. I feel sorry for the guy.
http://www.overclockers.com/forums/showthread.php?t=709200

Or these two?
http://www.overclockers.com/forums/showthread.php?p=6527116#post6527116
http://www.overclockers.com/forums/showthread.php?p=6796609#post6796609


You got a good basic list. Many great res options these days.
The 655 is almost outdated with the 35X out, a great PWM pump.
And if you must use compressions for pretty looking stuff, so be it.

No mention of fans (important, means you understand diff rads) or a liquid or a biocide.

EDIT. Resistors? Maybe you haven't heard of a 'fan controller' yet.........

 

[bluezero5]

good to see another interested in Water Cooling.

firstly, if quiet is your target, more rad is actually a good thing.
you can use the nice AP-15 fan, which is quiet at 1850 RPMs.
and to avoid having to go Push/Pull, what you want, is Rad with 'lower' FPI (fin per inch), if a RAD has too high FPI, then you will likely need push/pull to ensure a decent air flow.

Second, the pump.
MCP655 is a good pump, but it is really the 'last' gen's stuff.
people these days use the MCP350, MCP355 or MCP35X2, all of these have a 'lower' clearance, smaller, and better PQ curve. (which means, it pumps better even under higher restriction.) If you are looking for more Rad, then space might be a concern, so I will recommend you go for these pumps instead.

Tubing. General rule is 'larger ID, better flow', 'Thicker tubes, harder to kink', but of course, 'bigger it is, clumpsier it might be'. So it really depends how you cooling, but in general, 1/2" ID 3/4" OD will fit most Full Tower cases, that's what I use, that's what I will recommend. Don't use PrimoChill tubes, the more I read, the worse their quality has become since 2012.

oh, and GTX670 and GTX680 really small difference, GTX670 will give you better performance/$$, and GTX680 is really only if you are pushing some limits, I doubt you will see any difference on performance, go for GTX670. and yes, check the availibility of Waterblocks before purchasing, aside from EK, Koolance also makes VERY GOOD GPU blocks, that's what I use, that's what I will recommend.

feel free to ask questions.
in general, you want to know EVERYTHING before investing into watercooling, can cut your investment cost by A LOT if you do it right.

I have some pics of my setup under my Profiles' album, feel free to take a look, or PM me if you got questions about it.

 

[m0r7if3r]

FANS FANS FANS!
http://www.xtremesystems.org/forums/showthread.php?t=223391
http://www.xtremesystems.org/forums/showthread.php?t=261778

 

[bluezero5]

I have never been too crazy about XtremeSys's Fan research.
too many scientific errors.. which is a pet peeve of mine, given I work at a scientific lab.

to point out a few:

XtremeSys:

I think from a noise level only perspective...most fans (90%) perform pretty much the same per CFM level. Supposedly it takes 3dbA for most humans to barely perceive a noise level change, and that's about the spread I have found for most fans with a few exceptions. I'll list out some of my favorites below:

FALSE.

The Decibel scale is exponential, the difference between 20-23 and 60-63 is COMPLETELY different.

The power ratio at 20 is just x 100, but at 60 it is x 1,000,000
calling those 3dB the same.. is just..

View attachment what.bmp


then comes how they used the Anemometer...
.... which also.. makes me bite my lip and curse the gods.
but alas, at least that will yield 'fairly accurate' 'relative' results, so that's ok. but just picking up a anemometer and using it without math.. is just not right..

if they call that a 'just for fun' research, then OK... can read for the laughs.
but when they do things like CFM/dB... again, that's not even remotely an accurate thing to do. cause a fan with 200 CFM but 66 dB, is NOT the same noise penaulty with a 100CFM of 33dB, on that scale, you will ALWAYS wan to use the 100CFM.

and they KEEP charting CFM to dB without using Log scale for the noise ratio... and drawing conclusions from there..
sorry.. but that research was just hours of testing wasted into bad science. Just to mock them, I have asked a 'blonde' to check noise and CFM, can sure she yield better results than that test. (but apparently, not many people know I was doing that to mock that XtremeSys's research.. and some even took it seriously.. disregarding my disclaimer there.) ah well.

as for FANS. (enough of my ranting. sorry)

Quiet ones with good Statis Pressure
Gentle Typhoon AP15. very quiet, good pressure. but a little weak in CFM department, good for Rads with low FPI. can work to a degree to rads with higher FPI, but effectiveness depreicates quickly with more FPI.

Performance fans:
Scythe SlipStream 1900 RPM, good CFM, rather low in the Static pressure, requires push/pull to get optimal results. if used in just push or pull configs, On low FPI rads, it will perform just about the 'same' as AP-15, but louder. against high FPI, not very effective. But in PUSH/PULL, then static pressure problem solved, and it is one of the best fans. Noise tolerable.

Reeven 2000 RPM cold wing. Everything the SlipStream is, with a bit more Static Pressure, and quieter. but availibility a BIG concern, I haven't seen this fan out of Asia.

that be my fan choice 101.




anyway. I wanted to say that for a while.

 

[m0r7if3r]

Yea, but there's some weird thing with how humans perceive noise or something that makes the 3db a lot closer than you're making them out to be, if memory serves. I really don't know enough about it to offer a counter argument, but it's got a good scale of relative performance and links that allow the noise characteristics of all the fans to be analyzed.

 

[bluezero5]

Yea, but there's some weird thing with how humans perceive noise or something that makes the 3db a lot closer than you're making them out to be, if memory serves. I really don't know enough about it to offer a counter argument, but it's got a good scale of relative performance and links that allow the noise characteristics of all the fans to be analyzed.

This is a commonly misunderstand fact, it is ok.

0 db Threshold of hearing
30 db Whisper
40 db Buzz of mosquito
50 db Normal conversation
70 db Vacuum cleaner
100 db Subway or power mower
120 db Rock concert
130 db Jackhammer or machine gun
150 db Nearby jet plane

I will say my whisper is probably 5 times quieter than my normal speech. but on decibels.. just merely 50%. that linear range is only perceived through 10-20dB, anymore or less, if you think about it... it is not linear 'at all'.

Jackhammer 100-140 dB only. 2x from normal talking level if you think about it. but in reality, you think it must be 10x as loud as talking, if not more.

How about a chain saw? Lets call it.. 5 times as loud as your noisest fan? I guess that's fair to say right? buit it is only only at 110 decibels.

regardless. I am sure if XtremeSys plot their data on a Log scale, it will make a lot of fans they deem 'poor', probably a fair bit better.

 

[BlackEdition]

let me first qualify my statements by saying that i've never personally water-cooled a system... i have, however, read a ton about it recently...

you first really have to define what you're going for: quiet operation, middle ground, bleeding edge performance...

it seems like you've done that by setting out to build a quiet pc...

to do that, it seems the conventional wisdom is to use more rads with lower FPI and quiet fans... as far as quiet fans, there are a number of ways you can go. however, the most convenient is to use a fan controller so that you can control the speed at which they spin (faster = louder, generally)...

to ensure you've got adequate cooling, you have to calculate your heatload (i.e. how much heat all the components that you're going to be cooling put out)... i don't know if there is a hard and fast way of doing this, to tell you the truth... i'd bet you'd want to look up reviews of the components done by a reputable enthusiast site to see if they quote temp values etc... the sandy bridge processors don't run particularly hot (as far as i know) and the 670/680s are actually pretty good in the temp department as well...but don't take my word for it, and check out some reviews...

also, i think you should check out the excellent build log in a Switch 810 that was done recently by stainlineho:

http://www.overclockers.com/forums/showthread.php?t=708526

It sounds like it is very similar to what you'd like to accomplish. Maybe ask him about the volume of the rig?

good luck, and when you build, you should log it for the benefit of other watercooling first timers (like me...eventually)...

 

[m0r7if3r]

This is a commonly misunderstand fact, it is ok.

0 db Threshold of hearing
30 db Whisper
40 db Buzz of mosquito
50 db Normal conversation
70 db Vacuum cleaner
100 db Subway or power mower
120 db Rock concert
130 db Jackhammer or machine gun
150 db Nearby jet plane

I will say my whisper is probably 5 times quieter than my normal speech. but on decibels.. just merely 50%. that linear range is only perceived through 10-20dB, anymore or less, if you think about it... it is not linear 'at all'.

Jackhammer 100-140 dB only. 2x from normal talking level if you think about it. but in reality, you think it must be 10x as loud as talking, if not more.

How about a chain saw? Lets call it.. 5 times as loud as your noisest fan? I guess that's fair to say right? buit it is only only at 110 decibels.

regardless. I am sure if XtremeSys plot their data on a Log scale, it will make a lot of fans they deem 'poor', probably a fair bit better.

Duely noted. I actually wasn't thinking when I replied, because I know from the work I've done on guitar amps (don't ask, because I swear I will never stop talking lol) that 10dB ~= 2x perceived loudness...

 

[TheHutchTTU]

I have never been too crazy about XtremeSys's Fan research.
too many scientific errors.. which is a pet peeve of mine, given I work at a scientific lab.

to point out a few:

XtremeSys:


FALSE.

The Decibel scale is exponential, the difference between 20-23 and 60-63 is COMPLETELY different.

The power ratio at 20 is just x 100, but at 60 it is x 1,000,000
calling those 3dB the same.. is just..

View attachment 111406


anyway. I wanted to say that for a while.

The decibel is logarithmic and describes a ratio.

The decibel is a tricky unit because it describes so many different things in different ways.
If we are talking about a power ratio (irrelevant to what we do) than a 3dB difference is twice as much power.
When talking about what humans hear (relevant) than a 10dB difference is twice as loud.

If you look at the specs of a fan they will specify the fan has a dBA rating (not dB).
dBA is an absolute measurement, taken with instruments, and is what our ears hear.
For example, someone measures a fan to be 30dBA and then measures a different fan at 40dBA, that is a 10dB difference thus twice as loud.

Let's look at your power ratio example. You said 20-23 is not the same as 60-63.
True, but you are defining a range, not an absolute measurement.
If fan1 requires 100x more power than fan2, then fan1 requires 20dB more power.
If fan1 requires 1,000,000x more power than fan2, than fan1 requires 60dB more power. Head hurt yet?

Please don't take numbers off Wikipedia and assume what they mean.

-H

 

[bluezero5]

lol. due to work related reasons, I have to work with the decibel scale very often, so it comes quite naturally for me.

10dB being 'twice as loud' on perceived scaled, is tricky.

cause when you jump from 10 - 40, that's is not 4 times as loud perceived.
but x2 x2 x2 that's x8 times as loud perceived, and 10,000 the power ratio.
still not linear, and still makes charting those linearly incorrect. and no, that's not what dBA really means. dBA is just the measurement within a 'frequency range' that humans can hear, so the noise can be 150 dB, but if it is at 300k Hz, you can't hear it, so it is not there. dBA is a contour filter, meaning you measure to the human hearing range only. but 20dB still = 20dBA in 1000 hz. Here's a good read of it for those interested, http://personal.cityu.edu.hk/~bsapplec/single.htm


I came to realize this topic is fun to talk, but doesn't help the OP.
so lets take it to a new thread, or PM if we want to continue?