What watercooling kit ?

1337z0r

1337z0r

1337z0r

Associate
Joined
25 Nov 2006
Posts
1,906
Location
Birmingham
Hi,
Im thinking of going down the watercooling road, i havent overclocked before but im going to have a proper try at overclocking my e6300. Will be with the rig in my sig. Also i have a HIS X1950XT havent used it yet, cant till sunday as its my bday so i want a gpu block for that :D. One thing though, is i like to leave my pc on 24/7 so im just wondering if i wont be able to with watercooling. Also it has to be silent. I have a TT Armour case which should be big enough for a monster watercooling kit. Im willing to spend £150 max maybe a bit more if i have too.
 
i think you would be hard pushed to get a good kit for a £150 .you might be better off geting custom parts and building the kit this way .
 
Hi all,
Im also considering water cooling for my rig. Most of the time I leave my PC on 24/7. Is this safe with water cooling????
 
indeed i think it is, the only thing that can concievebly go wrong with a properly built WC system is the pump failing...if that happens there'll still be water in the system, so it won't overheat anytime soon, i have decided alphacool is the way forward afterall :D
 
by the time you add a GPU block to that kit u will be on about £200 .i think u can go custom. cheaper and better parts
 
Last edited:
I have got a decent kit for sale, for instance, its for sale at £140, that isnt overpriced and its affordable so you can get kits for the price range you are looking for. Its a Koolance Exos 2 kit, which is universal.

However to buy new kits at that price it may be hard pushed!
 
lots of people on here say koolance parts are bad, but i have been reading some reviews, and there block perform very well... :confused:
 
I have used it on a P4 3.2 ghz > Hit 4.3ghz with it. temps Idle 27 load 34

I have used it with AMD opteron 165+ 1.8ghz and hit 3ghz stable. idle 21 load 28

I have used it with AMD 4000+ 2.4ghz and hit 3.17ghz stable. idle 22 load 31

Koolance really arent that bad at all. the reviews are good and the results are just as good.

It hasnt ever leaked, made noise, had problem or caused hassle.

For the money i paid (£230) , this kit for a "noob" at watercooling did me fine! I was scared of water cooling at first, but this proved a great way to start as it was user friendly and you can add things to the loop (HDD, chipset, gpu etc)

I dont see what the problem is with Koolance? other than the price and its not custom (unless you alter bits)
 
so why all the forum dislike of koolance, there blocks look well made, well designed and are well priced and have good overall performance, plus they make a block for just about everything :confused:
 
I seriously dont know..I welcome a challenge from someone who dislikes Koolance for whateva reason?

Only reason im selling is cause I had to fund my move to conroe...shame that it's redundant and I cant fit to my new case :(
 
Koolance aren't terrible but their systems were originally built into cases and were therefore a hefty layout.

But Booner - I recognise you have a kit for sale and want to pimp the system but those temperatures you mention should really be placed in some sort of context. The numbers suggest that your Koolance achieved a +7C for an overclocked 165 which would easily be 100W at that speed. That places the block at 0.07C/W which simply isn't possible. No block in the world has ever achieved that sort of number.
 
MikeTimbers said:
The numbers suggest that your Koolance achieved a +7C for an overclocked 165 which would easily be 100W at that speed. That places the block at 0.07C/W which simply isn't possible. No block in the world has ever achieved that sort of number.
Click to expand...
Not quite. If you're looking at direct deltas without reference to ambient. Then the temperature delta corresponds to the change in power used. Not total power.

TR = (Tc - Ta)/P

where:
TR = Total Thermal Resistance
Tc = Tcase, cpu case temperature
Ta = chassis ambient temperature
P = total cpu power dissipation

Then rearrange and substitute for Ta

TR = Tc2 - Tc1 / (P2 - P1)

So a 7C delta and an idle power consumption of say 45W

7 / (100-45) = 0.127 C/W
 
I have a Koolance PC3-725 (the one in the Lian-Li V1000B case) and it is excellent. No issues with any aspect of it's cooling. It is not silent though. EXtremely quiet, yes, but silent - no. If you want silence then you need a passive system like the Zalman Reserators but then I think you would complain about the efficiency of the cooling being little better than a good air cooler.

One excellent kit (in my opinion) is the Alphacool Answer III 120 which is a complete doddle to fit and even comes pre-filled. The one I had was the basic one, at about £100, but there is a significantly better version at £150 which features a Laing pump and is supposedly very, very good indeed. This was completely inaudible with the case cover on and even with the cover off the noise was more of a purr than a buzzing of the fan.

The one gotcha with this system is that you need to have 20mm clear beyond the 120mm fan housing on the side of the case opposite the motherboard. That is actually quite a big gap and many cases don't have that much room. The Lian Li PC7+ definitely doesn't. The Akasa Eclipse does, but you have to fit it after you've put the motherboard tray back. There is a good diagram in the fitting guide which you can download from Alphacool's website. As Alphacool sell direct I can't link you, but I'm sure you can find it yourself.
 
fornowagain said:
Not quite. If you're looking at direct deltas without reference to ambient. Then the temperature delta corresponds to the change in power used. Not total power.

TR = (Tc - Ta)/P

where:
TR = Total Thermal Resistance
Tc = Tcase, cpu case temperature
Ta = chassis ambient temperature
P = total cpu power dissipation

Then rearrange and substitute for Ta

TR = Tc2 - Tc1 / (P2 - P1)

So a 7C delta and an idle power consumption of say 45W

7 / (100-45) = 0.127 C/W
Click to expand...

Why would there be an idle consumption of 45W? Opterons basically shutdown when idle. At 45W and 0.127C/W, that would place his ambient at somewhere around 15C! Is his computer outdoors?

Fornowagain, you're obviously a qualified guy. You clearly understand all sorts of physics that I can only marvel at, but computer temperatures are the greatest lie on the Internet. Time after time, load/idle temperatures are quoted without reference to any reality whatsoever. Computer manufacturers deliberately miscalibrate their BIOSes to let geeks like us think our after-market cooling solutions are ever so wonderful. Frankly, it wouldn't surprise me if manufacturers deliberately build crap stock coolers so we can all feel good about buying a £30 chunk of metal to "make our system better". Anyone who quotes BIOS reported temperatures is kidding themselves.
 
MikeTimbers said:
Why would there be an idle consumption of 45W? Opterons basically shutdown when idle. At 45W and 0.127C/W, that would place his ambient at somewhere around 15C! Is his computer outdoors?
Click to expand...
I've no idea on the ambient, its not given. I agree 15C is low, but possible. I can't comment, I don't have the data and neither can I say if the values are accurate. The derived formula refers to the change in power, even if there is a T offset (starting point) it still works for the power delta, as TR is constant (within a range). As for the idle value it was an example (given below), I don't have the TDP curves for the processor. I'm just trying to show how to correctly calculate the values.

With an overvolted/overclock cpu, even at idle its overclocked state pushes up consumption, that's not considering further losses from disabling power saving functions like wait states, CnQ etc. As many do to enhance overclocks.

Overclocked Watts = Default Watts * (Overclocked Mhz \ Default Mhz) * (Overclocked Vcore \ Default Vcore)²

Online version

E.g. if the original frequency was 1800MHz at 3000MHz. And again stock idle wattage was "say" 20W, overvolted from 1.3v to 1.5v, then.

20 * (3000/1800) * (1.5/1.3)² = 44.4W

Now, given the correct TDP values for the processor and reliable voltage's (DMM), if there's a known thermal resistance of a cooling system. Excluding losses/efficiency, estimates of temperature deltas can be calculated to give a rough (ball park) idea of the temperatures to expect. There's all kinds of reasons why the real values are always worse, never better. I'm suggesting that its useful to be able to calculate that quoted values are often possible/impossible.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom