[Mize]

Originally Posted by Rodéric

Interesting, I only consider watercooling for quiet and good cooling, having a ton of fans to compensate would be counter productive.
I assume you are overclocking.

For me it's mainly a geekfest project thing. with all the fans on low it's actually pretty quiet (more so than my HAF case with 1/3 the fans). Hell just getting rid of GPU fan noise is a big help as the spectrum of the Gentle Typhoons is much nicer on the ears.

I do overclock (2600k@4.5GHz, 2xGTX580s@850+) but not that much these days.

[Davros]

Originally Posted by Rodéric

having a ton of fans to compensate would be counter productive.

He, he, I've had many a discussion with Mize along the lines of "Mize you crazy fool, the whole point of water cooling is that you dont need 15 fans"

[lanek]

Originally Posted by Rodéric

I was considering water cooling too until I saw a comparison of CPU water cooling vs air cooling, although the CPU is indeed better cooled, the lack of a CPU fan means other MoBo components aren't cooled properly anymore, such as the chipset and the RAM, also the GPU if you don't connect it to your water cooling system.

Thermographic images of Water/Air cooling:
http://www.comptoir-hardware.com/art...0.html?start=7

There's some little thing i see in this article who is needed to watch a bit:

- The motherboard used is a Gigaybte x58A-UD3R, this mobo was good, but have a major problem of high temperature on Northbridge and CPU phase.

- When the CPU was overclocked, this was really not a good idea to dont have a fan for refresh the NB and remove the blue backplate on the cooler around the CPU.

- The watercooling system used is really not a beast and the fact it have only one fan in this position is really not good as the air pulled will not be enough.

Ofc have a Noctua cooler or a tower who have 1 or 2x 120mm fans close to motherboard hardware will ofc have an impact on the temperature of this zone. if you dont have a fan there, in a case with bad ventilation, the motherboard hardware temperature will be higher.

If needed, put just a low fan speed who push air on the motherboard and you will get the same result. A side case 240mm fan as in the HAF-X is just perfect for push air on the motherboard.

But anyway most watercooling setup are made in a way it is not even needed. On my system i have a 360 radiator on the top, with 3x gentle Typoon 1850rpm, The temperature of component is extremely low, there's a massive airflow in the case.

Now of you really want the max, you have just to watercool too the motherboard..



Hopefully i have measure really well the place over the ram .. ( EK 360 XTX. 64mm thick )

[Mize]

Originally Posted by Davros

He, he, I've had many a discussion with Mize along the lines of "Mize you crazy fool, the whole point of water cooling is that you dont need 15 fans"

You can cool for silence or for heat dissipation, but not both since those radiators need to be cooled. Like I said, fans on low = quiet, fans on high means overclocked 580s not breaking a sweat.

[Grall]

Daaaaammmn, that's a monstrous radiator. Holy crap. What does that thing weigh (empty, as it ought to be able to hold a considerable amount of water)?

[lanek]

1.5Kg, not so much finally. The photo make look it bigger and larger it is in reality.. ( due to the point of view and the luminosity was so bad .. ) http://www.ekwb.com/shop/ek-coolstream-rad-xtx-360.html

Some test:
http://martinsliquidlab.org/2012/04/...m-rad-xtx-360/

http://martinsliquidlab.org/2012/03/...reme-hf-plate/


Old install on the UD7 (big lol to gigabyte, the blue power button is installed the wrong way , anyway photo quality is not top ( taken with a SG2)

But im really glad of the performance of this rad... i will let it alone in the loop as it does the job really well . ( 2x gpu + 2600K@5ghz in the loop: 2x EK 7970 waterblock + 1 Ek SupremeHF full nickel plate 6, D5 strong+Ek Top rev.2 and res bay from EK too ( on final i think my setup is really 100% EK lol, was an accident ).

[Grall]

Originally Posted by Mize

You can cool for silence or for heat dissipation, but not both since those radiators need to be cooled.

Well, you actually CAN have caek and eat it too, just pile on more radiators (with slow fans). Sound volume does not scale linearly with the number of sources, and fans on low don't make much in the way of high-frequency noise, which is the most irritating kind.

Of course it will get more expensive, and it probably won't fit inside your chassis...

An interesting experiment would be to stack two radiators on top of each other, to save space. Assume airflow from left to right, with water flowing the opposite direction, first entering the "downwind" radiator, then moving over to the other, "upwind" unit. Not as efficient I'd think as having both rads breathe fresh air, but it could let you mount all the hardware more or less internally in a case.

[lanek]

bad idea the stacked radiators, martinlabs have conduct tests:

results: 2 radiators stacked cant even touch 1 radiator.. bad result ( the air of the first is heating the second )
http://martinsliquidlab.org/2012/06/...dwich-testing/

[Mize]

Yeah, you have to keep in mind it's all about heat exchange. A radiator with no air moving through it is worse than a passive heat sink with a proper convective orientation. This is why so many overclockers go with push/pull configuration on their radiators - it make a big difference.

[Grall]

Originally Posted by lanek

( the air of the first is heating the second )

You need to keep track of the way air and water moves through the system so this does not happen. Fresh air enters radiator A first, then B. Hot water enters radiator B first, then A...

You see stacked radiators (well, air equivalent) in basically all dual-GPU blower coolers, so it obviously works.

[Mize]

So B heats up the air enough that it doesn't do much over A. This is why air velocity is more important - faster moving air means cooler air in contact with the radiator fins.

AFAIK, stacked HSF designs are to increase air velocity with minimal noise increase. If you can create a stacked radiator that results in higher airflow through both radiators you have a winner, but air pressure is a bit different with high-fin-density radiators compared to the relatively low density of HSFs. Stacked radiators would be equivalent to just a much thicker radiator, thermodynamically, so you just need fans that can create enough pressure to get high airflow through the double radiator.

[lanek]

Martin have do all the tests, and they cant be contested .. 2 rads sandwiched,thisis a no go .. I think he have test absoluletely all the configuration possible for it, but the result of the tests are clear.. no
way, this dont work.... the air exchange of the first rad used in the second rad is like use higher temp for entry the rad, result the water temp is higher, and the system will equilibrate it, resulting in an higher temperature of the water on final..

For make it work you will need a negative delta T between the water and air on the rad number 1 ...

[Grall]

If you arrange the fans and hoses like I mentioned it's not possible for temps to go higher than when using just one radiator. At worst it'd be exactly the same, but as total radiator surface area is larger and there's additional air turbulence involved that reduces laminar flow (which is bad in a cooler), I'd think there'd be at least a slight difference. Maybe not a cost-effective difference, but it should at least be there. Your Martin buddy must have screwed something up, it's the only explanation.

[Mize]

Originally Posted by Grall

If you arrange the fans and hoses like I mentioned it's not possible for temps to go higher than when using just one radiator. At worst it'd be exactly the same, but as total radiator surface area is larger and there's additional air turbulence involved that reduces laminar flow (which is bad in a cooler), I'd think there'd be at least a slight difference. Maybe not a cost-effective difference, but it should at least be there. Your Martin buddy must have screwed something up, it's the only explanation.

Uh, sorry Grall, but you're incorrect.
You're assuming that the fans can move air through two radiators at a comparable velocity as through one. This simply isn't true. Radiators are fare more constrictive than HSFs and all that turbulence you mention - which is generally good - creates back-pressure that greatly reduces the flow created by fans.

I can't emphasize airflow enough here.

We can all agree that a radiator in a vacuum is near useless other than dissipating a small amount of heat through IR radiation (no air to which it can transfer the heat).

I think we can also all agree that air that is the same temperature as the water serves no use either.

Now let's talk about moving air at a lower temperature. The rate at which heat can be dumped from the radiator into the air is determined, in part, by the temperature difference between the water and the air. This difference is essentially a thermal gradient much like a hill - heat "rolls down the hill" to the colder side. The warmer my air is, the slower the heat transfer.

Anything that increases the dormancy of the air inside the radiator will decrease the rate at which heat is transfered because the air becomes hotter.

So if you want to double the thickness of your radiator or double the fin density or put two radiators together then you'd better switch your fans for those that can produce dramatically higher pressure such that the dormancy doesn't change.

[Grall]

All factors would have to be reasonably similar of course or any comparison becomes meaningless. That includes airflow, etc. I assumed you guys were with me on this.

Btw, if one was to stack rads, I'd want fans sandwiched in the middle as well as fans taking in air on top of the stack. High static pressure units, of course...

[Mize]

Well things just got a lot noisier then

[aaronspink]

Originally Posted by Grall

All factors would have to be reasonably similar of course or any comparison becomes meaningless. That includes airflow, etc. I assumed you guys were with me on this.

Btw, if one was to stack rads, I'd want fans sandwiched in the middle as well as fans taking in air on top of the stack. High static pressure units, of course...

Stacked radiators MIGHT work if you were using some of the "quiet" rad designs, but then you would probably be better off just buying one of the normal rad designs in that case. Basically, HWLabs already makes 3 different lines of radiators that pretty much satisfy all your possible needs.