Dual-loop (no, not like that - see inside)

[speculative]

Idea! What if you had a dual-loop h20 system where the 2 loops were completely independent of themselves. I'm talking about something like sandwiching 2 h20 blocks together and then sticking them sideways onto the CPU. So, you'd effectively have 2 h20 blocks instead of 1 on your CPU.

The benefits of this would be:

1. Lower temps (hopefully! :D )
2. 1 pump fails no prob, other pump handles the job enough so the CPU doesn't fry
3. 2 totally separate h20 cooling systems = twice the cooling power (I'm sure someone will correct me on that math though. ;) )
4. Benefit of 2 rads w/out the added resistance that using 2 rads w/only 1 pump would bring.
5. 1 rad could be the intake in the bottom front of case, or side of case, 1 rad could be out-take at top of case or back of case.
6. More thermal mass in the h20 block! :)

What do you think, am I on something? I would really like to see a retailer make one of these!

-spec

[lechumbl]

Hi Spec,

I don't see manufacturing these at a reasonable price..
Space would be also very large concern.
Most boxes don't have enough room for one water system.
Then you start placing the hardware outside, then you need larger pump to move water farther.
Sorry, just don't see it.

The idea is sound, but the logistics and size would be a major factor against you.

Now, let my opponents start............:2guns:

[Player0]

You'd need a special waterblock with dual channels to do this. Not sure if there would be too much added benefit, the core only has so much surface area.

[nitan]

Spec y not just have one block with 2 channels seperate channels cut into it (ie no need for sandwiching 2gether)....then have each channel fed by a seperate pumps and rad???.....as P0 sez the the core itself is only about 10mm sq (smaller i think)...

[Player0]

I think dual radiators would work better than this type of system, because the higher water temps make the rads work better.

But I dont know. Ive never seen something like this before.

[speculative]

Nitan: originally that is what I had thought - one regular-sized block w/2 channels. However, most h20 block channels are pretty long because you want the water to flow over as much surface area inside the block as you can, as this is what aids the h20 block's effectiveness at xferring the heat.

My other concern with that was that as Lee mentioned, there is only so much space to go around, and on my h20 block, for example, there simply is no room to fit another 2 fittings for another 2 tubes to it I'm afraid... Not unless the fittings were coming out the sides instead of the top. This might get in the way of things because the h20 block is so flat - but, if the h20 block were bigger, say the size of 2 stuck together, this would be much less of a problem. I'm sure it would be physically possible - but, I have clamps on my tubes and it's a cramped space as it is now.

Perhaps the fittings could extend quite a ways from the h20 block if necessary?

P0 - While my proposed h20 block idea is pretty far out of my reach :D I am seriously thinking about updating this h20 system, and perhaps adding another rad, which wouldn't be too hard. I really like the look of the bix rads, but I don't like that they seem to be 1/2" connections. I haven't researched them too much yet though, perhaps it's not too much of a job to connect them to a 3/8" setup.

Lee - I don't see why such an h20 block should cost much more than 2 h20 blocks. All the extra that's required is sandwiching them together.... Although, that would be a chunk of change.

Hmm... theoretically, it would be better to have 1 complete setup for the CPU and a second complete loop to chill the GPU. But, this is a pipe-dream. (Pun intended! :D )

-speculative

[Player0]

Spec, two laws of waterblock design make the dual channel block iffy...first, waterblocks work better the longer the water stays in the block, especially around the core. Because you have such limited room around this area, even if you use two channels, they will only provide 1/2 the time for water to absorb heat from the core.

Just had an idea...if you were to make one channel 1/2", in your standard spiral shape, and then inlay in to this chanel some sort of 1/4" copper tubing, you would actually have two channels in the same space which MIGHT actually have some benefit. Whether the two channels are working from the same 'loop' or not may not make too much difference. Huh. Too bad I cant build it :)

Anyway, before upgrading your water cooling system, try to spend some time figguring out where your limitation actually is. Maybe you should upgrade to 1/2?

[speculative]

Hmm... Not to go too off topic from original post, but is 1/2 really that much better than 3/8? I suppose there's more thermal mass, but that would be offset by the size of your reservoir (for total system mass). Would the channel inside the h20 block be larger, therefore creating more surface area for better cooling? I would think adding a 2nd rad would be more effective than going from 3/8 to 1/2 w/the same system otherwise... But, I don't have experience w/that yet. ;)

-speculative

[Player0]

1/2" gave me a big improvement, but that could be because switcing to 1/2" also meant I needed new pump and radiators. So, the whole upgrade gave me an improvement.

Hard to say how benefitial going to 1/2" was...but im sure its noticbly better.