US vs German systems - advantages/disadvantages?

[firtol88]

I don't think this has been mentioned but this could easily be a three way competition

Danger Den type 1/2
Swiftech type 3/8
Euro type 1/4

[Chew_Toy]

I agree. I would like to see how a traditional top end german style (10mm od tube low to med flow pump) would compare to the others.
If the performance is relativly close (2~4°c) to the others I would like to setup a system or two that way. But without an accurate test it is too much money to spend with just guess work. The only 2 changes I would make is probably use the C-Systems pump and a BIX for the rad, just for ease of construction in the systems I was thinking of.

[nigelyuen]

we should test each part at a time like
themal chill 120.2 vs airplex 240
alphacool xp vs cascade

something like that

[nigelyuen]

Originally Posted by Oberst Oswald

nigelyuen...Please suppy link to the Wasserkuhlung. Now that says German!

is that what you asking for?
www.1acooling.de

[BalefireX]

Doesn't seem very fair to restrict one type of system to a single manufacturer set kit while allowing the other type to pick and choose between different manufacturers. Also, BIX2 radiators aren't one type or the other really - they're made by a company in the Phillipines. If we're going to throw the idea of "premade kit" out of the window, why not MCW6002, Overvolted D4, Dtek Pro-120 Rad & Shroud, 1/2" tubing, or something of the like?

I'm interested to see a competition between a US manufacturer's set kit and a German manufacturer's set kit (Swiftech v. Alphacool or the like) or a competition between the "best of German products" and the "best of US products" but lets not mix those two concepts.

[Les]

maybe those interested could help pHaestus with the funding of these tests
think PayPal preferred
phaestus@procooling.com
hoping you (pH) received my July £10

[firtol88]

Originally Posted by Les

maybe those interested could help pHaestus with the funding of these tests
think PayPal preferred
phaestus@procooling.com
hoping you (pH) received my July £10

I'll pitch in if we can try for all the big names; DD, Swifty, Innovatek, Asetek...

[Les]

Originally Posted by firtol88

I'll pitch in if we can try for all the big names; DD, Swifty, Innovatek, Asetek...

Ah.
I have minimal interest in this "shooting match"
However am interested in the development of test-beds
It is this to which I donate, but each to his own.

[pHaestus]

Les:
Your donations have been most appreciated. I started setting up some of the equipment today in fact (Fluke RTDs) and as soon as one more package arrives I can begin to experience the joys of GPIB configuration and labview.

[nigelyuen]

Originally Posted by pHaestus

We don't have any problems here with Dangerden; I talk to Dennis and Jeremy fairly regularly and have always said they make well-built and inexpensive products for the mainstream. I had some altercations with DangerDen FANBOYS on other forums but that's nothing new.

Nigel: If thick base blocks are rubbish then I guess there's no need to test german waterblocks at all?

Pug: It seems to me that all we are doing here is arguing the same debate that led me to testing waterblocks with a large pump and flow control instead of with a variety of pumps and radiators. And when I said I'd make a dual radiator system with loud fans to win a competition for the good ole US of A you thought that was pretty unfair. Now that's pretty much what you are doing: "I could use a radiator more typical of the German style but then I'd switch to loud fans". I thought the concept of quiet cooling was stressed in German systems?

How hard is it to choose a complete kit put together by a German mfgr? Or barring that just send a few waterblocks on loan for me to test?

Or (here's probably the best solution) why not send me a German kit as provided by a German mfgr and then send the parts YOU would use instead for your hybrid wizard designs kit? That way we all see the tradeoffs (if any) involved in the German style, the American style, and then how much is to be gained from mixing and matching?

I am fairly confident I could get the missing parts to build either a DTek or a DD kit as well.

if a waterblock with a thick base, doesnt it mean, its bad performance?
doesnt the water transfer the heat faster with the pump on, than copper?

[nigelyuen]

thats how thin the base is for my block

[firtol88]

Originally Posted by pHaestus

Les:
Your donations have been most appreciated. I started setting up some of the equipment today in fact (Fluke RTDs) and as soon as one more package arrives I can begin to experience the joys of GPIB configuration and labview.

There is something wrong with anyone who can even jokingly use the word joy in the same sentence with any prepackaged automation HMI.

[Cathar]

Originally Posted by nigelyuen

if a waterblock with a thick base, doesnt it mean, its bad performance?

I'm sure that Les and BillA are shaking their heads about now.

Aside from that, water-cooling basically comes down to how well the block and the radiator are working, and to some degree, how much heat the pump puts into the loop.

Given water flow-rates above about 2LPM (0.5GPM), the thermal capacity of water ceases to be an issue of any real concern in the performance of a waterblock.

What I would have liked to have seen was whether or not a top-end "German" waterblock at 0.25GPM can indeed outperform, or at the very least equally match, a top-end "US" waterblock that's running at 1.5GPM.

If it can, then that's pretty much a conclusive decider in saying that US setups are inefficient and all the other blather that goes along with that side of the debate.

If that's provable, then that pretty much means that we can all switch to 6mm ID tubing, throw away our powerful pump, and stick an all-brass C-systems pump with push-fittings (or an Eheim 1046 for the Germans), and forget about the current line of development being followed in USA/Australia.

That, to me, would be more of a revolutionary ground-breaking shake-up and wake-up call to all the USA and Aussie block makers.

I want someone to prove to me that the USA/Aussie developmental approach has gotten it all wrong.

The closest thing I've yet seen is the Hydrocool Hydro-Stream HS-5 block (designed by a German engineer living here in Australia - but truly a "German" style block). For JoeC/Overclockers, at 0.3GPM the HS5 has a C/W of 0.139. For the Asetek Antarctica (White Water clone), it has a C/W of 0.118 at 1.5GPM. That's still a pretty handy lead for the "USA". Heck, extrapolating a bit of math with respect to pressure drop and holding pressure drop constant across both blocks, and the copper Asetek Antarctica holds a 0.015C/W lead over the Silver HS5.

I want a German block to show me the error of my ways. I want to not use a strong pump to achieve the best possible performance.

Originally Posted by nigelyuen

alphacool xp vs cascade

The Cascade is a 17-month-old about-to-be-totally-obsoleted design. How long has the Alphacool XP been out for?

[Edit: to fix incorrect units]

[nigelyuen]

Originally Posted by Cathar

I'm sure that Les and BillA are shaking their heads about now.

Aside from that, water-cooling basically comes down to how well the block and the radiator are working, and to some degree, how much heat the pump puts into the loop.

Given water flow-rates above about 2LPM (0.5LPM), the thermal capacity of water ceases to be an issue of any real concern in the performance of a waterblock.

What I would have liked to have seen was whether or not a top-end "German" waterblock at 0.25GPM can indeed outperform, or at the very least equally match, a top-end "US" waterblock that's running at 1.5GPM.

If it can, then that's pretty much a conclusive decider in saying that US setups are inefficient and all the other blather that goes along with that side of the debate.

If that's provable, then that pretty much means that we can all switch to 6mm ID tubing, throw away our powerful pump, and stick an all-brass C-systems pump with push-fittings (or an Eheim 1046 for the Germans), and forget about the current line of development being followed in USA/Australia.

That, to me, would be more of a revolutionary ground-breaking shake-up and wake-up call to all the USA and Aussie block makers.

I want someone to prove to me that the USA/Aussie developmental approach has gotten it all wrong.

The closest thing I've yet seen is the Hydrocool Hydro-Stream HS-5 block (designed by a German engineer living here in Australia - but truly a "German" style block). For JoeC/Overclockers, at 0.3GPM the HS5 has a C/W of 0.139. For the Asetek Antarctica (White Water clone), it has a C/W of 0.118 at 1.5GPM. That's still a pretty handy lead for the "USA". Heck, extrapolating a bit of math with respect to pressure drop and holding pressure drop constant across both blocks, and the copper Asetek Antarctica holds a 0.015C/W lead over the Silver HS5.

I want a German block to show me the error of my ways. I want to not use a strong pump to achieve the best possible performance.



The Cascade is a 17-month-old about-to-be-totally-obsoleted design. How long has the Alphacool XP been out for?

i know watercooling is not just a waterblock
but i just want to know having a thick base, does it affect the performance
with the same design, waterblock with thick base and waterblock with thin base, which is better, thanks

[Cathar]

Originally Posted by nigelyuen

i know watercooling is not just a waterblock
but i just want to know having a thick base, does it affect the performance
with the same design, waterblock with thick base and waterblock with thin base, which is better, thanks

The answer is "it depends", and it depends on the design, and it depends on the targetted pumping pressure that the design is being optimised for.

BTW nigelyuen, do you really have to quote my entire post to make a 3-line comment not particularly related to the bulk of my post?

[Turbokeu]

Originally Posted by Pug

I was fed up of going around minding my own business (literally), trying to help people make an educated and informed choice with their watercooling setup, only to be trashed by words like "don't buy that German stuff, it's crap".
^^^ Now that was an actual quote from a ProForum member on another forum, in the actual thread that prompted nightic to compose this thread in the first place (before he emailed me to inform me that he'd asked for you guys' opinion too.)

Yes, I remember, that was me on Bit-tech.net.
Yes, I said "don't buy that German stuff" (it's just my personal opinion).
But I didn't say "it's crap"...

As lots of other people, I think of german watercooling as a silent, good looking but not so well performing solution...

And I'd liked to be proven wrong.

CD

[Incoherent]

Originally Posted by Cathar

I want to not use a strong pump to achieve the best possible performance.

That would be an interesting design brief. To build a block that is inferior with a stronger pump.
My numbers say it is physically impossible. Maybe with forced cavitation or the like.

The only possible scenario is equal performance at all reasonable flowrates, even then, equal is a matter of measurement resolution.

[Les]

Originally Posted by Incoherent

. Maybe with forced cavitation or the like.

flow-boiling?
have toyed with for several years
however with water and temps 20-40c do not think can occur

[Cathar]

Originally Posted by Incoherent

That would be an interesting design brief. To build a block that is inferior with a stronger pump.
My numbers say it is physically impossible. Maybe with forced cavitation or the like.

The only possible scenario is equal performance at all reasonable flowrates, even then, equal is a matter of measurement resolution.

Okay, somewhat poorly worded I admit.

For a practical example of what I refer to look at the Swifetch MCW5000-A performance curve.

Near dead-flat above 1GPM. Sure it gets very slightly better, but not enough to ever warrant anyone's bother to chase for it, and certainly not with the extra pump heat that would be required to achieve it.

So basically a performance curve like that, where the knee is at 0.25-0.5gpm, is achievable with an Eheim 1046 and 6mm ID tubing, and leads against all "US" blocks at least up until 2.0gpm is being fed to the "US" blocks.

Now that would make me sit up and take notice (and throw the towel in).

[Les]

Originally Posted by Cathar

So basically a performance curve like that, where the knee is at 0.25-0.5gpm, is achievable with an Eheim 1046 and 6mm ID tubing, and leads against all "US" blocks at least up until 2.0gpm.

Think you will find the position of the knee changes with the scale of the x-axis.

[Cathar]

Originally Posted by Les

Think you will find the position of the knee changes with the scale of the x-axis.

(sigh)

Okay, let's be a little bit more nit-picky shall we?

I would like to see a block ("wunder-kuhler"), that when coupled with an Eheim 1046, outperforms any other "US" block given whatever pump you care to throw at them, and for our "wunder-kuhler" itself does not achieve more than a 0.01C/W gain for a 100mm^2 sized die when couple with, say, a Danner Mag 3, effectively making it pointless for any heat load less than around 200W to ever consider sticking a stronger pump than an Eheim 1046 on it.

I sure you knew the point I was getting at, but are probably just determined to make me jump through logical hoops for amusement factor I guess.

[Les]

Nit picky .yes
However was just a general point(re. knee)
Not related to this ridiculous pissing contest

[Cathar]

Originally Posted by Les

Nit picky .yes
However was just a general point about performance.

Guess that'll teach me for posting after enjoying a few drinks.

[Les]

Originally Posted by Cathar

Guess that'll teach me for posting after enjoying a few drinks.

when will i learn

[Cathar]

Originally Posted by Les

when will i learn

what have I done now?