US vs German systems - advantages/disadvantages?

[Blackeagle]

Titan151,

What pump, rad & blocks are in your set up?

[BillA]

Originally Posted by pHaestus

Let me guess....you guys are going to want lower than 0.5 GPM flow rates for these tests

the question is: are any (reasonably expected) European flow rates expected below 0.5gpm ?

one needs be mindful of the actual pump capability in these smaller ID systems, the friction losses are low specifically because the fluid velocity is also low

[Cathar]

Originally Posted by unregistered

the question is: are any (reasonably expected) European flow rates expected below 0.5gpm ?

one needs be mindful of the actual pump capability in these smaller ID systems, the friction losses are low specifically because the fluid velocity is also low

Agreed. The WCP testbed is a fairly bad indication as it uses a very high restriction radiator which challenges even the most restrictive waterblocks for flow-choking ability.

0.5GPM is perhaps a realistic minimum for any system that uses a heater-core style radiator. I could not imagine a system pushing less than 0.25GPM though, even with the continuous small-ID looped tubing radiators though.

For a lark I bought this baby 2.5W 200LPH pump that's about 40x40x30mm in size with 1/4" OD outlets on it (about 4.5mm ID) with a 65cm head height. Even that tiny thing could push around 1.5LPM (~0.4GPM) through two heater-cores in series, 2m of tubing and a Cascade SS.

[j813]

Yes an International evaluation.

Different types of categories, different types of setups, example: Single Rad category.
IMO using the same Fans is more comparable.
The Internal Diameter is already under observation anyway.
The results on evaluation could also be concluded on to Performance vs Price difference.


My 2cents

[nightic]

Originally Posted by Cathar

Hmmm, trying to find someone who sells it on-line via Credit Card. The CoolCases site wants you to do a direct deposit into their German bank account...

Watercooling.de sell it and accept international PayPal payments:


CF-1 Socket A/CF-1 Socket 478
(I think both mounting methods are included with the block anyway so the distinction isn't really necessary).

CF-1 Upgrade set

[Pug]

Originally Posted by unregistered

Pug
pHaestus has our std kit, H2O-8600, with 1, or 2, 80mm fan rads, 3/8"ID tubing, MCP600, etc.
- here we get into what you were describing, we sell that kit for 'silent' running (~18dbA@7V, 2 fans @~22dbA) though we do not say so
so right off I know that I will 'lose' the performance title, ah well

if you wish to hold to the spirit of this endeavor, I would suggest your 'silent' kit

Don't have much other than silent kits but can hold to that sort of noise floor.

Originally Posted by Blackeagle

Pug,

Any pics of the inside of the block you depicted as your first choice? It's a nice shiny one for sure.

Sorry mate, not just yet.

Originally Posted by Cathar

As a side note from this end of the world, Pug/Coolermeister are in the UK, which I personally perceive as being more on the US-side of the water-cooling market going by the content of the various forums over there.

Having the liberty to pick & choose easily from either side, you'll find users in both camps. CM tends to lean toward the US approach whilst I am more towards the German.
As you surmise, things have progressed toward reaching a more common ground... the tubing diameter has increased slightly for us and appears to be decreasing for the other camp. Hence the ability to consider this sort of head to head more easily, I'm sure.

Don't worry too much about sourcing the CF-2 just yet. As far as I am aware, the block I propose is more on a par with the Murks3 (not in mass-production - in the same way that your own blocks could be held in comparison to the Dtek WW).

Originally Posted by Meethoss

most guys I've spoken to seem to want a good balance between silence and power.

As mentioned, that's not just a US-centric concern.

Originally Posted by HAL-9000

I've also noticed, in all this international brouhaha we are having, that no one has mentioned the Japanese. Normally performance electronics and weird mod/tweaks on them are pretty common in Japan. Yet the country seems completely absent regarding representation in the water-cooling scene. Ever wonder why? I do, I think its strange.

I know of several users of German systems in Japan, fwiw. I'm sure there are proponents of both camps there though, much like the UK.

Originally Posted by Cathar

My best guess is 0.4-0.5GPM flow rates. Coincidentally they also put the RBX on par with the Cascade at those flow rates which certainly doesn't gel with any other reputable test (both public and private) - hence my strong desire to get the "best of the German blocks" over to Phaestus to test.

A figure of 52,1l/h @ 0,11bar was bandied about yesterday in regards to one of the high scorers in those tests fwiw.
Don't worry about footing the bill for the block when one is on offer - better to donate the cash to pH's testbench fund if you ask me. Rest assured, the block I propose will be in the WCP chart soon enough for your reference.

Originally Posted by pHaestus

Let me guess....you guys are going to want lower than 0.5 GPM flow rates for these tests

Originally Posted by unregistered

the question is: are any (reasonably expected) European flow rates expected below 0.5gpm ?

See above. I'll let you guys use your better judgement on that one (your testing experience far outweighs my own, obviously )

Originally Posted by pauldenton

i would suggest some reasonably desirable mid-tower (lian maybe?) on the basis that any kit that requires a pc70 or similar huge case is doomed to marginal appeal.....

Originally Posted by pHaestus

...Other than that I'd have to buy a case.

Well, with the level of expertise here, I think it would be fairly straightforward to assess potential ease of installation without actually purchasing said case. again, rather said necessary funds were funnelled towards testbench components for future benefit to all.


By the way - for those still obsessed with working to a price point ...

Originally Posted by Pug

a production kit that can compare to one of our sub-£200 ones

(emphasis mine) was the original statement. We've moved on from there in the interests of practical achievement to

Originally Posted by Pug

a similar kit to what we propose

I'm not going to get silly about it - many of you know I have a production radiator on my shelf that has mounts as standard for fifteen 120mm fans ...


side note - no sleep last night for me so sorry if I missed out any pertinent points...

[BillA]

a tester's perspective, with a cup of coffee:

it is clear form the above, and my own investigations of mini WCing systems, that very low flow rates do exist, <60 lph (0.3 gpm), and that wbs could well be designed to operate in this range
- is this not the crux of the German/US 'debate' ?

this being the case, 0.5 gpm is too high a minimum

I personally have always included 0.3 gpm (68.1 lph) as the initial datum
- this is at the extreme of a magnetic flow meter's capability with a 1/4" (10mm) flowtube due to the low fluid velocity

to those not familiar with low flow test results, rude surprises are in store

[pHaestus]

I guess people are getting a taste of it w/ the 0.5GPM numbers for Danger Den wbs Bill.

[BillA]

and JATO assist at 0.3gpm

[pHaestus]

I am not sure I follow that comment (JATO has to do with Automobile specs right?)

0.3GPM is below the 1% range of my flowmeter...

[Blackeagle]

I would think that real small pumps, 8mm lines, Papst low noise/pressure fans, and water blocks designed to work with them would make for very compact, and very silent systems. As someone else posted, easy to fit in mid tower cases.

The MCP6000 that pH just tested was the best to date (pH's testing) at low flows. Was part of it's desinging done with the EU market in mind?

[DickNervous]

Wow! That was a hell of a lot to read.

While I am not very active here, I thought I would put in my 2cents worth. I have several WC setups and just posted a reivew at Gruntville of the Alphacool Cape Kit like the one at the top of that page.

I have also built/reviewied a "typical American" setup with the PolarFLO SF block, Eheim 1250, BIX2, 1/2" tubing, etc. In addition, I have had on this setup the MCW-5002 and Whitewater.

Now, I admit my testing is not nearly as scientific as what the gang here does. However, I do think that it is at least accurate enough to give you an good idea of how this stuff performs in the real world. And since I have experience with both types of setups, I can sum up the differences pretty good, depending upon your situation. So here goes....

If you are an Exteme Overclocker....

.... then brute force is what you are looking for. Every degree counts. So go with a big 1/2"ID system and the PolarFLO/RBX/WW/MCW6000 block of your choice. Put the biggest pump you can in there and a radiator from a Kenworth and go to town D:

For the casual overclocker and hardcore gamer (we all know most games don't like the extreme overclocks)....

... then you have a choice to make. The top German kits perform much better then "adequate" and will keep your system nice and cool and look good doing it. However, since only the Innovatek and Asetek kits seem to be easily available in the US, it might be cheaper to go the "big" route. It will come down to costs and aesthetics.

For the quite PC person....

... you have a similar choice as the casual overclocker, but would lean more towards the German-style kits 'cause they are quieter.

I happen to really like the German kits for water cooling a smaller case, like an Antec P160 or even a SFF system. Heck, look at the Y2K Bug that Mashie dd. You would never fit a 1/2" system in that thing.

I guess the bottom line is that each person much prioritize what they are looking for in a water cooling kit. Think of performance, noise, ease of use, and aestehtics. How you rate each one will determine which approach works for you.

[killernoodle]

Errm... JATO is a solid fueled booster rocket designed to be strapped to a large cargo plane for short takeoffs with heavy payloads... I guess bill thinks it will be quite restrictive

[BillA]

correctly stated DN

every product has a 'feature set', and its how the consumer weights the particular attributes that determines what is 'best', for that particular user

I do not believe even in the notion of a universally best kit (or anything)

Blackeagle
no and yes, the P-Q curve is a consequence of the decision to go with a closed impeller
(lower flow at low head, higher at higher)

pH
JATO = jet assisted take off
as in a C130 lifting off in 150' or such, film clips are stupendous

- have to work on finding you a good flow meter
I have a surplus Sparling (2), in pieces, not working; do you have access to a good tech ?
-> is there a reader who could assist ?

[bigben2k]

JATO: Jet Assisted Take Off. It's indeed a solid fuel booster rocket. There's a myth that someone strapped one to a car, then crashed into a cliff...

As for the US vs German points of view, what I see in general is the stereotypical "bigger is better" for the US, and "less is more" for "Germany" (which is actually more European, not just German). Personally, I prefer the latter, but the way we get there, is by first pushing the extreme of "how big can we make it?".

For example, we've discovered that a massively large pump actually decreases performance, as it adds a significant amount of heat into a loop.

Similarly, I'm sure Zalman is discovering a number of interesting things from developing the huge "flower" HSF.

[pHaestus]

I have a friend in chemical engineering (none in civil regrettably). Let me drop him an e-mail and see whether they have techs there who could rebuild.

[Meethoss]

Sorry, just going back to the original thought of why US typically uses 1/2" and why Europe uses smaller - I went to a plumbing shop recently and found that nothing was sold in 1/2" and instead was usually smaller. Maybe that's something to do with why we originally started down this route? (I might even be totally wrong, I was probably looking in the wrong section or something :P)

[kronchev]

Originally Posted by bigben2k

JATO: Jet Assisted Take Off. It's indeed a solid fuel booster rocket. There's a myth that someone strapped one to a car, then crashed into a cliff...

they tried that on myth busters. they got an old heavy car to about 160 but it didnt take off. 160 is kind of slow too, im dissapointed. shouldve used a stripped CRX or something, THAT would take off

[bigben2k]

Originally Posted by Meethoss

Sorry, just going back to the original thought of why US typically uses 1/2" and why Europe uses smaller - I went to a plumbing shop recently and found that nothing was sold in 1/2" and instead was usually smaller. Maybe that's something to do with why we originally started down this route? (I might even be totally wrong, I was probably looking in the wrong section or something :P)

You're right.

Lower flow rates don't warrant the use of 1/2" tubing: 3/8" is quite sufficient.

Here, state side, 1/2" fittings are more common (at least within my experience) read cheap, and aren't really an inconvenience to use.

[Cathar]

Flow rates of 0.75-1LPM eh?

Doesn't that then get into the intent of a design? The impingement designs totally rely on adequate pumping pressure to get their cooling performance.

A large number (almost all?) of the "USA" designs, mine included, are totally inadequate to cope with those sorts of flow rates. Crap, I don't even test at anything below about 2LPM, and even then such a result is just a curiosity for me as that sort of flow rate is already well outside of my targetted flow rate range, upcoming designs included. That the blocks that I make with ~1mm bases perform anywhere near as well as they do at such levels is perhaps somewhat of a total surprise to me.

Getting back to a car analogy, this is perhaps somewhat equivalent to running said Ferrari on diesoleum.

This becomes more of an issue of the focus of the testbed. A good testbed will test blocks across a range of flow rates, especially including those for which the blocks were designed to operate at.

I am now left with no wonder as to why no one understands what the sorts of performance differences between the US vs German types of blocks are. At 1LPM at 100W though, and I have almost no fear in stating that the USA blocks when run at their intended flow rates will take the lead.

[HAL-9000]

Originally Posted by Cathar

Flow rates of 0.75-1LPM eh?

Doesn't that then get into the intent of a design? The impingement designs totally rely on adequate pumping pressure to get their cooling performance.

A large number (almost all?) of the "USA" designs, mine included, are totally inadequate to cope with those sorts of flow rates. Crap, I don't even test at anything below about 2LPM, and even then such a result is just a curiosity for me as that sort of flow rate is already well outside of my targetted flow rate range, upcoming designs included. That the blocks that I make with ~1mm bases perform anywhere near as well as they do at such levels is perhaps somewhat of a total surprise to me.

Getting back to a car analogy, this is perhaps somewhat equivalent to running said Ferrari on diesoleum.

This becomes more of an issue of the focus of the testbed. A good testbed will test blocks across a range of flow rates, especially including those for which the blocks were designed to operate at.

I am now left with no wonder as to why no one understands what the sorts of performance differences between the US vs German types of blocks are. At 1LPM at 100W though, and I have almost no fear in stating that the USA blocks when run at their intended flow rates will take the lead.

I think you are right that "USA" style blocks will start beating out the low-flowers at higher flow rates. But thats because most US blocks now use two things that I think you basically invented: put the water-in nozzle over the core, and create impingement by breaking up the waterflow as it enters the block, as it comes in contact with the core. Strange thing, just about nobody puts the nozzles on the corners of blocks anymore. Wonder why?

You need high power pumps to do that effectively. What gets me is now you see some really well made and designed iterations of that philosophy in German blocks like the NexXxos 3 from Alphacool.de, yet they persist in using 8mm tubing. I wonder how well that block would perform with 1/2" tubing? To again return to the car analogy, its like putting a 2-barrel carb on a 7 liter dragster motor: for either subject, you need to "feed the beast!"

[pHaestus]

Well that's one thing that bench testing of wbs will show presumably. This graph might be relevant:

http://www.procooling.com/~phaestus/temp/wbcomparo3.jpg

I finished running the rev3 today. I put 1/2" plastic hosebarbs on it for ease of use (and for lowered flow resistance). No big surprise as far as performance goes I think considering its design.

[BillA]

you forget so quickly
considering its age (i.e. the competition), it was an Excellent wb (and I said so)

[pHaestus]

as did I (used one in my main rig for many months in fact). I meant its performance wasn't a surprise partially because I've seen you post results before.

I have a HeatKiller 1.5 here too. Trying to figure out how to keep it from being a CPUKiller; if I get comfortable w/ its mounting I'll run it this week too.

[BillA]

misunderstood, sorry
a very under-appreciated wb