Printable View
Check this website
http://www.watercoolplanet.de/index....ow=1&pagenum=1
Cascade was number one but it's now number two
I've never heard of number 1 and 3!!!
LRWW is 6th
microchannel seems to be the hype now
ps It's in german but can someone check it and tell me how they tested the blocks and what system are they using?
Old news. There was a lengthy thread over at procooling on this found here: http://forums.procooling.com/vbb/sho...&threadid=8005 The tests are suspect, the heat die was changed mid test and they test with a 1046 pump. Way too weak for the Cascade and they changed the barbs.Quote:
Originally posted by antipop
Check this website
http://www.watercoolplanet.de/index....ow=1&pagenum=1
Cascade was number one but it's now number two
I've never heard of number 1 and 3!!!
LRWW is 6th
microchannel seems to be the hype now
ps It's in german but can someone check it and tell me how they tested the blocks and what system are they using?
Not for me :)Quote:
Originally posted by nikhsub1
Old news.
I got some surprise looking at the score of some blocks there, and i couldn't find on the website how they did the tests
I've checked the link and we're not talking about the same block this time
Issues with the tests at that site:
1) (major) At some stage the heating elements were changed invalidating many results
2) (potentially major) An Eheim 1046 pump is used, running through a continuous tube loop radiator of around 26 x 180 degree elbows - flow rates would be lucky if they were above 2LPM before a block is even attached. Need to assess whether or not this represents the type of setup that you have.
3) (major) The blocks are only mounted once, with no control to measure the variability in mounting, or for thermal paste settling.
4) (minor) The test-bed favors smaller blocks as the heating element pipe is uninsulated allowing for more convective cooling of the heating element
Specific to the Cascade tested:
*) The stock barbs were changed
The test is one of a mosaic of tests out there, but it is important to realise the vast chasm of differences both in hardware and testing methodology between this test bed, and a test bed like BillA used, despite there being superficial similarities.
The same issues as last time.
Note that the other block they said they found to outperform the Cascade is not even to be found in the top 20 now! I have to wonder if Old Thomas asked to have it removed.
#1, 3 & 4 in Cathars post are the ones most glaring to me. They could retest the Cascade with the new heat element they are using in tests now. And why fail to insulate the heat element for most consistant results?
And they have got to know that only with multiple mounts can they claim to minimize variations in the TIM effects on outcomes. They chose to disregard this fact in favor of saving some time. This should cause anyone (everyone?) to have second thoughts on the remainder of thier testing methods.
That sites failure to correct these glaring errors in thier testing is, to me, indicative that they do not care about improvements in accuracy. The web is to crowded already with such "good enough" testing sites, we don't need any more.
I really don't fault them for the other points Cathar makes. The use of smaller lines, and smaller pumps are very widespread in europe. It's only reasonable for them then to use the pumps and lines (fittings) that users in europe would be very likly to use.
While I can understand Cathar not liking the changing of the barbs in the Cascade, would a line size adaptor (8 => 12mm perhaps?) just before the block change the affect of small lines on his Cascade?
Anone know why they are still making use of a rad of that ultra restrictive type. I have to doubt such rads are still much used in europe. As they have recently changed the heating element they should upgrade the rad now as well. Along with some insulation for the heating element and multi mounts of the blocks in the future they could greatly improve the accuracy of thier tests.
Then do some retests of at least thier top ten.
With regards to the size of the lines and the radiator used, that is fine. My point was just that one needs to be aware that such is not really common amongst enthusiasts outside of Europe, and when reading the review, one needs to look at the setup that they own, or that they are considering on purchasing, and then decide if the test equipment is an adequate indication of what they intend to use.Quote:
Originally posted by Craig
I really don't fault them for the other points Cathar makes. The use of smaller lines, and smaller pumps are very widespread in europe. It's only reasonable for them then to use the pumps and lines (fittings) that users in europe would be very likly to use.
While I can understand Cathar not liking the changing of the barbs in the Cascade, would a line size adaptor (8 => 12mm perhaps?) just before the block change the affect of small lines on his Cascade?
With regards to the different barbs, there is a small plenum chamber atop the jets inside the Cascade. Now depending on how the barbs were installed, this plenum chamber could be artificially restricted without the installer of the barbs being aware of the importance of how it's meant to be setup. Now this may or may not have been an issue, but when reviewing someone's block, it is usually common to at least ask the block's maker if a modification passes approval, which is not what happened. I was unaware of the test until the results were posted. Generally the Cascade will work just fine with smaller barbs and tubing, so long as the barbs have been installed correctly, which in most cases, requires cutting the barbs down as I do.
they need to drop a heatercore, a rainbow lifeguard quiet one, some 1/2" tubing and fittings in that loop and re-run the test.
I already know where I'd bet my $ under those conditions. :)
IMO changing anything in desgn of blocks is counetrproductive and misrepresenting - why not drill some dimples in the base then or maybe plain it a bit to make it thinner? ;) - ridiculous!
Btw, there's an interesting Eheim mod there, have you seen it?
What does it do? (Eheim Mod) My German sucks :p
In france in one of the most frequented oc website, tehy swear by those test! I asked the question here because i would get flamed there :(
In france a lot still use bix and alike, small tubing and weak ass pump! It's hard to get into an argument there
Cathar, can you tell me your opinion about the micro channel wb, they seem to be the last hype in france?
ps have you noticed that number 1 and 3 are german?
Micro-channel is one design aspect. True micro-channels, which I coin as channels that have one dimension <0.25mm is a very powerful design, but one needs to keep an eye open on the restrictiveness of the design.Quote:
Originally posted by antipop
Cathar, can you tell me your opinion about the micro channel wb, they seem to be the last hype in france?
ps have you noticed that number 1 and 3 are german?
True micro-channels blocks excel at ultra low flow rates (<2lpm), which is co-incidentally the sorts of flow rates that this test is targetting, and that the Germans tend to design for. When only targetting 1-2LPM, it's pretty hard to argue with true micro-channels, but true micro-channels don't lend themselves well to higher flow rates and the benefits that come forth-with.
Mini-channels, being 0.5mm->1.5mm in width aren't as good as micro-channels at lower flows, but are far less flow restrictive. Coupled with an impingement effect (ala White Water), a hybrid mini-channel/impingement block can compete very well with true micro-channel blocks at lower flow rates, and in my opinion, will start to pull away as the flow rate is increased as the impingement effect really starts to come into its own. This is what makes the White Water design so powerful, and that is its flexibility to work well across a wide range of scenarios, and its ability to support such.
The WCP test-bed as it stands present an extremely narrow view of the water-cooling performance spectrum.
As so often happens in any industry, if all that a company has to independently measure its performance by is a single benchmark, then they will optimise for that benchmark. Is it then any wonder then that the blocks that perform really well on Germany's "most trusted" test-bench are all made in Germany, and are all true micro-channel designs?
Without a greater spectrum of flow rates being represented, how can any other design show its strengths?
The test-bed, as it stands, is akin to basing all determinations of how good every car is by their 0-50kph (0-30mph for you US lads) acceleration time.
Will get back to you on thet one later since I don't want ot mislead you (my partner is German :) ).Quote:
Originally posted by Belgian_Dude
What does it do? (Eheim Mod) My German sucks :p
As far as could make out of it it increases response of the pump (it looks like the impeller is slightly loose there and b adding some PTFE (teflon tape) they mnaged to make t very tight.
It may shorten the life span of the pump since in this way impeller section will be geting much higher momentum loads but since it's magnetic driven impellr it porbably doesn't matter.
Anyway, as I said will get back to ya in ~7 hrs :)
Cathar -> nice analogy mate :)!
Btw, wouldn't it be better to produce more than one version of a block to suit different flow rates?
As in every engineerig/construction jobs targeted purpose designed solutions are always the best ones :)
Thank you Cathar for those informations,
Jabo i'm waiting to hear more about the mod
Now Ladies and Gentlemen, prepare yourselves for a brilliant bit of ingenious German engineering.
Got this article explained to me and here is what was done there.
This mod is aimed to simply prevent rattling and clacking noises (as they put it ) coming out of the pump when some air enters it and loiters around :D :D :D :D :D :D :bounces: :bounces: :wierd: :wierd: :wierd: :wierd: :am: so much for oversophistication and trying to be more scientific than Science itself !!!!!
I almost pissed myself when my partner translated it to me !!! :D :D :D :D :toast:
Good , eh?
:rolleyes:
afaik, ehiems, cos of pond usage, have a very slightly movable impeller so there is some give if a piece of grit or whatever els eyou might get in a pond, hitting it. As a watercooling sys will be jsut water, no grit nothing else, then the feature is not needed. Teh fact that the impeller can move means it can rattle and cause water sounds as it moves about, didn't check, but i assume thats what this mod is. Think a bunch of people jsut glue the impeller down properly, can't hear mine so never bothered.
Anyway, just as a quick q, as there seems to be a lotta mention of europe + certain rad's, pumps being used. haven't paid much attention to watercooling lately, are there plenty of good new rads out, that by the way you said, aren't really available in europe?
The last few month I'm working on getting the European people to see the other side of the coin.
Those Germans with their small pumps and stupid 8mm hose make me mad.......
In the beginning of this year I've started out reselling Innovatek......I've sold alot of sets but had to get better stuff....
So the next step was Dangerden.....nice product and they're going like warm rolls..$$$$$
But still I wasn't satisfied with the preformance..... so started looking for better stuff..... the WW came and finally I had some good stuff....
Now Silverprop has joint the list and a Cascade auction is going at the moment.....just for promotional purpous and letting the dutch and Belgian people get to know some good stuff.....
Also importing Tygon and that sells like crazy.... so there's a market....but still it is amazing small....
Aquacomputer and Asetek with their waterchill are f*ckin-up the market with their low-end sh*t.... :upset: this week I've removed Asetek from my website for that isn't IT.
Still, here in Holland, every one believes flowrate doesn't matter.... I try to proof them otherwise..... so next week a salesmanager of Iwaki comes to visit me and I'm starting to offer the "über pumps" besides Hydor and Eheim.... btw the 1048 is the best selling pump in Holland....I sell only 2 1046 a month.
Now I've come up with a total new idea...... I'm starting up a Distribution channel for high-end preformance watercooling products (nice puzzle word)...... And I'm working around the clock to get in contact with watercooling companies which have outstanding products.......
Now do you guys have some pionters for me on which radiators to carry besides hwlabs..????
JPiperfomance is in the picture but he has some production delay.......
Now I know alot of good stuff comes from Down-under.... are there some otherbrands I should look at???
Best Regards,
Jerome
http://www.tweakstore.nl
Cathar,
I mistook your first posts point on the barbs used. Thought you were only concerned with the diamiter, not seating depth. I can well see how incorrect seating depth could screw up results.
With those Iwaki's you'll have one of the best pumps made for watercooling.
Heater cores are unbeatable for price/performance, perhaps do a bit of research and find a core that is both high performing and at reasonable cost to yourself.
As to getting out the word better, perhaps some independant test of one of your set ups vs whatever is decided on that is representative of the small lines, small fittings & rad system with a low power pump.
Cathar's Cascade is still number ONE in my book! Give me a pump like Niagra Falls, 1/2 tubing, my Chevy truck heater core with some screaming 120mm fans and Cathar's Cascade and I'm a happy Duder. :D
OMG, I just saw some of the prices offered at the auction.Quote:
Originally posted by The TweakFreak
a Cascade auction is going at the moment.....just for promotional purpous
I wonder what an "Antiques Roadshow" assessor would make of one? :eek:
hi all
hey the tweek feak, i have a question about the silverprop evolution evo, how is it compared to the maze4 and the swiftech mcw5002
Quote:
Originally posted by Cathar
OMG, I just saw some of the prices offered at the auction.
I wonder what an "Antiques Roadshow" assessor would make of one? :eek:
"Antiques Roadshow" assessor in the year 2563:....'
Mhhhh...what we have here is a sculpture from the early 2000's......after some research we're sure it is an original 'Cathar'...in that time a famous australian artist...
In those days people didn't have CPU-inplants yet and cpu's still produced heat.....:D
As the for the prices offered at the auction...... please Stew don't forget that getting blocks from Australia costs a 19% extra, due to taxes.... so the amounts offered at the auction is about the same I've payed for them( incl. transport)
It's the first for me either :DQuote:
Originally posted by TheDude
Cathar's Cascade is still number ONE in my book! Give me a pump like Niagra Falls, 1/2 tubing, my Chevy truck heater core with some screaming 120mm fans and Cathar's Cascade and I'm a happy Duder. :D
The block performs like a best in my comp and i've not even completed the whole setup
I get 43°C under load with a 1700+@2.4GHz and 2V (ambient 25°C)
Hi all!
Do you have EVER build a test like WCP and try different pumps? I don't think so...
The typic american style 'bigger is better' is f***ing b******it! OK, Blocks like the Cascade can get better with stronger pumps - but the 'more' effekt will be killed by the reading-out-errors.
One very useful reply, can someone explain me the last sentence?Quote:
Originally posted by Nirwana
Hi all!
Do you have EVER build a test like WCP and try different pumps? I don't think so...
The typic american style 'bigger is better' is f***ing b******it! OK, Blocks like the Cascade can get better with stronger pumps - but the 'more' effekt will be killed by the reading-out-errors.
Have you seen and tried cathar blocks, it's not huge at all!
ps i'm sure that you're german :)
Yes, I am a german - my english is not very good at all.... When you test something, the sensors do not work with 100%, the produce errors, something in 0,1°C, nothing special. But if you test a stronger pump, this 0,1°C can be very important - Cascade blocks do not get 3°C better with an 1250 eheim...
I don't see why the 0.1°C does make a diff with higher pumps but doesn't with other pump?
Then the cascade performs with a 1048 as well as with a 1250
And thats the Problem: Most Americans use 1/2 " tubes and eheim 1250 oder 1060 - they think, that theire blocks perform better. But thats not realy right - the better °C is 'killed' by the errors at the teststation!Quote:
Originally posted by antipop
Then the cascade performs with a 1048 as well as with a 1250
So you're saying that the teststation the website have put up is the best?
They're a lot of people doing test on waterblocks and some do a wonderful job (like the guy @overclocker.com but i forgot his name)
Oh man, don't start with this. It is a proven FACT that more flow is ALWAYS better provided the pump heat does not overwhelm the system. Just go look at ANY of BillA's waterblock tests. The higher the flow, the better the block performs, end of story.Quote:
Originally posted by Nirwana
And thats the Problem: Most Americans use 1/2 " tubes and eheim 1250 oder 1060 - they think, that theire blocks perform better. But thats not realy right - the better °C is 'killed' by the errors at the teststation!
nikhsub1, i don't see what is his point? I don't understand how more flow equals to more temp error
I don't either but I get the feeling he is saying that 1/2", high flow systems are a myth and really don't cool any better than an eheim 1046 with 1/4" ID tubing with ten or so 90 degree fittings to boot :rolleyes:Quote:
Originally posted by antipop
nikhsub1, i don't see what is his point? I don't understand how more flow equals to more temp error
Here in Finland the usual thought is also thet the flow means nothing to the MAx preformance of WC system....People suggest 1048 is the pump to have! Even I have a Hydor L20(coz its the cheapest decent pump) mut I would realy want a Gundforss centrall heating pump....the average lifespan of those mothers is somethin like 25 years so it will outlast yor Computer easyly...and the flow and silence is next to none!
man, here we go again. i tell you, the mass of those little @ss german block can only do so much, anyway there is only so much flow that can get through those small blocks do to the sheer lack of size:stick: get with it german dudes! go out, get an iwaki and hook it up to a real water block with the aid of a big bad @ss heater core or a thermochill 120.3 and you'll see what we are talking about in this side of the world:stick:
...and you're basing your comments on what testing information?Quote:
Originally posted by Nirwana
Cascade blocks do not get 3°C better with an 1250 eheim...
Actually the difference between 1.5LPM (choked up Eheim 1046), and >6LPM (Eheim 1250 with a non-restrictive radiator) would be approaching 3C at the Cascade block itself for a hot CPU.
Much above 4LPM and you get diminishing returns. As you get below 4LPM, the losses do become more noticable.
An Eheim 1048 with a heater-core style radiator is the least that I personally recommend with the Cascade if you want to be staying on the correct side of the performance knee. One would see a little over 4LPM with such a setup. The Cascade block still performs very well even at very low flow rates, but it's just that at >4LPM you're a lot closer to where it was designed to be used.
This is exactly the point. I've run the Cascade at 1LPM right through to 10LPM and measured the differences in my testing, and yes, approaching a 3C gain is not out of the question between 1.5LPM and 6LPM. As you move past 6LPM the gains do start to drop off with less than a 1C difference seen between 6LPM and 10LPM on a hot CPU.
The block and pump have somewhat to do with the cooling capabilities, but the cooling fan cfm and positioning is what is going to make the biggest difference. I'm currently running a Thermaltake aquarious kit on a KD7 with a xp2400 running at 2.025 volts and it's maintaining a constant temp of 35c under full load with ambient temps in the room of 70 - 75F.
Now the reason I state that is because the TT kit is running on 1/8" ID tubbing, a 3" x 3" dbl pass radiator, and I really don't know what the flow rate of the pump is, but I could spit more than the pump puts out, but I've got it mated up to a 120v furnace exhaust blower that puts out about 350 to 400 cfm and the exhaust of the fan is aimed directly at the motherboard to keep the mosfets, capacitors, and nb cool at the same time. Now this setup is not mounted in a case, it's actually screwed to a peg board in my basement, with the rest of the components laying on a shelf. :cool:
Just thought I'd throw that small thought in there for you guys. :D