Nice work Cathar. As an engineering grad, I love it when math/physics is used to prove/disprove a statement of fact.Originally Posted by Cathar
Provided your math is correct, I'd love to see Koolance and their fanboys weasel out of this one. Spin as they may on the testing conditions & the application of those results towards PC watercooling, falsifying data is as unethical as it gets.
Last edited by ranker; 06-17-2007 at 06:49 PM.
MM Extended U2-UFO CYO (Duality front, Standard back, Horizontal Mobo brace) Anodized Black || eVGA X58 || Intel i7 920 || 6 GB Corsair Dominator PC3-12800|| eVGA 295GTX || Asus Xonar Essence STX || VisionTek 650 TV Tuner || 1 300GB WD Velociraptor || 1TB WD's Black Ed. || LG 22X DVD-Writer || Lite-On 20x DVD-Writer || Corsair CMPSU-1000HX PSU
CPU Loop: DDC-2 w/ XSPC Reservoir Top -> Swiftech GTZ -> Thermochill PA120.3 (Push: 3 Scythe S-Flex G)
GPU Loop: DDC-2 w/ XSPC Reservoir Top -> (Koolance VID-NX295 FC block) ->Thermochill PA120.3 (Push: 3 Scythe S-Flex G)
Chipset Loop: DDC-2 w/ XSPC Reservoir Top -> Alphacool Silentstar Dual HD waterblock enclosure -> Bitspower Black Freezer eVGA x58 MOSFET-> Bitspower Black Freezer eVGA x58 NB -> Thermochill PA120.2 (Pull: 2 Scythe S-Flex G)
2 x Dell 2408FPW LCD || Logitech Z-5500 Digital 5.1 || Logitech G9 Mouse || Logitech G15 LCD Keyboard || Logitech Quickcam Ultravision || Sennheiser Headphones
It's all irrelevant.
Check this post if you haven't seen it already.
Either they're completely lying, or the test is completely invalid. They can choose which, and then decide how they might best like to withdraw their false claims.
So, the question is in what regards. Is it a complete lie or just an exaggeration? Guess it doesn't really matter .
Using the formula you used. If you bump the air flow up to 10m/sec the exhaust air would have to be 50.5. So, is it possible they posted results from a higher air flow rate? ( that is if I did my math properly.)
Dateranoth
EVGA X58 mobo
6gb RAM 1600mhz cas9
Core i7 w/ D-Tek FuZion @3.9Ghz (43C idle) (65C load) (27C ambient)
EVGA 680GTX SC w/ MCW-60 and stock Heat Spreader (With some Modification) (32C idle) (53C load) (27C ambient)
Water Cooling Loop:
1/2" Tygon Tubing
Laing DDC w/ Petra'sTech DDCT-01s Top >Swiftech MCR-320 w/ 120mm Yate Loon D12SH-12 x3>D-TEK FuZion>MCW-60>Swiftech MCRES-MICRO
My Worklog
Shop Petras
Does anyone see the AirOut tempt or at least the dT of AirIn to AirOut on the KATECH report?
EVGA X58 mobo
6gb RAM 1600mhz cas9
Core i7 w/ D-Tek FuZion @3.9Ghz (43C idle) (65C load) (27C ambient)
EVGA 680GTX SC w/ MCW-60 and stock Heat Spreader (With some Modification) (32C idle) (53C load) (27C ambient)
Water Cooling Loop:
1/2" Tygon Tubing
Laing DDC w/ Petra'sTech DDCT-01s Top >Swiftech MCR-320 w/ 120mm Yate Loon D12SH-12 x3>D-TEK FuZion>MCW-60>Swiftech MCRES-MICRO
My Worklog
Shop Petras
Looking through their results in the PDF
@ 3m/s air-flow, & 5LPM flow rate
Water-in = 84.86C
Water-out = 58.55C
Air-in = 24.59C
Dissipation = 8860kW
Munch the maths, and the exhaust air-temperature mus be 105.56C
i.e. the air would have magically heated up to >20C above the water-inlet temperature.
@ 5m/s air-flow, & 10LPM flow rate
Water-in = 84.14C
Water-out = 63.89C
Air-in = 24.28C
Dissipation = 13760kW
Exhaust air-temp must be: 99.73C. i.e. Exhaust air, by all existing physics on the known thermal properties of air, is now magically 15C above the water inlet temperature.
Cathar to the rescue, as always. Awesome stuff.
All things aside, I was thinking of another variable that might need to be considered in terms of why the results are impossible: mistranslation. Surely if the test was done in Korea, then test results were published in Korean and then translated. Maybe there was a mix up in the translation.
Case & PSU:
Lian-Li PC-G70 Black w/ Full Window :: Modified for Wire Routing
PC Power & Cooling 510-SLI PSU :: 5x120mm Yate Loon 12SMs
System Core:
Intel Core 2 Duo E6600 @ 3.4 GHz (1.45v - Stable) :: Watercooled
Abit AW9D-MAX @ 378 MHz :: 2x1GB G.SKILL PC2-6400 (2.1v 1:1 4-4-4-10)
Storage & OS:
WD 4x250GB Caviar SE16 (RAID 5) :: WD 2x250GB Caviar SE16 (Single Drives)
Windows XP Pro w/ Service Pack 2 :: Lite-On DVD/CD-RW & Lite-On DVD+RW/-RW
Multimedia:
eVGA 7900GTO 512MB @ 700/1530 MHz :: LG L246WP 24" LCD Flat Panel
bluegears b-Enspirer HD 7.1 :: Creative Labs MegaWorks THX 2.1 550 Speakers
Watercooling:
D-Tek Fuzion CPU Waterblock :: Swiftech MCW60 GPU Waterblock
Thermochill PA120.3 Radiator :: Swiftech MCP655 Pump :: EK-250 Reservoir
I don't believe that. That PDF is a scanned document with stated units of measurement on it, which are international engineering standards regardless of what language the operators are, with original signatures. Koolance, in scanning the original document, are no doubt trying to demonstrate that the results are exactly as received in the report from KATECH.
Edit: I personally think it's far more likely that they're going to try to claim operator error by some junior employee as a result of the fallout on this.
Ooooh I didn't notice that (eh I didn't even look through the PDF lol, I just saw that the water was 84C or something and the 9kw and 5kw figures, which brings us to a 50something C water discharge temp)
Lenovo Thinkpad X220 - Core i5 2410m, 4gb
waiting on 28nm video cards...
Hi Willy - (am assuming that's you?) - Monday morning arrives and the blurry goggles of Mr J.Daniels have now worn off... (cor what a weekend... my wallet's lost a HELL of a lot of weight... TT Highlights, then British Superbikes, then F1 forcing me to spend the entire weekend on the sofa with a bottle or three, a glass, and all phones unplugged - the heartburn may take a while to subside - should start drinkin' it with milk!)But according to your numbers, are the Koolance figures accurate or not? Cathar alluded to a lack of accuracy.
We certainly see discrepancies from our end.
Yes, discrepancies in abundance without clear definitions of the "how". Have amended my original statement once more... however, were any of OUR figures (ie: Yours and ThermoChills) produced by an identical testrig to Koolance's - or indeed, anything remotely similar? I'm willing to accept that a testrig of such a relatively huge scale will produce unseen data that will conflict with that gained from the methods we employ, so looking ONLY at their data, I have no hard reason to question it's accuracy due to it's irrelevance, and it's internal accuracy is only relevant under those conditions on that specific testrig (if you get what I'm aiming at... the eyes aren't blurry, but the head still may be).
We've always said data produced on different rigs cannot be compared directly for such reasons, altho previously the differences between two rigs have never been exemplified in such a vast manner...
Could really do with BillA making a brief reappearance to add in his thoughts on the matter, as he is our "tester of choice" and would be able to comment on the differences in hardware and methodology that may account for such differences... altho even he may struggle to explain why laws of physics are being mysteriously warped as they pass thru the Katech continuum.
Last edited by Marci; 06-18-2007 at 02:38 AM.
Brass that we use is 151W/m-C (87.23k). That's 85%Cu, 15%Zn - specific heat of 380J/kg-COriginally Posted by MaxxxRacer
Last edited by Marci; 06-18-2007 at 03:09 AM.
barometric pressure (mA) and relative humidity are both measured and accounted for in testing done by BillA... just not noted on the published graphs as the relevance of such to the enduser is minimal.Originally Posted by V2-V3
David's (Rosco) summation - http://www.google.com/translate?u=ht...&hl=en&ie=UTF8
Okay, I'm going to have to withdraw this statement, as it's false. I had heard it a long time ago, and always believed it. Someone wrote to me an asked me to prove it. I broke it down to first principles and established a logical progression sequence of water molecules travelling top-to-bottom on a 2D plane, and air-molecules travelling from side-to-side acrosss the 2D plane, and then constructed some software to simulate it.
It turns out that it's quite possible for the average exhuast air-temperature to asymptote towards the water inlet temperature, while the water discharge temperature is significantly less than the average exhuast air-temperature. The disparate thermal capacities of the two mediums means that the water molecules that the air first strikes (i.e. at the entrance of a radiator) can transfer so much heat that the air warms up to pretty close to the water inlet temperature by the time it reaches the exit of the radiator, and so while the water molecules at the front air-entrance-side of the radiator are being cooled, those at the air-exhaust-side are barely being cooled at all 'cos the air has already heated up. In this way, it's possible for the air exhaust temp to approach the water inlet temp, while the water discharge temp may be less.
It is (of course) impossible for the air-temperature to ever be more than the water-inlet temperature though. So the proof that they are defying the laws of physics still holds true based upon the analysis in this post.
i.e. What they're claiming is still impossible for various of their data points, just that the singular data point that I first pointed out is theoretically possible.
My apologies for the confusion.
Last edited by Cathar; 06-18-2007 at 01:47 PM.
I was leaning towards air air leak myself, until some posted that the numbers corrolate with 10m/s.
It's easy to mess up the mount of the core, at the other end of the blower, in a test rig like that. The pictures don't inspire any confidence, for me, on that point.
i think they had too much sangrea and just made it all up to make koolance look like lemons *nods*
after the PSU i'm planning on watercooling the ball bearings in the fans...
Isn't that guy a certain called [FLOW] ? Ten hours ago, he showed me your replies here in commentaries on CM and I explained him exactly what you are talking about because I saw that lack in your explanations. I checked directly their values too into Excel to see if any aberration could be possible because of their results. Calculations are good but a bit too simplistic face to reality, it's only an average energetic equivalence from water to air, but reality is slighty different like on my quick drawing below (I let you imagine the more complex 3D repartition of the exhaust air temperature field for a 2-pass rad). But as you said rightly, the mean calculated T° is enough to make some preliminary assumptions, but for the Koolance rad, these values are effectively strange when we compare them to inlet water T°. The average exhaust air temperature value is already above the max temperature possible (other values for TC & Hwlabs are physically possible, I saw no problem IIRC). Max exhaust air T° can't never be above water T° at inlet, it will be anti thermodynamical, you'll create a blackhole lol
Last edited by rosco; 06-18-2007 at 04:27 PM.
*:-.,_,.-:*'``'*:-.,_,.-:*'``'*:-.,_,.-:*'``'*:-.,_,.-:*'``'*:-.,_,.-:*
GTZ --> MCW-NBMAX --> EK FC --> PA 120.3 --> PA 160.1 --> 2x DDC Ultras in Series --> Custom Clear Res
"Artificial intelligence is no match for natural stupidity."
*:-.,_,.-:*'``'*:-.,_,.-:*'``'*:-.,_,.-:*'``'*:-.,_,.-:*'``'*:-.,_,.-:*
Don't know a FLOW guy. People use different names all over the place online. Might have been?
Yes, Rosco, the diagram that you just drew is pretty much exactly what I just modelled in software. As air-flow is increased, it's far less likely for the air to get close to the water temperature. It all depends on the relative thermal capacities of the air-flow and liquid-flow, and the shape of the 2D array. It's a very complex arrangement.
The simplified maths are enough, however, to show that the average of the air-discharge temperature is above the water inlet temperature, and that cannot ever be possible.
Agreed. I also computed for HWLabs and Thermochill, and all those results are possible. Also computed for the water temp deltas matched to the flow rate and claimed heat dissipation, and found that for all radiators that the results are in alignment.
It's just once we consider the air-flow aspects is where we find discrepancies.
It is possible, I guess, that the air-speed meter is wrong, and the air-flow is even much higher than what we've calculated. ie. in the order of 600cfm or so, instead of 330cfm, when they're talking about 5m/s air velocities. That 5m/s might be totally wrong, or may be being measured at a different point within a larger chamber?
The problem with that though is that the air-pressure drops across the radiators then make absolutely no sense. It is, as if, the air-flow was a lot ligher (>600cfm) and there were air-flow leaks meaning that a lot of air-flow was bypassing the Thermochill & HWLabs radiators. Then it'd make sense. The only problem with that though is that at 600cfm, the ventilation resistance for the Koolance radiator doesn't make any sense at all.
Then again, that might be explained by a faulty air-pressure sensor, and that, coupled with a faulty air-speed calculation, coupled with faulty/leaky mounts of the radiators, would then explain how the results were arrived at. In which case, there's so many systemic errors that the results are totally invalid anyway.
No matter how you draw the bow, the results are rubbish.
guys why dont we ask petra or billa to make tests about this issue... i liked the way petra did test for all the pumps and alex is a honest guy and im sure their test will be valid and bias free
i see the koolance response as a joke and a stunt to make them look like the best radiators out there.
Incoming new computer after 5 long years
YOU want to FIGHT CANCER OR AIDS join us at WCG and help to have a better FUTURE
Bookmarks