Koolance's Response to Hardware Labs "The Copper Radiator Myth"

[_G_]

Wow, plz don't feed the trolls.

click me

[Cathar]

Oh, something I wanted to ask to Cathar (I believe he had something to do with the design of the PA series rad) and to Marci.

I was the primary designer & tester of the PA120 radiators in terms of providing guidance on what to change to make them better. I ran competitive performance tests, ran analysis of prototypes, devised theories for the ideal mechanical attributes for the radiators, researched fan power, noise, back-pressure targets, and through a series of iterative prototype analyses provided the design suggestion feedback that determined the path of the product. Marci handled the manufacturing & engineering side of things. Through his team, he was responsible for doing what had to be done to take the set of design guidelines that I was putting forwards, and determine how to best map them into a physical reality using the highest quality components. Some of what I was suggesting wasn't possible, so Marci and I collaborated until we came to a mutual understanding of how to best combine the physical and theoretical realities into a real item.

How would an all aluminium PA series radiator compare - a few adjustments can be granted, but essentially the same product.

In other words, is there anything at all in Koolances statements about aluminium. I'm happy to accept - thanks to the studies done - that a combination of their materials, manufacturing process and design do not produce a superior radiator.

This is NOT an easy question to answer. We must always factor the properties of the coolant into the overall performance aspect. While it may be possible to design a superior aluminium radiator for pure water use (I really don't know - I didn't research that) the reality is that copper/silver makes for the best performing affordable waterblocks, and we simply cannot mix aluminium and copper together in a pure water system without incurring devastating corrosion. There must alway be some amount of anti-corrosion protection added, and this will always negatively impact the overall performance of both the radiator and the waterblocks.

Further, since it cannot be guaranteed that users will always take the required steps to provide corrosion protection for their systems, this then automatically invalidated the choice of aluminium as a potential radiator material, and that is why I never researched it.

Without knowing the amount of corrosion protection required, and the level of performance degradation incurred as a result at the waterblocks, it is impossible to state which is the higher performing item. We may conduct short-term tests using pure water, but the reality is that an all copper system can be run pure-water-only, while a mixed-metal system must be run with coolant. This is then another flaw in any testing methodology that does not account for these requirements. If we're testing aluminium against copper, we should also be running 50:50 coolant/water for the aluminium scenario, and pure-water for the copper scenario, because only THAT will reflect real-world use.

So, you see, the question is not easy to answer, and due to the non foolproof nature of the use of aluminium in a non-pre-packaged product, we didn't even consider it as a realistically viable end-user solution.

[migueld]

However, don't ever think we're sitting on our laurels. There are distinct advantages to not relying on a sub-contractor to do your engineering and manufacturing work for you. It costs a substantial sum for a set of tools to generate each new fin geometry. Speaking of which, they will be arriving soon.

I'm looking forward to the new designs. I'm also glad you are investing in R & D and are willing to test the competition. I'm often surprised how some if not most of the w/c manufacturers rarely take time to study the competition.

[Jimmer411]

I hope it's not off millerlite.

Ive never considered adding anyone to my ignore list before today. The ignorance in your posts is borderline stupidity




As for radiators, I dont choose sides. I run what ever offers the best price/performance. Ive never considered Koolance because their products don't suit my needs. Awkward size radiators, plastic water blocks that aren't aesthetically appealing IMO as well as a higher price tag for the performance that you are getting. Plus I was hooked on the whole 1/2 > 1/4 ID tubing thing... Even with the latest tubing info, I wouldn't switch over to Koolance. Not because of its PR, or the community dislike towards it but because the performance that it offers isn't enough to justify the difference in price between upgrading my current custom water cooling setup VS purchasing their kit.

Which is why I just replaced my dying Mag3 and DD Dual heatercore with an MCR320 and MCP655.

[Halk]

So, you see, the question is not easy to answer, and due to the non foolproof nature of the use of aluminium in a non-pre-packaged product, we didn't even consider it as a realistically viable end-user solution.

Honestly that reads like a cop out. Which is really strange, because from both your recent and your historical posts you're quite free with your knowledge.

I was under the impression that flow rates (past a certain point) had little or no real world impact. Thus if an additive negatively impacts on the thermal properties of water, then it's going to be along the same lines of a minimal drop in flow rates? (Right? I don't know if I am, I think I am... :P)...

So from that it can be said that if there's a real world difference from using all aluminium in a radiator rather than the brass/copper solution then there must be a superior and inferior way.

So is there any chance that using all aluminium, assuming an optimal design, is better than the current way?

Or, and this could well be the case, I've taken what Koolance said for granted, and it's not the case that ThermoChill/other competitors use a combination of different metals that compromises something.

Edit : As for people who don't know better using aluminium. Well, should the rest of us be subjected to inferior (if indeed they are) products because Joe Public doesn't know to add an anti corrosive? It's like the massively over quoted "Censorship is denying me a steak because a baby can't chew it" (paraphrased because I can't be arsed looking it up).

[hwlabs]

An aluminum equivalent would need to be larger, unless you believe what Koolance says.

[Halk]

An aluminum equivalent would need to be larger, unless you believe what Koolance says.

Ok. I don't understand why. But, ok it's larger. What's the performance like. Are there any benefits?

[Bbq]

To SDatl404: We know, koolance, under those conditions, will perform better. That's not the point though: Answer me this: Who runs their watercooling systems with 300+ CFM fans, with water at 85c + ? The tubing isn't even rated to that high. Crank it down to realistic values: 35c water, <100cfm fans. You'll clearly see that the HWLabs and Thermochill will outperform the koolance.

[migueld]

You'll clearly see that the HWLabs and Thermochill will outperform the koolance.

Prove it, I'm tired of people making bold claims with no first-hand data. I don't like aluminum, but if you are gonna bury them show me some data.

[Cathar]

Honestly that reads like a cop out. Which is really strange, because from both your recent and your historical posts you're quite free with your knowledge.

I actually take offense to that.

I stated that I didn't know the answer. I stated the difficulties with determining the answer if it was considered. I stated the realities of mixing metals in an uncontrolled system environment. I stated that the final decision was not to pursue aluminium based upon the difficulties of ensuring trouble free operation. That is no more of a cop-out than deciding not to make radiators out of a vastly superior full brazed pure-silver construction because it'd cost too much for anyone to buy. Is it now a manufacturer's fault for not offering something which it deems to not be a viable and cost-effective long-term solution? It's a design decision based upon a cost-risk-benefit scenario.

Truly, what you just stated is offensive. I could not be any more transparent with my knowledge and the reasons for the decisions. I can not and will not answer what I don't know, and I won't speculate.

I am sorry that's not good enough for you.

[hwlabs]

We will need to validate Koolance's claims with testing using their own parameters.

How a similarly sized heat exchanger with a higher pressure drop on both the liquid and air side can perform over 50&#37; better is very very very improbable.

[lee143]

Trolls are green right? Kinda like the color of that pc....



I don't under stand either why koolance didn't do any realworld test. Maybe because you can't see a differance in all the radiators until it gets high up!

The reason there isn't many new koolance user in here:

1) Everyone here constantly talks smack and says they are junk 24/7. 99% of them know nothing about radiators and have only used one in their life. Of coarse this would be a thermochill, because every likes them sooo much.

2) Most users here like to build there own custom units. Koolance is more designed around the ready made solutions for people with less time or less knowledge on building things.

How about you take your own advice and bring proof that koolance is better?
1. I guess you're part of the 99% because you've only used koolance.
2. I will openly admit that koolance probably makes it easier for first-timers to use their products, HOWEVER, that is the tradeoff for performance.

[Cathar]

We will need to validate Koolance's claims with testing using their own parameters.

How a similarly sized heat exchanger with a higher pressure drop on both the liquid and air side can perform over 50% better is very very very improbable.

Well, I already pointed out that the Koolance stated liquid pressure drops for the PA radiators is not even remotely in alignment with independent tests, nor is it for internal test data. You might like to look at your own in-house data for the GTSealth and determine if their stated pressure drop data is valid. That'll give you a personal inkling as to validity of even these stated values.

[hwlabs]

Actually was going through our numbers earlier and they are just as remote.

Lost in translation?

[Jedda]

Strewth you've got to wade through some crud to find the few posts worth reading.
This thread especially would've benefited from active moderation.

[septim]

bah this is goin on and on and on with you say this he say that.

koolance and fanboi/s, (no offense to KATECH and their testing methodologies for car radiators that is) maybe a change to car radiator manufacturing would be better (or worse for automobile industry's sake) ...
since:
1. testing with that big airflow generator, obviously invalidates it immediately,
no sane pc enthusiast would use that high cfm/db noisemaker inside his bedroom or living room.
2. if you want independent, then sponsor or send one of your radiators to a third party enduser, who'll do the test in real pc environment, not car environment. this goes the same for hwlabs. send it to an independent end user
3. information is ammunition, if you don't want people complaining this and that, then i suggest redoing your test using real computer 120mm fans, not some monster cfm generating car radiator tester.

*halk, a cup of coffee would do you good for that hangover before posting again...

[V2-V3]

1. Test real world application relevant to a PC radiator. Air flow, Coolant Flow, Temperature. Some parameters like this:

*30C coolant temperature
*120CFM of air flow
*10LPM Coolant flow
*20C ambient air temperature

2. Make absolutely sure we are testing only one variable. This should be the radiator

3. Do we want to compare thermal dissipation per cubic unit or do we want to test overall thermal dissipation of each product? Do we want to make an overall comparison based on a 100 point scoring system that includes all performance factors such as: Air flow resistance, Coolant flow resistance, Product size, Temperature differential of coolant supply and discharge?

We need to ask our selves "what data do we need" so we may compare the results ourselves.


We also need not to forget such things as air density, humidity, atmospheric composition, as these things are mostly neglected in most reviews/tests and can have an impact on test results.

[septim]

well why didn't they do it in the first place...

instead of going through all this and feeding the less informed, you could have used some fans recently tested by Vapor over on air cooling...

[ColonelCain]

This is going in circles. It is getting really quite annoying with all of the Koolance fanboys pointlessly fighting with the people who know just about all there is to know about watercooling.
I say that those fanboys STFU, accept that members such as Cathar, Marci, and hwlabs know what the fu** they're talking about, and stop posting in this thread.
Or, I vote that the mods delete this thread.

If those fanboys really want to argue some more, they can wait for when tests are reproduced under realistic conditions (i.e. a 60-80cfm fan, and 20C or less liquid). Then, if they really want to, they can argue some more, in which case it will be even more pointless.

[1Day]

I have found the technical discussion by the more knowledgeable members who are replying and discussing the topic at hand to be very helpful. Thanks guys it is much appreciated.


Edit: Decided not to be y.

[V2-V3]

well why didn't you do it in the first place...

instead of going through all this and feeding the less informed, you could have used some fans recently tested by Vapor over on air cooling...

I have nothing to with the test was conducted or who it was conducted with.
I do not work for Koolance or any other PC related manufacturer.
I only posted the article.

Here is my opinion of the test:
Koolance consulted with an automotive R&D firm to test the thermal dissipation capacity of 3 radiators made by 3 competing manufactures.

Koolance most likely used an automotive based R&D firm because it might be the only industry that has a lab designed to test Liquid to Air Heat Exchangers and has equipment precise enough to control and measure important factors of the test and provide useful scientific data.

The down side to this is the fact that their equipment is designed for the Automotive industry, meaning such things like air flow and coolant temperature are tuned to simulate the environment which an automotive heat exchanger is used.

KATECH may not have been able to perform the test with lower coolant temperatures and air flow rates because their equipment was simply not designed to operate that low. It is also possible that Koolance did not provide sufficient environmental parameters for the test and relied on normal Automotive testing procedures.

This does not mean the data collected is not valid it simply means further testing is required with parameters to simulate the real world environment associated with that of a PC radiator.

[septim]

well then my mistake,
why ever did they not do it right the first time around?

simple reason is probably because they wanted more marketing hype for their products...
take advantage of the gullible...
take more profits for themselves...
etc etc...


*already edited above post*

[ranker]

However, after giving it some thought, I absolutely and without any doubt state that Koolance knew these results were not typical to anywhere in the same ball park, PC watercooling. There is zero chance they didn't immediately recognise that the temps and airflow were far, far in excess of what is the absolute outside of what happens.

Given that they A) Did not acknowledge it, and B) Did not offer any kind of comparison to the figures that are commonly quoted in PC watercooling...

Then I deem, and I'll say it in court, that they were deliberately out to deceive. Which in my book is on the same page as lying.

It is not possible to judge if they intended to go out and seek these results, or if they ended up with these results and intended to publish them anyway.

But stuff it. They are attempting to deceive consumers, and that may not be illegal.

Koolance, this is the bit you'll want to take me to court for libel over : Your radiators in any conceivable circumstances are inferior. Your products under perform your competitors. Official Koolance product descriptions are misleading. Finally that Koolance are lying to customers

The statement " At best, we are left with the conclusion our original claim of outperforming the other brass/copper heat exchangers by only 39-49&#37; was conservative." is a lie. It's only applicable outside of any reasonable application, something which Koolance should be well aware of. And by stating it they are saying that their products are superior to other products on the market. Since their products are inferior, then this will cause a negative impact on their competitors sales. This is libelous.

Anyway, Koolance. Sue me if you disagree.

Amen. Strange that me and halk agree. Very well said.

[Marci]

http://www.xtremesystems.org/forums/...4&postcount=42
Edited... *yawn*

Back off to bed now, the effects of Jack haven't worn off enough yet...

[Fairydust]

A lot of posts criticize the coolant temp of the test, I don't quite see why though. Higher coolant temp might not represent typical conditions for our application but how does it change the validity of the results? Afaik it only accentuates the differences thereby minimizing measuring tolerance which appears to be a good thing to me. Maybe someone can enlighten me to the error in my thinking.

Don't have much else to say about the test, if ThermoChill and HWlabs both state the pressure drop figures are completely wrong I have to question the accuracy of the results (not that results are universally valid for pc appliaction in the first place).