Coolgate Rads

[Church]

Hmm, never thought of using alternative inlets for different flow resistance, interesting idea. Though imho it won't matter much in big picture, as flow resistance of rads usually is small enough relative to waterblocks for that to be ignored, and one probably can freely use inlets based on however it's simpler to route tubing with.

[skinnee]

I had to admit it, but churchy is right. Using the top barb ports won't mean more or less restriction vs the front ports.

As for FPI, yes your eyes are correct. It averages 7-8FPI when you measure across the length of the tubes.

[Church]

Hmm, averages 7-8? And IIRC RX rev2 had a bit lessened FPI, so looks like about the same.
That's also one of ways to follow for newcomers @market. Copy best but sell cheaper.

[rheydt]

I am quite curious, if 90% copper tubes will bring more performance, can´t wait for skinnee´s review any more!

[Ogge_swe]

toothpaste anyone?

[hwlabs]

The difference in terms of thermal resistance between thin wall brass tubes, and copper, when clad with lead-free solder is nowhere near as significant in performance as air and liquid side pressure drop.

Sure you can reduce the fin pitch to a point you may compromise the core structure but if you look at the performance numbers, the Thermochill PA pretty much equals it.



For reference from the same reviewer:

[Church]

Oh, another one LC vendor rep on this forum? Nice to see.
hwlab: IIRC PA was discontinued though, and replaced with TA, with worse cooling performance though.
Speaking about all copper - i'd love to find out why magicool plexi elegants performed that much better then the rest at both low and high fan rpm range. Was it because of non painted fins or because of copper? If it's because of not painted - maybe it's worthy idea to use by other rad manufacturers too?

[Waterlogged]

Oh, another one LC vendor rep on this forum? Nice to see.
hwlab: IIRC PA was discontinued though, and replaced with TA, with worse cooling performance though.
Speaking about all copper - i'd love to find out why magicool plexi elegants performed that much better then the rest at both low and high fan rpm range. Was it because of non painted fins or because of copper? If it's because of not painted - maybe it's worthy idea to use by other rad manufacturers too?

I believe that's the site, not the vendor.

[cdawall]

these look like they would be pretty easy to stack like the stackable MCR stuff from swiftech maybe these will end up instead of a bunch of swiftys in my upcoming build

[Church]

Stacking in most cases is bad. Why bother?

[Waterlogged]

Stacking in most cases is bad. Why bother?

That testing was done on rads that have a much higher FPI than these rads. I'd really like to see this testing revisited someday with a bit more thoroughness.

[Church]

Waterlogged: those higher FPI rads are bit of like RS/BIP/MCRs, they partially compensate middle FPI with lower thickness. Skinnee also didn't get much of improvement from stacked sandwich of MCRs over single rad. And in both of these tests they note that higher speed fans are needed for their higher pressure to compensate increased airflow resistance. - Now what does that reminds me of? - exactly! High FPI rads. Same worse performance with quiet fans and extra cooling capacity at higher rpms (just like Naekuh's suggested sandwitch with nutcase san aces in between). So what advantage might bring using two lower FPI rads to get something similar optimisation/performance wise as one high FPI rad like GTX? Takes more space, price goes up, result same. And what both rad sandwitch and high fpi rads have common - bad choice for quiet setups.

I have to admit, just to make sure, extra testing never hurts. Skinnee, you don't happen to have two RX or SR1 arround?

[SpuTnicK]

I believe that's the site, not the vendor.

no, its vendor. Member registration date (year) does not match the date of start of corresponding site.

Stacking in most cases is bad

I would also ad another rad config to compare: fan-rad1-fan-rad2-fan. Overall 9 fans over2 rads, and see how well it would perform.

And what both rad sandwitch and high fpi rads have common - bad choice for quiet setups.

well the time those test were done, no one knew that GT fans will appear to be the best fans performance/noise wise.
I suggest, the test should be repeated with different fans (like GTs: low, med and high speed) and with different FPI rads (preferably low ones)

[kirakisyou_2000]

The difference in terms of thermal resistance between thin wall brass tubes, and copper, when clad with lead-free solder is nowhere near as significant in performance as air and liquid side pressure drop.

Sure you can reduce the fin pitch to a point you may compromise the core structure but if you look at the performance numbers, the Thermochill PA pretty much equals it.



For reference from the same reviewer:

Hello

The people who write this review.
Thank you for watching my test.

In my test results, the fan speed - speed of the fan performance was slightly better than a PA120.3.

This radiator was like wow.

http://translate.google.de/translate...tml%23readmore

[bundymania]

Thanks for posting this here !

[cdawall]

Stacking in most cases is bad. Why bother?

i am also stacking with more fans than he used in any test...not to mention higher speed ones

[subtec]

Something that occurred to me today that I never see mentioned is the effect of the inlet/outlet placement. It seems to me with the inlets placed out at the corners of the radiator the water has an equidistant path regardless of which way it goes through the radiator.

Conversely, with the inlets closer to the middle, the paths vary in length, with the water at the edges traveling a longer distance. The longer path will naturally be more restrictive, meaning an overall lower flow rate towards the outside edges of the radiator.

I'll illustrate the point:


It's also worth considering the effect of fan deadzones, which will be directly over the center of the radiator - the area with the highest flow rate in an unequal-water-path radiator. In other words, the greatest thermal transfer between the water and radiator fins is in the middle, while the greatest thermal transfer between the fins and the air is toward the edges. Moving the inlets to the corners serves to improve the overall thermal transfer.

Come to think of it, considering the airflow distribution and deadzones of typical fans, it may be beneficial to design some restriction into the channels closer to the center (or alternatively, make the outer channels less restrictive) in order to take greater advantage of the increased fin-air thermal transfer at those locations.

Anyway, if these Coolgate radiators perform better than average, the corner inlets might be one reason why. It could also be one reason the Swiftechs perform relatively well for their thickness.

[skinnee]

I have to admit, just to make sure, extra testing never hurts. Skinnee, you don't happen to have two RX or SR1 arround?

You guys are



Well, current radiator testing (rads already on hand) is planned out through the 3rd week in March. There is some slack in there, but not much... each radiator goes through 96 hours of thermal testing.

After that I was planning on doing the the fan orientation testing and then 240 and 480's or the Mora/Nova testing. I can bump the 240's / 1080's and get stacked testing in if you would rather, but that means the 240's / 1080's will get bumped quite a bit. Stack testing is going to take about two-three weeks to do it properly, and I would need to pick up another SR1.

Your choice, just don't complain after you've made your choice.

[Waterlogged]

You guys are



Well, current radiator testing (rads already on hand) is planned out through the 3rd week in March. There is some slack in there, but not much... each radiator goes through 96 hours of thermal testing.

After that I was planning on doing the the fan orientation testing and then 240 and 480's or the Mora/Nova testing. I can bump the 240's / 1080's and get stacked testing in if you would rather, but that means the 240's / 1080's will get bumped quite a bit. Stack testing is going to take about two-three weeks to do it properly, and I would need to pick up another SR1.

Your choice, just don't complain after you've made your choice.

I just want it on the record that I never mentioned your name. I was kind of hoping Martin would be intrigued enough by this that he would take up the challenge,

I'd rather see the other testing done first, we've been needing solid testing of that stuff for a while. This can wait a bit longer IMO.

[skinnee]

No worries, it is something I need to do, just wanted to let you know timelines I am working with. Martin is on a noise endeavor... which I am happy to see him take on.

[Waterlogged]

No worries, it is something I need to do, just wanted to let you know timelines I am working with. Martin is on a noise endeavor... which I am happy to see him take on.

It's all cool.

Did you watch those links I sent you ?

[skinnee]

a few seconds of them, about all I cared to.

[rheydt]

can u share these funny links with us, waterlogged?

[Waterlogged]

can u share these funny links with us, waterlogged?

I knew I shouldn't have said that in public.

Sorry but no, I can't.

[BlueAqua]

can u share these funny links with us, waterlogged?

http://www.youtube.com/watch?v=aA3dHi_o7Yw