Overclock.net banner

21 - 25 of 25 Posts

·
Shooting down fallacies
Joined
·
8,074 Posts
From the pics and blue prints, it sounds like an inpeller simular to Asetek or the other brands. The pump has only 2000rpm as well.
I'd like to see some actual specs/tests on the Arctic, purely out of curiosity. It isn't doing anything magic with the rad or coldplate, but it's making the Aseteks look like toys in tests. That leaves the pump, which we all know is Asetek's biggest weakness (not that any part of them isn't a weakness, mind you). Cooler Master used essentially identical rads and coldplates to the Asetek on their in-house Neptons and got better performance with higher flow pump (maybe you should try to find a leftover or lightly used Nepton?). You also need to keep in mind that speed in different pumps means little in terms of flow, so dismissing it due to it's top rpm may not be prudent.
 

·
Registered
Joined
·
323 Posts
I'd like to see some actual specs/tests on the Arctic, purely out of curiosity. It isn't doing anything magic with the rad or coldplate, but it's making the Aseteks look like toys in tests. That leaves the pump, which we all know is Asetek's biggest weakness (not that any part of them isn't a weakness, mind you). Cooler Master used essentially identical rads and coldplates to the Asetek on their in-house Neptons and got better performance with higher flow pump (maybe you should try to find a leftover or lightly used Nepton?). You also need to keep in mind that speed in different pumps means little in terms of flow, so dismissing it due to it's top rpm may not be prudent.
I watched a video from Gamer Nexus and the pump is similar to the CoolIT and the one on the Eisbaer LT, the latter being bigger. From the rpm, I don't see it standing out from the pack. I do think the CooliT block is clearly superior to the Asetek one, so perhaps the Arctic is a bit better, but it will still be rated as a 50/70 L/h, since it is still smaller than the Eisbaer LT.

Believe me, I have been testing several AIO water coolers over the years, and they all perform more or less the same if you use the same fan and the same thermal paste. The only one that stood out from the pack was the H45 from Corsair (forgot the name of the OEM), which runs the pump (very loud) at 4200rpm and performs like a 240mm rad, although being a 120mm rad. The only problem is that the pump is next to the radiator, not on the block, so I can't use it.

Then I think the Arctic result could be a mix of the coldplate, a better impeller than the Asetek one, but it is still a regular AIO pump. Perhaps the radiator is different. Perhaps they made the tests with the stock thermal paste (it can vary from brand to brand).

So from what I have seen, for an AIO unit to be clearly superior, either the coldpate / pump is groundbreaking or the flow is faster. Mounting pressure and the flatness (or not) of the coldplate is also important. Having said that, think the Enermax and the Heler blocks have more potential. And the VRM fan on the Arctic is a bit of an (ugly) gimmick.

Anyways, 2ºC difference (Gamer Nexus tests) could be due to test error, especially if they are using different fans and thermal paste. I could be wrong, but the Arctic unit might tie with any of the CollIT units from Corsair - and I'm not sure if CooliT fixed the backplate gap thing with the new models.
 

·
Registered
Joined
·
380 Posts
Discussion Starter #23
Hi Contiusa and guys, sorry for my late replies, I was asked to write a paper that requires researching sources and combing through extremely large amounts of bla-bla (I just can't call those sources) takes an overwhelming amount of time, one of the videos is 6 hours long. Arrrrgh. I'll have a look at my Helor, but it is installed right now in my main rig on a 10600k. I might take it out after all for an inspection, as I actually pulled the RAM out of my rig - I got an Alphacool Eisbaer 280 with a rusty radiator (!!!) a day ago and managed to test it on a 9700k system, during the night, cutting short my research (but I didn;t have any spare RAM, so the Helor is not working anyway). I am writing from this 9700k bench right now and it is 01.42hrs at night in NY. I'll regret it tomorrow. Anyway, my testing with that Asetek and Corsair H105 (38mm radiator, just like the Arctic CLC) before modding them did not result in what I expected, I'll have to update you guys on specifics as soon as I can. Sorry for the delay and see you later. Be safe!
 
  • Rep+
Reactions: Contiusa

·
Registered
Joined
·
380 Posts
Discussion Starter #24
A quick update #1

I have tested the Cougar Helor 240, Corsair H105 (240mm) and Asetek 591LX this week, when I had time for breaks. Here's what I noticed:

On the two 240mm AIOs the testing was done with Gentle Typhoon 2150rpm PWM fans - 9700k @5GHZ -2AVX, 1.3vcore, Asus z390-a (with a 120mm Fractal Design 1200rpm fan over the VRM - the VRMs on this mobo are not cooled properly, so it turns out that Buidlzoid did that too on this exact mobo - the traditional fan over VRM, just as the Arcitc did with their AIO coolers). Anyway:
  • Cougar Helor 2800rpm pump, 27mm rad
  • Corsair H105 1850rpm pump, 38mm rad
I cannot find the screenshots of the testing with GT fans, but iirc Corsair outperformed the Helor by 2-3c. I do not have the means to measure the ambient exactly, so probably the real difference was closer to 1-2c maybe. This seems to indicate that the thicker 38mm radiator on the 240mm Corsair H105 might be the reason for a marginally better result over the faster 2800rpm pump (but) with the 27mm radiator on the Cougar Helor 240, despite the sluggish 1850rpm pump on the Corsair H105. Thicker rad with slower pump won, marginally.

Remember that btw the Arctic cooler has a 38mm rad and a VRM fan, which adds to the effect (GN was the latest to confirm, indeed).

I also did the testing on that Asetek 591LX 360mm (with three 3-pin Raidmax fans that had no speed or other markings and two of those were wobbling and squealing of old age) - I was hoping to use this AIO's pump for my mod. But the pump is... 1500rpm!!! Ands the cooler seems to have about 50% of the liquid, after I did a shake and listen "measurement" test. This time, I found a screenshot for the comparison between the Asetek and Corsair, but with their native fans. I again used my 9700k @5GHZ -2AVX, 1.3vcore (z390-A with the 120mm fan at 1200rpm over VRMs) under Prime95 SmallFFTs with AVX instructions ON for one hour on each cooler (after a restart and about 10-15min cooldown of the system after the restart). So, this Asetek, with those 3 x 120mm old and I seemingly failing 3-pin 1700rpm fans reached peak 87c, while the Corsair H105 240mm CLC, with all its 38mm thick radiator and super loud fans at 2600rpm, reached peak 85c. So, a half-filled 360mm cooler with failing fans was only 2c warmer than a 2years old 38mm rad, 240mm CLC, with a sluggish pump and fast/screaming fans. I felt that if I refill the Asetek and use better fans, things may change to the opposite - it would win over Corsair.

Despite these results, presumably showing that a larger dissipation surface - @ciarlatano ;) - wins over pump speed, I am still uncomfortable with using the 1500rpm pump from the Asetek on a copper thick 240mm rad (EKWB 240PE with a 40mm thickness or XSPC RX240 with a whooping 56mm thickness and I didn't even determine if it fits in either mine or my elder's son case).

Also, as I said, I got an Alphacool Eisbaer 280, tested it and it reached peak 88c with those sluggish 140mm Be Quiet airflow 1300rpm fans and the 2600rpm pump. Unfortunately, the radiator fins on the Eisbaer are super rusty. No, it's not copper coming out and it's not green oxidation, it's just good-ol rust. I couldn't believe it, but it's real. I wonder what the fins are made of... My elder son said he wants it with a new radiator and clear tubing, so I just might disassemble the Eisbaer and use its pump and fittings with one of the 240mm (EK or XSPC) thick radiator that I've mentioned, instead of my original project. Well, unless, some cheap and okish pump comes my way:) The alternative is to buy an Alphacool pump, I know a place where they still have a few left, despite the pandemic.

I am undetermined if I should scratch my "tinkeritus" (as someone here on OCN put it) and still buy a super cheap, but this time good and fast pump, for the sake of this experiment or go the sure Alphacool pump route.

@Contiusa - about the Cougar, I'll try to reply next days hopefully. It's fully functional and I might give it to my younger son or keep it for reserve, but I will definitely not disassemble the Cougar yet. By the way, I think Reeven Naia 240, Cougar Helor, as well as a Thermalright Turbo Right 240 AIO share the same pump, with Naia being slightly modified:



Ok. Its 03.16hrs in the night here. I need to catch some sleep. Talk to you, hopefully, soon.
 

·
Shooting down fallacies
Joined
·
8,074 Posts
Despite these results, presumably showing that a larger dissipation surface - @ciarlatano ;) - wins over pump speed, I am still uncomfortable with using the 1500rpm pump from the Asetek on a copper thick 240mm rad (EKWB 240PE with a 40mm thickness or XSPC RX240 with a whooping 56mm thickness and I didn't even determine if it fits in either mine or my elder's son case).
With adequate flow, larger surface area will always show greater gains than increased flow on the smaller rad. I.E., taking two identical systems that both have 0.75 gpm flow rates with identical 240mm rads, changing to a similar 360mm rad (or to a much lesser extent a thicker 240mm rad) on one would should a greater performance increase than upping the flow to 1.25 gpm on the 240mm rad. However.....we are talking CLC pumps here....if the flow rate on both was 0.25 gpm, you would see a distinct improvement upping the flow rate to 0.75 gpm. Both aspects are obviously huge factors, my comments on flow rate were assuming the same components in the loop. But, all of that is just stating the obvious.

Also my comment about rpm not being a sole barometer of dictating flow seems to have been greatly misunderstood. Different pumps have distinctly different capabilities. In a small loop a D5 at 1500 rpm is going to flow a lot more water than an old Asetek at 3000 rpm. While that is an extreme example, it points out why I said that using rpm as a sole means of judging a pump's capabilities is shortsighted. Perhaps a more applicable example is the Asetek pump pushing a whopping ~50 lph (or be crazy and say 100 lph since we don't have the exact) at 3000+ rpm, while the Enermax is 450 lph at 3000 rpm. And that is why I'm really curious to see some measurements on the Arctic pump.
 
21 - 25 of 25 Posts
Top