Depending on the AIO its design will limit far more. As Lenny says, something like the TRUE 120 scales far better than an AIP as heapipes cunduct the heat far better than watercoolers do, and if you increase the delta T they work better! So better fans would show better cooling far more, the TRUE always showed this.
I get that but “better cooling” was not the intention of the test. I fact if you use a less efficient cooling method then you should expect to see a bigger difference in water temp between the best fans vs worst fans. The idea of the test (at least in my head) is to compare the fans "cooling ability" against each other. This should be easier if you have a bigger range of possible temperatures that you can accurately measure. My thoughts were that some of the fans on test are so good, they wouldn’t have much between them all so with margins of error, would in effect be equal.
Not if the cooler starts acting like a insulator. You need something that will be as open as possible with as high a cieling as possible
My first 38mm fan: https://www.silverstonetek.com/product.php?pid=711 Seemed like the right place to post it, albeit it's for a DIY extraction system rather than a PC. On paper at least 173.5 CFM should give me a bit of oomph.
Sweet - I've always wanted to test that Silverstone fan as it's the "only" 140mm fan in 38mm flavour. Bit pricey for my liking, but I'm sure I'll end up getting it one day. Meanwhile, I made a quick vid of the GFC just for the lolz. The other fan is an absolute monster in its own right, weighing in at 405g. Maybe now it's clear why I'm scared of this fan!
@Bloody_Pete, yeah I've been following his stuff. Some of his ideas are pretty cool, but often it's "silly science" which isn't always my thing. Today I got 120mm fan goodies in the post from @Vault-Tec (thank you kindly!) and will be testing these this week. I've always been keen to see just how effective these Corsair fans are, and I guess we'll find out soon. Also, the eLoop is just a beautiful fan - it just oozes quality.
I should really dig out some of the fans I have for you! I have some OG GT's, that 1850rpm perfect ones, before they got some acustic weirdness. They're the ones everyone things of when they think GT's! I also have some I painted, so it'd be interesting to see how paint affects them!
Dope dude ! can't wait to see the results I was always told the Cougar were the best of the time, and they are very strong fans but yeah no data lol. I bought five boxes of those extensions, then realised the sodding Eloops come with it all in the box lmfao.
Haha yea, the eLoops have a lot of accessories with them. I had a bit of time this evening to do some thermal testing with the fans, and I'm pretty surprised by the results. I tested the Noiseblocker eLoop, Corsair SP120, and the Cougar Vortex. Performance-wise on the heatsink, they are all within half a degree of each other. I re-tested the Noctua A12x25 because I was using that as my baseline in the last round of testing, and each other fan beat it by over half a degree. I'm not convinced that there's any significance in that, but make of it what you will. One thing worth noting is that the Noctua is just ludicrously quiet compared to all the others. Even the eLoop, which I like a lot, had an irritating hum when running on the heatsink. The Cougar was very good. Delta temps Noiseblocker eLoop: 32.3°C Corsair SP120: 32.2°C Cougar Vortex: 32.3°C Noctual NF-A12x25: 33°C As for maximum static pressure, the Corsair SP120 stomped all over the others with a pretty mad 4.7mmH2O, but it's a 2,350rpm fan and may as well be a Delta at that speed. Both the Cougar and the eLoop were a fraction below spec. I'll do a round of tests with the sound level meter when I get some time and some peace - it's been raining all day and it's so frickin' loud, plus the seagulls won't shut up. Kids are away on holiday for the next two days so I'll have plenty time to mess around with fans!
I'm surprised I must say. Then again I'm not totally surprised. It's been known for years that Corsair's SP fans are very, very good (that's why they still use them on non RGB coolers). The others? again, were rated as the best or among the best when tested (and that Cougar is about 10 years old now !). Good to see how they really perform though ! I've never ran the NBs at anything even resembling full speed. In fact, I don't think I have ever run a fan at over 7v.
I did some sound level testing with the eLoop and the Cougar, but I made a mistake with the video so I'm going to have to re-shoot it... that's what I get for doing it late at night when I'm tired and can't concentrate! I'll be recording again this evening and will hopefully have something to share shortly thereafter.
(Slightly off-topic) Seems that every time I try to make progress with the fans, something else happens that gets in the way. This time it was a baby! But I'm still around... just sleepy lol.
Melted fan connector anyone? I found this yesterday on one of my lower-powered (0.8A) Deltas, and I have no recollection of the fan failing. It looks like there's been a short between the + and - pins which fried the housing. I have actually managed to do some testing, all of which I'll be sharing in due course.
OK... so I finally managed to get some more testing done, this time with the more up-to-date (but still old) Revision C of the Thermalright cooler. There are fewer fins spaced a little further apart than in previous models, and it runs about 1-2 degrees cooler - most likely because there's less resistance to airflow. I tested a bunch of fans at 1,250rpm using 20 rounds of Linpack Extreme, and in some cases I duplicated the tests because I was getting strange readings from the ambient thermometers. For this round of testing I wanted to see how well this crop of 25mm fans compared to a trio of 38mm server fans at the same speed. Noise is not a consideration (yet) and I'm pretty happy with the consistency of the result - the extra thickness obviously helps a lot, and the server fans are comfortably ahead of the bunch. I did run a few of the fans again at 1,000rpm and 800rpm, but more on that later when I log the sound level as well. I think we've always known that thicker fans are generally accepted to perform better as heatsink or rad fans, however I will be finding out for once and for all if they can be both better and quieter. I'm now thinking that I need to get a 120mm AIO so that I have a good range of results across different styles of coolers. Here we go... the blue columns are the 38mm thick fans.
Amazing really that they are all so close. Surprised that the CM SF is so good tbh. I never considered those !
Yeah the Sickleflow performs well on both heatsinks, although you get what you pay for (I paid £3) - a cheap, buzzy motor that will probably drive you up the wall if you have sensitive hearing. Here's a quick sound level summary video of the two best-performing server fans vs. the Noctua. Again I'm surprised that the Noctua just wipes the floor with them... even at 1,250rpm it's barely louder than the server fans at 800rpm. Performance-wise the Noctua isn't stellar, but the noise : performance characteristic is just remarkable. That's not to say that the server fans are not good...they are actually very quiet for 38mm fans, and the Foxconn in particular has a very pleasing sound profile. In saying that, both of them had really nasty vibration issues, which is why I have the cooler seated on a big block of silicone rubber. And here's a chart showing the relationship between noise and thermal performance. There's not much in it when you dial the speed down to 1,000rpm and below... most fans I've tested are at the 32 dB(A) mark, which is extremely close to the threshold of ambient noise in a near-silent room. The Cougar Vortex results aren't on the chart because they were identical to those of the Delta, which is surprising... it is not a quiet fan, registering 36.5 dB(A) at only 1,250rpm. The eLoop fared considerably better on the noise front with a respectable 31.9 dB(A) at 1,000rpm, but sadly the heatsink performance is what let it down.
