120mm Fan Testing on an MCR120 Radiator Round 6

[Philwong]

Thanks for testing out the AC F12, Martin.

I also find these budget fans excellent for value and performance. What really seals the deal for me is the PST function or PWM daisy chaining. This allows me to tap the CPU and GPU fan headers to throttle my radiator fans to achieve optimum noise vs. cooling balance.

I've four F12s wired up to a single CPU fan header using only the PWM and RPM rails. Power is pulled directly from the PSU via a simple mod. Works fine for my 5870 GPU header as well.

Some folks have highlighted that the CPU/GPU temp's fluctuation may be higher than the coolant, but this approach still works relatively well for the F12 without audible noise issue from my seating position.

I was previously using YL SL and some of them tend to tick at lower RPMs even without PWM control.

The thin cables are a flimsy, but they can still hold up pretty well electrically.

Phil

[Manicdan]

Perhaps because this fan has sat on the shelf for a year plus,

sorry but..

hope the next GT15 are looking just as good as test 2 results,
im really loving mine sofar

[Church]

Martin: Oh, so maybe Vapor was responsible for recent world-wide AP15 shortage?

[somethingclever]

But yeah, I am a bit short on 140mm fans, so I'll be looking to round up anything in a 140 pretty soon. I'll probably be ordering some as well, but I have a few in shipping right now highlighted in orange on the fan list.

This is fantastic, I can't wait!

[Martinm210]

sorry but..

hope the next GT15 are looking just as good as test 2 results,
im really loving mine sofar

Sorry but what?

Is there an expiration data printed on your fans somewhere? Is there documentation somewhere I can read that confirms fans can only be stored for 1month, 6months? How long are they stored at the vendors before you buy them? Oh and yes I'm stubborn and enjoy learning the hard way sometimes...

I do know that my Coolermaster R4 fan has a date printed on it of 9-13-2009. I would assume that is manufactured date, so that fan was over a year old when ordered brand new.... Should I RMA that fan for being over a year old too..?

Actually, we don't know much at all about date or ages or what if anything goes wrong with a fan in storage or even how long they are in storage before we get them..... Also this fan could have been defective from day 1...or it could be that all GTs behave randomly like this...it's all speculation. All I know is the sample I currently have performs randomly and has been since I've been working on the latter rounds of testing where I've actually tried to retest many times and have been seeing bigger variences. In round 2 my test bench was a bit lacking to pick up the find details. We really don't know anything more than this sample is currently behaving inconsistently. I will know more soon enough though as I get my hands on some other samples to check against.

I suspect there is actually quite a bit of sample variance out there. I've seen it with pressure drop testing of blocks, so I'm sure it's there with fans as well. Always a bit of a gamble, and most people would never know they got a less than ideal fan unless you tested it in detail and had another sample to compare against which nobody does...

Me too, the GTs are obviously really good, I just don't think they are perfect. I suspect they will be the fan to beat in the 1000-2000RPM range, but there are other factors to consider. For those looking to run under 1000RPM there may be better options and certainly lower cost options. But who knows....I'd like to see some good competition myself. I'm seeing stores bumping prices because supply isn't keeping up...and I don't like that.

Martin: Oh, so maybe Vapor was responsible for recent world-wide AP15 shortage?

No kidding, I'm just happy he's willing to share in all this AP wealth...

This is fantastic, I can't wait!

I am looking forward to doing some 140 testing. I've been speculating that it's a direction we'll be going toward in general, but who knows. Maybe 120 is the sweet spot? We do know that you get more CFM per RPM, but we don't know too much about noise of 140s on a radiator.


Anyhow, I have two more recordings done that I'll process in the next couple of days, then I'll plan on batch #3 this coming weekend.

[Martinm210]

Cooler Master R4-C2R

Thanks to vhaarr for the sample on this one..

NOISE LEVEL
Was about the same as the medium yate at the low end, but a fair amount better at the top end. I suspect this is due to the greater number of blades and forward sweep which seems to really cut down on the air type noises created by the blades.

NOISE QUALITY
I'd say average to less than average primarily due to the motor tick which makes the overall sound a bit rough and grainy or mechanical rather than a smooth air flow. Also seemed to fight a little bit of resonance as indicated by the irregular noise curve. It's too bad this fan does not come with the motor on the excalibur, it would probably make a great combo since I really didn't hear any motor noises coming from that one.

EFFICIENCY
Very similar to the yate medium, if anything a little better at the high end

CFM per RPM
Surprisingly was a little worse than the yate medium. Perhaps the blades are not deep or dished enough to get that full scoop type action, but regardless it is what it is.

RANGE
Overall a pretty very good power range from 960RPM at 4V to 2055RPM at 12V, that's good.

OVERALL
While noise level is fairly good and actually pretty impressive at the high end, I think it suffers a bit on the noise quality front which seems to come from the motor. It's an good value fan with lower noise level, but it's not the smoothest sound compared to other options.

VIDEO LINK
http://www.youtube.com/watch?v=A2Z2izxVYDY

[Martinm210]

Thanks for testing out the AC F12, Martin.

I also find these budget fans excellent for value and performance. What really seals the deal for me is the PST function or PWM daisy chaining. This allows me to tap the CPU and GPU fan headers to throttle my radiator fans to achieve optimum noise vs. cooling balance.

I've four F12s wired up to a single CPU fan header using only the PWM and RPM rails. Power is pulled directly from the PSU via a simple mod. Works fine for my 5870 GPU header as well.

Some folks have highlighted that the CPU/GPU temp's fluctuation may be higher than the coolant, but this approach still works relatively well for the F12 without audible noise issue from my seating position.

I was previously using YL SL and some of them tend to tick at lower RPMs even without PWM control.

The thin cables are a flimsy, but they can still hold up pretty well electrically.

Phil

I do like the F12. I reminds me of the yate in terms of value but with a step up in quality/features without getting outside of the $10 a fan budget..

[Martinm210]

NOISE BLOCKER PL-2

This fan was also sponsored by vhaarr

PACKAGE
This was very premium in terms of packaging, all sort of goodies included. Custom rubber gasket system with studs and thumb nuts. Also comes with two sleeved wire lengths. It really went all out in accessories.

NOISE LEVEL
Fairly similar to the yate medium, perhaps a hair lower at the bottom end and a bit higher at the upper end, but the differences were less than 1dBA which is probably within testing error and within the 3dbA (barely perceptible) range. This is good

NOISE QUALITY
A fair amount better than average. No abnormal resonance peaks, no motor ticks, just a very smooth wind sound type noise. I think it's better than the yate in it appears they use a flatter blade type to soften the noise chop type sound.

EFFICIENCY
Not as good as the yate medium

CFM per RPM
Also not as good as the yate medium, flatter/straighter blade type may be more tuned for case type fan duties.

OVERALL
Cost is a real bummer on these, but they do have very good noise qualities and a load of packaging goodies if that's important to you.

About average noise level, better than average noise quality, premium packaging, and high price.

VIDEO LINK
http://www.youtube.com/watch?v=1rsMKjBZHo8

[Church]

Hmm, few thoughts regarding result representation.
Watt and Rpm vs CFM can be seen from test results spreadsheet. Am i wrong to claim that only graph chart that really matters in fan choice is CFM vs Noise with other two being redundant?
After all, if some fan is quieter at 1400 rpm then other at 900 while pushing same amount of air, then does that rpm number mater at all? In such case it's just artificial tech specs number not really usuable when choosing/comparing fans.
Same about power consumed. Only things from wattage that should matter to user - maximum power consumed when working, maximum startup power draw and minimum startup voltage, to know how much fans can be fed from single channel of motherboard fan header/rheostat/fan&pump controller, and down to which voltage fans can be safely down-volted. These maximum numbers can be just looked up in test data sheet and used when summing up daisy chained fan total power draw, and rest - easily discarded. Not worth separate graph chart as well imho.

[Martinm210]

Hmm, few thoughts regarding result representation.
Watt and Rpm vs CFM can be seen from test results spreadsheet. Am i wrong to claim that only graph chart that really matters in fan choice is CFM vs Noise with other two being redundant?
After all, if some fan is quieter at 1400 rpm then other at 900 while pushing same amount of air, then does that rpm number mater at all? In such case it's just artificial tech specs number not really usuable when choosing/comparing fans.
Same about power consumed. Only things from wattage that should matter to user - maximum power consumed when working, maximum startup power draw and minimum startup voltage, to know how much fans can be fed from single channel of motherboard fan header/rheostat/fan&pump controller, and down to which voltage fans can be safely down-volted. These maximum numbers can be just looked up in test data sheet and used when summing up daisy chained fan total power draw, and rest - easily discarded. Not worth separate graph chart as well imho.

I agree, I'm just looking for any trends in those latter two, not much there really other than debunking many myths I had (38mm fans always producing better performance per RPM,etc.), but it might come in handy for comparing radiator performance levels. If a radiator is tested at one RPM, you could convert using the RPM/CFM chart.

I was also curious to see if power consumption is related things like noise level and or bearing quality/friction? In the end, I'm not sure anything correlates there.

I would still argue that noise level, even with this much more appropriate A weighted version is only half of the picture. Unfortunately there really is no way to quantify noise quality other than try to include some notes in a narrative and provide the audio for users to listen to.

That's just my thoughts though, still remains fairly subjective regardless of the tools/methods used.

The ultimate to me is a combination of noise level per CFM and noise quality. The noise level is easy enough with the meter, but noise quality remains in the audio track to listen to..

[Martinm210]

That's it for batch #2, the fans highlighted in dark green are all I have on hand that haven't been tested so far.

My bigger PPCs order won't get here until next Friday, although I might get a couple more in from SVC before the weekend I can include in batch #3. My SVC orders usually get here in only 3-4 days, where my PPCs order it going to take 9 LONG days. That's alright, I should probably have 10 or so fans to test this weekend, I was just hoping for more since I've got some extra days off for the holidays to play. I just might have to got cut some firewood or something productive instead..

[MagisD]

Great work martin!

I was convinced it was a wonky gt-15.

Sad news is I got my batch of them in and there different. Its subtle about the fan itself looks like they started cutting corners. Allso No sleeveing, no heatshrink, stickers seem cheaper. Overall they seem like a knockoff of the orginal.

It could be the fc-2 I'm useing but seems to have a different harmonic.

Gona test new vs old on same setup see if i can tell the difference.

Sent from my X10a using Tapatalk

[Martinm210]

Great work martin!

I was convinced it was a wonky gt-15.

Sad news is I got my batch of them in and there different. Its subtle about the fan itself looks like they started cutting corners. Allso No sleeveing, no heatshrink, stickers seem cheaper. Overall they seem like a knockoff of the orginal.

It could be the fc-2 I'm useing but seems to have a different harmonic.

Gona test new vs old on same setup see if i can tell the difference.

Sent from my X10a using Tapatalk

I should have listened to you earlier, but my stubborn self couldn't resist..

Sorry to hear about the changes, I've got some coming too.

[Martinm210]

MASTER NOISE vs CFM CHART
Updated 11-8-10

I'm only going to update this after getting each batch done. What is obvious to me is how extremely tight the results are. Per my reading, it normally takes about 3dbA for most people to perceive a change in noise level...that's about the spread of "Most" of the fans here. But if we're splitting hairs...here is the chart for your viewing pleasure.

Just note this is NOISE LEVEL only. I think this is only half the picture, noise quality is what you get by listening which points out the things like motor tics, and other less that smooth sounds.

[antiacid]

MASTER NOISE vs CFM CHART
Updated 11-8-10

I'm only going to update this after getting each batch done. What is obvious to me is how extremely tight the results are. Per my reading, it normally takes about 3dbA for most people to perceive a change in noise level...that's about the spread of "Most" of the fans here. But if we're splitting hairs...here is the chart for your viewing pleasure.

Just note this is NOISE LEVEL only. I think this is only half the picture, noise quality is what you get by listening which points out the things like motor tics, and other less that smooth sounds.

Cmaster r4c2r is a winner in that first+second batch, surprising.

I sort of didn't read through that whole discussion about the gt15. Is anybody going to sponsor a 2nd one for you?

[iandh]

I should have listened to you earlier, but my stubborn self couldn't resist..

Sorry to hear about the changes, I've got some coming too.

I'm really, really hoping this is just a bad batch, and not an improper response to increased demand.

I know there's been a shortage for some time on the GT's, I pray that's not due to retooling for higher production at a reduced quality level.

The superb quality of the GT's is what has made them into the ambassador of fans, I hope that Nidec servo isn't letting profit and demand prevent them from seeing that.

I'm wondering if there was a shortage on the high quality bearings that they were using, but I can't imagine how that could be the case being that they are a standard of the server market.

[MagisD]

I'm putting my build on hold till i have time to strip out my old fans out of my old build and play with them side by side.

But I'm very sensitive to harmonics certain fequency/tones give me mirgraines. The new gt's seem dirtier (Sound wise )to me if that makes any sense?

Sent from my X10a using Tapatalk

[Martinm210]

FYI,
Got my SVC order in today, they are so fast... Thanks to Ratdog and twich12 for these two..

Batch 2 GTs have no more wire sleeving as others have indicated, a change in rear sticker, and what appears to be a change to the support frame...possibly for better circuitry cooling?

I'm curious if that opening up will add noise though..?

[Waterlogged]

Martin, I have 4 brand new in the box batch 1 AP14's and can say without a doubt that sleeving is not normal. I know for a fact that P-PC's sleeves fans without the customer asking (you also have little choice in the matter), two of my AP14's came from there and are sleeved (and put back in the package backwards), and the other two came from FCPU a few weeks later and those are not sleeved. I also have some AP13's from FCPU that are going in my current mini-ITX HP build and those came unsleeved as well. FCPU will sleeve it, if you pay extra for it.

Weird thing I noticed is, I have AP15 and AP14 that have the same sticker as your AP14 but the housing of your AP15 and I know both of these are batch 1 fans.

The blades on that Cooler Master fan look a bit like the blades on the Akasa Apache and Viper fans, both of which I just got. Apache Black (maybe a AP13 or AP14 competitor) is a little bit weaker but fairly quiet and the Viper (seems to be in AP15 territory) is a good bit stronger (than the Apache) but also a good bit louder with kind of a higher pitch whine to it. . .I'll be sticking with my GT's. The Viper just seems to be a souped up Apache with different colored blades as noise and flow (all subjective test methods used) seem to match up RPM wise.

[Martinm210]

Martin, I have 4 brand new in the box batch 1 AP14's and can say without a doubt that sleeving is not normal. I know for a fact that P-PC's sleeves fans without the customer asking (you also have little choice in the matter), two of my AP14's came from there and are sleeved (and put back in the package backwards), and the other two came from FCPU a few weeks later and those are not sleeved. I also have some AP13's from FCPU that are going in my current mini-ITX HP build and those came unsleeved as well. FCPU will sleeve it, if you pay extra for it.

Weird thing I noticed is, I have AP15 and AP14 that have the same sticker as your AP14 but the housing of your AP15 and I know both of these are batch 1 fans.

The blades on that Cooler Master fan look a bit like the blades on the Akasa Apache and Viper fans, both of which I just got. Apache Black (maybe a AP13 or AP14 competitor) is a little bit weaker but fairly quiet and the Viper (seems to be in AP15 territory) is a good bit stronger (than the Apache) but also a good bit louder with kind of a higher pitch whine to it. . .I'll be sticking with my GT's. The Viper just seems to be a souped up Apache with different colored blades as noise and flow (all subjective test methods used) seem to match up RPM wise.

Interesting, I thought it was factory... I believe my 15 came from PPCs, it's been a while now, but Honda left me a big old box of fans last year and I'm pretty sure it was PCCs.

[Martinm210]

Nidec Servo (Scythe) Gentle Typhoon AP-14

Special thanks to RatDog & Cisco Systems for giving me a general fund to pick from, I've had the old batch one AP-15 for a while and wanted to try a new batch AP-14. The AP-15 was driving me nuts with it's inconsistency, so I was really curious to get my hands on a brand new sample and also have a chance to look at the lower RPM variety.

Vapor is sending me a couple of batch two AP-15s I'll give a run on too when they come. Then I'll have an opportunity to get some good runs on some new fans from batch 2.

This fan is brand new however, so I feel more confident it's at least not suffering from shelf life.

OK, on with the results.

BATCH 1 vs BATCH 2 FANS
These fans have changes slightly, the good news is, I don't think noise was affected by the changes from what I can tell.


PACKAGE
Fairly average. Comes with the fan without any sleeving on the wires and it comes with some case screws and a molex connector in a plastic box. Overall, about average.

NOISE LEVEL
At the high end this fan spins around 1450RPM and is a full 8 dBA lower in noise level on my particular test rig than the yate medium, that is significant. At 7V or roughly 850 RPM is is about 3 dbA lower in noise level than the yate medium. 3dBA is considered barely perceptible, so that's not so significant, but still worthy of note. Nothing short of exceptional here I mean REALLY OUTSTANDING noise level.

NOISE QUALITY
This is a difficult one to evaluate. On one hand, the exceptional noise level is amazing, but the GT has a way with sound. The normal noise of air movement from the fan blades is nearly absent. What you are left with is very small amounts of bearing/motor noises and those small noises are not quite as smooth as the sound of air movement. Also while I didn't pick up any measurable noise spikes from harmonic issues, I could very faint amounts of tone developing over the voltage range from time to time.

In the end, I guess I've generally rated noise quality as exceptional, if it was purely a smooth air like sound where motor noise was absent. In this fan, there is no air noise, all you hear is the faint motor/bearing working. That's not quite as pleasing as smooth air. However, the fan is very quiet and there is not any motor tick that's generally the most annoying type of sound. I'll give noise quality an average rating, although I think that's totally subjective.

ELECTRICAL EFFICIENCY
How is .5 watts, the GT is absolutely the most electrical efficient fan I've seen. Even the GT15 keeps power under 1 watt. It's about twice as efficient as the yate loon. That'll keep the stress off of overloading your fan controller. Another exceptional by the GT.

CFM per RPM
About average in this test. The more I test and compare this chart, the more I think I'm just seeing variability in RPM sensor accuracy. I don't think I'll worry too much more about this comparison.

COST
I've seen this fan sell for about $15 to $19 a piece, so it is not a bargain fan and not going to be cheap to fill a case full of a dozen of these. I think it's probably justified if you need the full RPM power, but perhaps not if you're looking to undervolt down in the sub 1000RPM range.

VISUAL
This is pretty subjective, but I've seen many nicer looking fans. I suppose I appreciate the fact that they focused on performance only on this fan, but it would be nice to have a few color choices. It's unfortunate when you see people painting the fan blades, as that's a pretty good sign folks want something other than grey and black. I also wouldn't ever recommend painting, could very easily throw off blade balance if you're not careful. I'd give visual an average.

OVERALL
Jaw dropping noise/cfm ratios from 1000-1450RPM, it flat dominates in this area. Average noise quality, exceptional electrical efficiency, average to high cost, and average visual.

I'll give it three thumbs for it's dominating cfm/noise ratio on a radiator, it's really exceptionally quiet at medium speed RPMs.

VIDEO LINK
http://www.youtube.com/watch?v=Yuj7GZqNpOs







And for giggles, I compared it to my wonky GT15, while my Batch one (lower result) didn't check out well, it did compare very well to the Batch 2 GT15 test..



And this was my retest of this fan in batch 4 compared to my results in batch 3. About 1CFM difference, but very similar.

[Manicdan]

keep up the great work man

btw, why would someone buy a GT14 or less? it seems identical to the 15 except the range is limited. is it really just so people who have little control over speed can pick the speed they wanted?

[Martinm210]

keep up the great work man

btw, why would someone buy a GT14 or less? it seems identical to the 15 except the range is limited. is it really just so people who have little control over speed can pick the speed they wanted?

Thanks!
Good question, it's pretty hard to tell with the 14 vs 15, but at least with other fans there is sometimes less motor noise/tick if you can keep the fan fairly close to 12V. Some other fan families like the Ultra Kaze is clearly that way. The UK3 has a really strong motor tick at lower volts, but the UK2 is a fair amount smoother at the same performance level.

[nex_73]

Thank you for all the hard work/testing! Really appreciate it!
Seems like i have to switch my 2 excaliburs =)
Looking forward to the Scythe SS SL12H result as i have 2 SL12SH and they seem a little bit better at the same dBA regarding to:
SPCR

[Martinm210]

FYI,
I got about a dozen or so fans tested or retested last night. It's a bit cooler 18C as opposed to 20-21C on this batch and this is how the batch checked in using the yate medium as my check and baseline fan. It seems noise level vs air flow remains relatively good, however the temperature does affect the air flow per RPM a bit more than I expected, I suppose because the air is slightly more dense?)

Since we're really just looking for noise vs cfm, I suppose this is still ok, but something to be aware of. There will be some error in the max CFM at 12V depending on temperature. (32.5CFM vs 30.5CFM or about a 2CFM difference in max at 12V.)

On the noise vs air flow chart, it was generally fairly close within about .5dbA, with a couple of bumps that were about 1dbA different.

I'll accept that and just acknowledge repeatability on the noise vs CFM is typically better than 1dBA, and CFM per RPM is repeatable to about 2CFM.

Plenty to start processing for a few days..mostly retests, but it should add to the round database..