It is currently Thu Apr 17, 2014 2:03 am

All times are UTC - 8 hours




Post new topic Reply to topic  [ 72 posts ]  Go to page 1, 2, 3  Next
Author Message
 Post subject: Noctua NF-P12 ... vs ten other fans. Done.
PostPosted: Wed Jan 16, 2008 6:34 pm 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
While on paper, this Noctua should handle resistance better than most fans, but not with the testing today. :( This fan is Noctua's newest and according to the hype, it is made to handle restrictions better than other fans.....because of the blade design apparently. This means it should blow through tightly spaced fins, or suck through filters better than a more normal fan design. Didn't happen. Here's the setup I used. The duct to the Ninja is removed in the photos, and the filter is pulled out. This filter can be removed while the computer is running, and the change in temperatures and RPM observed.

Image

Image

The fan is controlled by a NoiseMagic NTM3 controller......which will raise/lower fan speed according to the CPU temperature. Ideally, the temps and the RPM would not change when the filter was pulled out.....that is if the fan was not affected by the filter restriction. In the real world, some change is to be expected. To compare this Noctua to a different fan with the same test today, I first used a Scythe 1600rpm S-Flex. With the S-Flex the CPU idled at 31C@650rpms. With the Noctua the idle was 32@750rpms. This was with the filter installed.

When I pulled the filter out at an idle, the S-Flex dropped about 45rpms, as the controller readjusted the fan speed.....the temp remained the same. When I pulled out the filter with the Noctua, the temp dropped 1C and the rpms dropped 55. These numbers indicate the Noctua was more affected by the filter than the S-Flex......opposite from what I expected.

The numbers when the CPU was being stressed by CPUBurn were even more in favor of the S-Flex. The Noctua was affected 80rpms by the filter, and the S-Flex only 50.

These are real world tests conducted on a particular system. A bench test with different example, might return different results. But.....doesn't look good for the Noctua. It costs almost 2x the S-Flex, sounds about the same at similar rpms, and is more affected by restriction. YMMV.....but I am very disappointed in this fan.

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Last edited by Bluefront on Sat Jan 19, 2008 10:56 am, edited 5 times in total.

Top
 Profile  
 
 Post subject:
PostPosted: Wed Jan 16, 2008 7:50 pm 
Offline

Joined: Mon Jan 01, 2007 9:57 am
Posts: 530
Location: US
How's the noise quality of the new Noctua? Did you have one of the high-pitch fans to compare it to?

I have S-flex fans myself and love 'em. I'm just curious - is there a noticable improvement between the old and new Noctua's?


Top
 Profile  
 
 Post subject:
PostPosted: Wed Jan 16, 2008 9:07 pm 
Offline

Joined: Tue Jan 15, 2008 7:39 pm
Posts: 29
Location: Colombia
Are you speaking about the Scythe S-FLEX SFF21F? If that's the case, I'd like to see some more examples. I'm really interested in reading some more comparisons like this.

I'm currently looking for this. Why? Because I'm about to buy three 120mm fans, and my choice right now is between the Scythe S-FLEX SFF21F and the Noctua NF-P12.

For the Scythe we all have read comparisons and seen its performance, so it's a known fact. For the Noctua, all reviews I've read give it a high score, but unfortunately none compares it with a Scythe.

If someone could do more tests, I'd appreciate it.

_________________
Thermaltake Tsunami VA3000BWA | Seasonic M12 600W | Gigabyte GA-P35-DS3L Rev. 2.0 BIOS F9 | Intel Core 2 Quad Q6600 | G.Skill PC2-6400 2GBPK 800 MHz 4-4-4-10 @ 2.0v | 2x1 Hitachi Deskstar P7K500 500 GB | EVGA GeForce 8800GT 512 MB | LG L1753T 17" | System picture


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jan 17, 2008 3:30 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
Yeah....that's the S-Flex model I used. And after considering the results again, I've determined to re-do the test. This time I'll try four common fans with similar CFM ratings.

Yate Loon ..... D12SM12 ..... 70.5cfm @1650rpm
Scythe S-Flex.... SFF21F ......63.7cfm@1600rpm
Noctua ....... NF-P12.......53.3cfm@1300rpm
Scythe Slip-Stream.... SY1225SL12M.....68.5@1200rpm

Like the first test, this next test will be focused on how the different fans respond to the same restriction......in this case a simple foam air-filter, shown in the photo. It's about 6"x12", located about 2" from the fan blades, in a relatively big chamber. In order to eliminate variables, I'll dis-connect all other fans, except the PSU fan. The ambient temp in the test room stays about 23/24C. The MB is an Intel D865PESO......and the temps and the rpms are very steady using SpeedFan 4.33. Best I can do for now....

One would think the Noctua would be affected the least by the filter, considering the way it is being hyped. It's maximum rated CFM is slightly lower than the other fans, but that should not make any difference in this sort of test. Remember.....I'm testing for restriction effects, not cooling ability.....although the two are closely related.

If anyone can think of anything else to add to the test, speak up...... :lol:

Oh...I do not have the earlier Model Noctua to test (never heard one either). The actual sounds from these fans are very similar at the same rpm. The Yate Loon runs faster at 5V than the other fans. The fan control varies from 5-12V depending on the CPU temp. These are all quiet fans motor wise.....the cfm determines most of the noise.

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jan 17, 2008 7:25 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
Here are the results of this second test.....you might interpret them differently from the way I do. But the Noctua failed a second time.....this time finishing last to the three other fans, in a airflow resistance test. No doubt of the results for me. The S-Flex and the Yate Loon handle the resistance the best. Here are the numbers.....using only the single fan, and only judging the CPU temp (all the other temps were similar).

Yate Loon_____________Idle/no filter=28C@870rpms
_____________________Idle/filter=29C@880rpms
_____________________100% use/no filter=42C@1175rpms
_____________________100%/filter=43C@rpms@1265rpms

Scythe S-Flex__________Idle/no filter=30@580
_____________________Idle filter=31@595
_____________________100%use/no filter=42@1020
_____________________100%/filter=44@1105

Noctua________________Idle/no filter=30@590
______________________Idle/filter=31@670
______________________100%use/no filter=44@1025
______________________100%use/filter=46@1160

Scythe Slip Stream______Idle/no filter=30@760
______________________Idle/filter=30@825
______________________100%/no filter=42@1010
______________________100%/filter=44@1105

Here's what I call the Total Airflow Resistance Effect (adding the rpm effects at an idle and at 100% for each fan...lower is better)

S-Flex________90
Yate Loon_____100
Slip Stream____195
Noctua________205

Of course there other factors to use when buying a fan, but this resistance factor is important to setups with filters, close fins, etc.

The performance of the Slip Stream was expected.....looking at the thin blades. The Yate Loon was the surprise.....handled the filter well, and sounded about like the other fans. I thought the S-Flex would be the winner of this second test....it was.

The Noctua was the disappointment......as it did not handle the filter well, and returned the worst temps at a higher RPM than the S-Flex.

The best value.....the Yate loon by far @$4 vs the Noctua @$21.
The Yate Loon D12SL12 would probably perform about the same as the D12SM12 used in this test. As always.......YMMV

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Last edited by Bluefront on Thu Jan 17, 2008 12:22 pm, edited 1 time in total.

Top
 Profile  
 
 Post subject:
PostPosted: Thu Jan 17, 2008 10:14 am 
Offline
Moderator

Joined: Mon Dec 09, 2002 7:11 pm
Posts: 7335
Location: Maynard, MA, Eaarth
Hello Carl,

Good test -- thanks for doing it. I guess that either your filter is higher "drag" than any advantage the 9-blade fans have -- or they have no advantage.

Is the increase in RPM's from no filter to having a filter only due to the filter's drag? :shock: Does the sound quality change between no filter and having the filter?

I gotta' say that 1100RPM fans are making more noise than I would like, especially if there is a "beating" due to the restriction from the filter. Is this the least restrictive filter you can use? What about running w/o a filter and just cleaning it regularly?

_________________
Sincerely, Neil
http://neilblanchard.blogspot.com/


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jan 17, 2008 12:08 pm 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
Neil....what happens when you reinstall the filter is that the airflow is reduced, and the temperature of the CPU starts to go up. At that time the fan controller raises the RPMs to compensate for the reduced airflow....that's why you see the rpms higher with the filter.

Now if you do not have the fan attached to an auto-controller, when you put a filter over the fan, the rpms go down. I did a bunch of experiments with different filters......this foam type is one of the least restrictive. This is the type of foam used in those little fan filters sold various places. It's ok, not the best for stopping dust, but it works. As big as this one is, it shouldn't need to be cleaned too often, and it's easy to get out anyway.

Keep in mind what I'm cooling here.....a 100w CPU, and keeping the temperatures very low to boot. With practically any modern CPU, the maximum fan rpm would be much lower. And the setup operates on just the one fan.....less noise potential. Plus the fan is inside the setup. You'd have to hear it to see how quiet it is. At anything other than bench-marking, you wouldn't know it's running.

As far as the Noctua is concerned....it's very quiet, but just doesn't handle restriction very well, unike the Noctua ads have you believe. Maybe somebody else can do similar testing with the thing. I'm not impressed at all. :(

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jan 17, 2008 12:22 pm 
Offline

Joined: Wed Nov 28, 2007 1:26 pm
Posts: 322
Location: Netherlands
So basicly, the new Noctua's are ok, except when using filters?

I was considering using these, but they are rather pricy here, much more then the s-flex, nexus and slipstreams. If they are not noticably more silent, and on top of that won't handle filters better, then i see no reason to pick them.


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jan 17, 2008 12:29 pm 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
Yeah that's my opinion. For the price of one Noctua, we here can buy five Yate Loons......low, medium or high speed. And get better temps, with about the same noise. In the last test I used a medium speed YL.....the low speed model might be slightly quieter at 5V.

Actually the S-Flex stands out, what with it's superior bearings, and good performance.

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Thu Jan 17, 2008 7:34 pm 
Offline

Joined: Fri Feb 23, 2007 11:55 pm
Posts: 410
Blue, have you given a thought to 120x38 fans? they should provide more presure at the same rpms, at least theoretically. I know it's not the point of your post, so...

i've got recently a pair of DFS123812-1000, at about 6-800 rpm they are pretty smooth, as far as the noise.

there are higher rpm models of the same, like DFS123812-2000. that one will need to be dialed down for sure.

hopefully it will fit, not sure about the clearances from the picture.

i also found that delta tri-blades low, even severely undervolted, will push a lot of air through pretty much anything.

_________________
dothan 730 @ 2.6Ghz l p4c800ed l 4x512MB bh5 2-2-2-5 5:4 ~3v Antec Neo 480
e6600 @ 3.6GHz l P5W64 WS Pro l 4x1GB D9GMH (B6-3) 3-3-3-3-12 4:3 2.2v l PCP&C 750


Top
 Profile  
 
 Post subject:
PostPosted: Fri Jan 18, 2008 3:48 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
I have considered using 120x38mm fans, but have never tried many. Lately there has been an increase in the number of that size available, so I may try some.

Concerning this restriction thing.....to over-come restriction effects on a fan, you simply need to increase the RPM somewhat. You can see this happening when using that NTM3 controller. It senses a slight temperature increase at an idle, and it raises the RPM slightly......gradually in .1 volt steps. That's why I like these controllers so much. There is one big problem with it that you can see in the test results. There is no way to adjust the base idle speed of a particular fan. That Yate Loon fan for instance.....it runs faster at 5V than the other fans (not sure why), but I cannot lower the idle speed, even though it could idle at 4V and still push as much air as the other fans. A resistor would do it, but would also lower the maximum rpm.

As far as the testing procedure I used.....works for me. The fan is being restricted by the removable filter, but it is also being restricted by the Ninja, and the other pieces in the airflow path. And there's no mistaking the restriction effects as you watch the temperature and the RPMs change.

You could duplicate this setup somewhat on the bench, by rigging up a fixture with a box with a ninja on one end, a fan in the middle, and a removable filter on the other end. You'd have restriction on both sides of the fan, and could judge each fan by the loss of rpm at various voltages when the filter was removed/installed. I made such a setup a while back....and used it to test various filters. A similar setup could be constructed to test various fans for the resistance effect.

Felger Carbon has also done similar testing.....but I don't think he's used filters in his procedures. And that's what concerns me the most when picking a CPU fan. We already know the Ninja is the least restrictive heatsink......what we don't know is which is the best fan to use to overcome the added resistance of a filter. I thought the Noctua had a chance to be the best.....apparently not.

So who is going to volunteer to duplicate this resistance testing? I can do it, but I only have a limited number of 120x25mm fans to test......and most of the ones I did not test, have noisy bearings, which eliminate them before any other testing. :lol:

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject: Noctua in Lian Li PC-S80
PostPosted: Fri Jan 18, 2008 5:28 am 
Offline

Joined: Fri Aug 31, 2007 1:56 am
Posts: 11
Location: Worldwide
Hello,

After buying a Lian Li PC-S80 case (against SPCR advice, I just liked the last Lian Li I owned) I wanted to quiet it down as far as possible and gain cooler temps. It has a restrictive airflow from two side vents near the back (right and left) that pass the air through a narrow channel between the case side and an inner liner to fans at the front. The air then is drawn through two fans (in a stack) over the harddrive tower and on into the mainboard section of the case.

Because of this "restricted" air flow I felt that Noctua P-12s (based on numerous reviews) would be a good choice to further cool and quiet my machine.

I measured the temps at idle in two configurations. Config 1 with stock fans in place. Config 2 P-12s in the front two 120mm fan locations. And then the same configurations for burn runs with Orthos.

What I discovered disturbed me. I had to have the fans on the same or higher RPM to gain the same cooling I had with the stock fans.

At idle my temps were one degree better in some areas. At burn the same story. I have an excel spreadsheet but it fell apart when I pasted it in.

So in the PC-S80 with my system the Noctua fans are a costly disappointment.

I have since started testing with a TRUE on this motherboard but am still looking at that as well.

Questions?

_________________
P182, A8N32 SLI, 4800 X2 & TRUE, HX620, 2G Corsair , 74 Gig Raptor, 500 Gig Samsung, XFX 7600 GS


Top
 Profile  
 
 Post subject:
PostPosted: Fri Jan 18, 2008 12:47 pm 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
That's about the same conclusion I made.....a costly mistake. What I don't understand is the Noctua certainly follows accepted theory about blade shape/size and resistance. It should be good with resistance..... :?

If anyone can figure a better way to compare fans with regard to the ability to handle resistance, I'd like to hear about it.....maybe try it. Come up with a resistance test where the Noctua beats an S-Flex......

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Fri Jan 18, 2008 2:16 pm 
Offline

Joined: Wed Mar 21, 2007 10:45 am
Posts: 223
Just to be clear here, are you measuring performance with RPM fixed or with the noise level fixed? Because different fans produce different noise levels at the same RPM, and nobody cares about RPM, they care about the noise.

So for instance, with the RPM of the Slipstream and the P12 both tweaked to make them sound the same, does the Slipstream still outperform the P12?

PS: A way to test flow is to place an empty bag on the output of the obstruction that the fan is blowing onto. So fan->grill->bag. When you know how much volume is in the bag, and you measure how long it takes for the bag to fill up, you can tell how much air volume was moved per second.

So a standard test platform to do this would have a mockup of a PC chassis with a grill or filter for the fan to blow through. A bag of known volume would be attached to the other side in such a way that all air is easily removed from it before the test.


Top
 Profile  
 
 Post subject:
PostPosted: Fri Jan 18, 2008 2:36 pm 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
To repeat myself.....the tests I performed were in the computer, with the fan hooked up to an automatic controller. I noted the rpm change needed to maintain aprox the same temp with/without the filter.

If for instance it took an increase of 100rpms to maintain 31C when I installed the filter, the score would be 100. But I also measured the increase to maintain the same temps running 100%, and added that to the idle score. There are other ways to score this.....but the final result would be the same.

I've heard of that "blow up a bag" test.....just not accurate enough. How do you tell when a particular bag is completely full of air? It would only be a guess. The test I used involved the actual computer using sensors and rpm measurement.....very accurate IMHO.

I have no accurate way to judge noise volume.....all four fans sounded very similar at the same rpm, very quiet at an idle, and audible when the CPU was being stressed.

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Fri Jan 18, 2008 7:53 pm 
Offline

Joined: Tue Mar 11, 2003 5:17 pm
Posts: 268
Location: Canada
Errr wish I didn't buy 3 of them awhile back. :evil:

I was planning on to use them on a radiator but I guess S-Flex is the way to go.

_________________
Asus P5Q-E, Intel Q6600 stock, D-Tek Fuzion v2 w/ Quad Nozzle Kit, OCZ PC2-6400 (2x2GB), Enermax Modu82+ 525W, 8800GTS w/ Accelero S1, Lian-Li V2100B Plus


Top
 Profile  
 
 Post subject:
PostPosted: Fri Jan 18, 2008 7:54 pm 
Offline
Moderator

Joined: Mon Dec 09, 2002 7:11 pm
Posts: 7335
Location: Maynard, MA, Eaarth
Hello Carl,

Can you post a picture of the exhaust on this case? I have seen this Dell case before -- are you just using the stock grill, or did you add some more vent area? I'm assuming that you are working this with positive pressure?

It might be that pushing air through a case -- even w/o the filter is already developing enough back pressure to overcome any fan's abilities. Positive pressure by definition means there is some back pressure in the case. And therefore, some of that pressure must push back out through the fan itself...which lowers it efficiency.

Pushing air is simply harder to do, than is pulling air. If you reduce the pressure inside the case, the surrounding air naturally moves in to equalize the pressure, and this in turn causes more air further and further away from the fan to move. The further away you get, there is less velocity, but the air keeps moving.

Looking back at your numbers, the Yate Loon has the highest RPMs of all the fans in all conditions -- do you prefer it? It only "beats" the two Scythe fans by 1 degree, and it has ~160RPM more...

The Noctua almost matches the S-Flex RPMs at every level except 100% load with the filter -- how do their noise levels compare? The Slipstream does virtually match the S-Flex for RPMs -- and temps, in all conditions.

I guess my point is, all these temps are completely fine, so it should come down to the sound they each make vs their cost. (Obviously, the Noctua is very expensive...) The Yate Loon is the least cost, followed by the Slipstream? The YL runs significantly faster -- it must be louder?

_________________
Sincerely, Neil
http://neilblanchard.blogspot.com/


Last edited by NeilBlanchard on Sat Jan 19, 2008 6:02 am, edited 1 time in total.

Top
 Profile  
 
 Post subject:
PostPosted: Sat Jan 19, 2008 1:40 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
Since all four fans in the test were sucking through the same filter, and blowing into the same positive pressure setup, with all four tests conducted within three hours at the same ambient.......I'd say the testing was fair, giving no advantage to any of the fans. Here's what the only exhaust in the entire case looks like....it's non-restrictive. The little sparkle PSU is the only restriction at this point.

Image

And here's what the setup looks like with the duct to the fan installed....

Image

But to remove all the variables from the test, I can do it over. This time the fans will be blowing into free air, outside the case, with the voltage fixed. I'll test for an rpm drop when a filter is placed over the intake side of the fan. I'll do the tests at about 5V and about 12V and add the two scores. The numbers will be different, but the relative differences should be the same.

Now this is not a "real world" test of a fan. But it should give you an idea of how this Noctua handles an intake restriction. According to the Noctua ads, this fan should excel at this sort of test, if I'm reading the ads correctly. We'll see. I'll also try several different filter types, and some other "not so quiet" fans.

Neil....all the Yate Loon fans spin faster at lower voltage than many other fans, and definitely faster at 5V than the other fans in this last test. I cannot explain this. Whether it makes any difference in this sort of test.... I don't know. But the Yate Loons do push more air at 5V.....and sound slightly louder of course.

FWIW......right now that computer used in the test has the S-Flex 1600rpm fan installed. It's running about 590rpms with the CPU@31C. The Yate Loon cannot run that slow @5V. So it would be slightly noisier.....but still well below my ability to hear it from 3 feet away.

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jan 19, 2008 3:33 am 
Offline

Joined: Thu Feb 26, 2004 3:40 pm
Posts: 424
Location: EU
Good luck with the new test, very informative test. :)


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jan 19, 2008 4:17 am 
Offline
*Lifetime Patron*

Joined: Tue May 15, 2007 1:39 am
Posts: 1889
Location: Finland
The more tests the merrier - and better for accuracy.

Thanks for the effort; am currently looking into a heatsink fan for a friend, and this helped.

_________________
Case: FD Define Mini
Parts: P8Z77-M Pro µATX, N650Ti-1GD5/OC, G.Skill 2x4/1600/CL9 DDR3U, Xonar DX, WD G 1 TB, m4 128 GB, RX-5300 PSU
Cooling: Noctua NH-U12P SE2 + Scythe SS PWM, 2x Noctua NF-P12
Extras: Eaton UPS, Dell 24" EIPS, Ducky TKL kb, SteelSeries m, Synology DS213j 3+3 TB RAID1
idle & load: CPU 32 °C & 44 °C @ 300/600 & 600/800 RPM, GPU 35-65 °C @ 1200-1650 RPM


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jan 19, 2008 5:04 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
Ok....Here's the line-up for this next test, along with a photo of the test box. A 120mm fan is held snugly/tightly on the one end. The other end holds the filter, removable of course. The filter size is 5"x7" with the filter being held about 1.5" from the fan blades. The fan can draw through the entire area of the filter. For now the winner of the last test (1600rpm S-Flex) is running the computer, and taking it's place is an 800rpm S-Flex. From the upper left to the right and on down.....

1. Yate Loon D12SM12
2. Scythe Slip-Stream SY1225SL12M
3. Evercool EC12025M12BA
4. Noctua NFP12
5. Scythe S-Flex SFF21D
6. Enermax A1212025MULF
7. Yate Loon D12SM12C (a 120x20mm)
8. Mad Dog (a re-badged YL about 1800rpm)
9. Yate Loon D12SL12
10. ThermalTake A1225L12S (came with a Big Typhoon HSF)
11. Globe S1202512L-3M (older, rare, the sensor shorted)
12. Antec (120mm 3 Speed)

Image

I have no idea which fan will win the test. From looking at the blades, it looks like the Noctua should win, with the TT coming in second. My guess would be the Globe.....nice quiet fan that can run 1800rpms.

Remember I'm testing for the rpm drop when using a filter....nothing else. What's your guess here? I suspect I'll finish this afternoon. I hope the Noctua does better this time, since it cost the most by far, and is the only fan claiming to handle restrictions well.

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jan 19, 2008 6:18 am 
Offline
Moderator

Joined: Mon Dec 09, 2002 7:11 pm
Posts: 7335
Location: Maynard, MA, Eaarth
Hi Carl,

I was not saying that it is "unfair" to any one fan -- I was wondering if the positive pressure overwhelmed any advantage that any one fan might have? The only way to test this would be to install the fan as an exhaust, and pull the air through the filter, duct, etc.

My hypothosis is that pulling air through a case is easier to do/more efficient than pushing it through the same case. This would be because of the way air flows and the way it reacts to differences in pressure.

_________________
Sincerely, Neil
http://neilblanchard.blogspot.com/


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jan 19, 2008 7:36 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
Neil....this test in progress eliminates positive pressure effects, and is just sucking through the filter. This mimics most case fan setups better than the previous test. This test is acting like a negative pressure setup, with the fan blowing into free air.

Actually if a fan handles resistance well, you would think it wouldn't matter if the resistance is on the intake side, or the exhaust side of the fan. When you hang a fan on a heatsink, blowing through the fins, the resistance is on the exhaust side. Using the same fan on the case blowing outward, the resistance is on the intake side.

I had to remove the Antec fan from the test. I forgot it doesn't have an RPM wire. Dumb Antec fan...... :lol:

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Sat Jan 19, 2008 10:45 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
I wish I could say the results of this second test were easy to interpret.....they are not. The 1600 YL that did well in the case test, did poorly in this free air test. The Noctua did much better this time......finishing 3rd out of eleven fans. The final score was determined by adding up the RPM drop in each of four different tests......each fan at 5V and 11.5V with two different filters. The first filter is the black foam type used in the first testing. The second filter is a pleated furnace filter. This is more restrictive than the foam, but because of the increased surface area of the pleats, it seems to provide the fan a similar restriction to the flat foam type. Here's the final score....lower is better. I listed the RPM of each fan at 5v and 11.5v without a filter.

1. Scythe Slip-Stream SY1225SL12M 800/1230rpms............Score 55
2. Evercool EC12025M12BA 1105/1880rpms.......................Score 60
3. Noctua NFP12 730/1285rpms.........................................Score 70
4. Scythe S-flex SFF21D 400/840rpms................................Score 80
5. Enermax A1212025MUL-F 454/845rpms...........................Score 80
6. Globe S1202512L-3M 985/1815.......................................Score 100

7. Yate Loon D12SL12 820/1340..........................................Score 140
8. ThermalTake A1225L12S 735/1250..................................Score 160
9. Mad Dog (no model #) 1010/1710...................................Score 215
10. Yate Loon D12SM12 1030/1610.......................................Score 265
11. Yate Loon D12SM12C 1050/1660.....................................Score 330

The scores of the first six fans are close enough that another set of tests would change the order of finish. One of the slower fans finished first. The fastest/loudest fan finished second. The Noctua finished well, in the middle of the first group, along with the slowest fan, the 800rpm S-Flex.

The Yate Loons were the surprise. The 120x20mm YL finished last. Who would have guessed that these quiet fans have trouble with restriction? Anyway.....this is a single set of tests done by me today .
A different set of tests done by somebody else using different examples of each fan,might give much different results.....but I doubt it. I was careful and checked everything numerous times.

My own personal take on all this.....I'm still saying the 1600 S-Flex is your best bet in a setup with restriction and filters. I saw what it did in a real-world test with my own eyes. The Noctua ? Needs more testing by more/different people. How it preforms in your own case is what matters. It's plenty quiet, but so are the Scythe fans, all three I tried.

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject: RE:Noctua NF-P12 ... vs ten other fans
PostPosted: Sat Jan 19, 2008 12:32 pm 
Offline

Joined: Mon Dec 17, 2007 10:30 pm
Posts: 39
Location: California, USA
I was ordering some NoiseMagic controllers and saw they had the Noctua NF-P12 that looked like a very good design so I ordered one. Then I read this topic. Doh! :shock: I have some S-Flex fans already so maybe I should have waited for more reviews. Oh well, it arrived so I started doing my own tests. I have an S-Flex 1200 rpm on my CPU cooler that I try to run below 800rpm because I can start to hear it above that. I was hoping the Noctua NF-P12 would give more cooling (not that I really need more cooling) without getting louder.
My setup:
P182 case with no front vent door, dust filter or metal grill on front or rear (no restrictions).
S-Flex case rear at 710 rpm. No top fan and covered.
VGA exhausted out the back.
Intel E6850 stock with Thermalright 120-Ultra Extreme (with S-Flex E (1200) and for testing the NF-P12)
Prim95 (two instances of Small FFT's for 2 cores)
Ambient 22c

So I ran with the S-Flex at 774 rpm (manual with Fanmate) and Prim95 for 20 min. and then turned off the PC to let the heat sinks cool for an hour then installed the NF-P12 at 779 rpm and run Prim95 again.

S-Flex at 774rpm:CPU(Tcase)= 42c, Core0+15c=50c,Core1+15c=50c

NF-P12 at 779rpm: CPU(Tcase)= 42c, Core0+15c=50c,Core1+15c=50c
Update: tried some more tests and I get maybe one degree difference either way so I cant really say one is better than the other. It could be the difference of accuracy or ambient temp.

So they showed no difference, and that’s a bit disappointing because I wanted the Noctua to do better.
But wait, this is just the rpm vs. temp test. I want to test it some more to compare noise to temp. It sounds like the NF-P12 can run to 1000rpm and maybe 1100 (in my PC with sound acoustic foam) without being as loud or at least annoying as the S-Flex. I turned the rpms up to 1100 and it was a bit louder but more of a whoosh sound, but the temps did not come down. Maybe the pc needs to cool off again.
Need to do more testing but no time right now.

When running the fans in free air they sound like this:
Both quiet and about the same noise below 800rpm? (need to connect tach )
S-Flex above 800rpm starts to make a hum or buzz like motor sound but maybe the blades going past the motor struts. It gets louder with rpm but probably no more db's than the NF-P12, but I don't like sound compared to the NF-P12 noise. When I hold a grill to the fan outlet, the hum sound gets less.

NF-P12 above maybe 950 rpm begins to make a slight hum, but more a whoosh and whirl (slightly higher pitch) sound. The whoosh/whirl sound gets louder with rpm but not as irritating as the S-Flex IMO. Placing a grill to the outlet takes away whirl high pitch sound and leaves the whoosh sound.

Vibration test. I placed the fans on top of a cardboard box (the MB box) and the vibration that would be transferred to a case is heard if solid mounted. Rubber fan mounts would reduce most of this noise.

S-Flex has a low hum that is faint at medium speed and is then drowned by the fan noise at higher speeds.

NF-P12 has a low rumble tic-tic noise (almost like an air cooled VW bug engine) at medium speed and that is a bit louder than the S-Flex. The noise can be heard at all but the highest speeds. However, I did not hear any of this noise when it was mounted solid to the Ultra 120 cooler.

So far I have not seen the NF-P12 perform better temperature wise than the S-Flex with my testing, but if I want to run a fan above 800rpm then I like the way the NF-P12 sounds better.

Another note, with the Fanmate at its lowest setting and the fans on the CPU cooler the S-Flex was 420rpm and the NF-P12 was 574 rpm

I was wondering if the NF-P12 fan blades are too close together and creates cavitations of air when pulling the air, especially at higher speeds.
More testing needed.
And a real CFM vs. rpm and db noise like Felger Carbon's would be interesting. Hint, hint :wink:


Top
 Profile  
 
 Post subject:
PostPosted: Sun Jan 20, 2008 6:41 pm 
Offline

Joined: Wed Mar 21, 2007 10:45 am
Posts: 223
Bluefront wrote:
I've heard of that "blow up a bag" test.....just not accurate enough.


I disagree but there is another way to get better results! http://www.google.ca/search?hl=en&q=digital+anemometer


Top
 Profile  
 
 Post subject: Noctua in Lian-Li and Antec Cases
PostPosted: Sun Jan 20, 2008 8:51 pm 
Offline

Joined: Mon Feb 05, 2007 4:14 pm
Posts: 218
Location: Forkbomb, New South Wales
pd230soi wrote:
After buying a Lian Li PC-S80 case...

Because of this "restricted" air flow I felt that Noctua P-12s (based on numerous reviews) would be a good choice to further cool and quiet my machine.
...
So in the PC-S80 with my system the Noctua fans are a costly disappointment.


Same here, (File Server) PC-A16, EX-34 drive cages. While the drives are all happy at 35-40C, the Noctua fans provide the same amount of cooling (as judged by the temperatures reported via SMART) with the same if not more noise than the Yate Loons they replaced. Both brands were throttled by an NMT2 thermocontroller.

I also tried them in my (HTPC) NSK2480, in place of two 120x38 Scythe Ultra Kaze (1000 RPM) fans. The cooling was much worse, and much louder. Both brands were BIOS throttled.

I don't think calling the Noctua NP-12 fans costly or a disappointment is strong enough- hopefully, they'll last a while, but I wouldn't recommend them to others. They're overpriced junk as far as I'm concerned.


Top
 Profile  
 
 Post subject:
PostPosted: Sun Jan 20, 2008 10:29 pm 
Offline

Joined: Mon Dec 17, 2007 10:30 pm
Posts: 39
Location: California, USA
I wanted a fan to replace my S-Flex 1200rpm fan for my CPU hsf. I wanted a fan that would either push more air at the same speed or spin a bit faster to push more air but without more noise (when controlled by the MB).
Well the Noctua NF-P12 pushed about the same amount of air as the S-Flex, but it spins faster for the same voltage from the MB and without much more noise, or at least to me the noise is a little less annoying than the S-Flex at higher rpm.
So the Noctua NF-P12 works OK for what I wanted.

BTW, its interesting that the fan spins faster with the same voltage because e the amp rating is less than the S-Flex (NF-P12=. 09ma, S-Flex=. 15ma) and I would think that would make it spin slower.
Either the fan blades of the NF-P12 move less air or the motor or bearing is a lot more efficient.
It seems that the leading edge of the fan blades have less angle of attack as the S-Flex.


Top
 Profile  
 
 Post subject:
PostPosted: Mon Jan 21, 2008 3:05 am 
Offline
*Lifetime Patron*

Joined: Sat Jan 18, 2003 2:19 pm
Posts: 5316
Location: St Louis (county) Missouri USA
That "Anemometer" probably could be used to measure fan/air speed accurately.....but you'd still have to construct something around it to serve as an exhaust restriction, and something to serve as an intake restriction.

The first test method I used (with the single fan blowing through a heatsink, and sucking through a filter) gives you a good idea of how a particular fan handles a "real world" situation....in this case the exact use for which the fan was intended.

Don't get me wrong here.....The Noctua is a quiet fan, but it doesn't live up to the way it is being advertised. Perhaps Noctua just means this NF-P12 handles resistance better than their earlier models. Maybe it does......but it certainly doesn't do better than the Scythe fans. YMMV

_________________
"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke


Top
 Profile  
 
 Post subject:
PostPosted: Mon Jan 21, 2008 4:57 am 
Offline

Joined: Tue Jan 15, 2008 7:39 pm
Posts: 29
Location: Colombia
It'd be nice if you and the guy that did this review shared experiences and methodologies. I found in that review the comparison I wanted: against a Scythe S-FLEX SFF21F, just like you did here. Since results may vary, I hope you both get in contact.

PS: In their forums, they give a lot of positive points to the Noctua NF-P12 using it on a Thermalright Ultra-120 eXtreme.

_________________
Thermaltake Tsunami VA3000BWA | Seasonic M12 600W | Gigabyte GA-P35-DS3L Rev. 2.0 BIOS F9 | Intel Core 2 Quad Q6600 | G.Skill PC2-6400 2GBPK 800 MHz 4-4-4-10 @ 2.0v | 2x1 Hitachi Deskstar P7K500 500 GB | EVGA GeForce 8800GT 512 MB | LG L1753T 17" | System picture


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 72 posts ]  Go to page 1, 2, 3  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: Google [Bot] and 0 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group