Martinm210

So after cleaning out most of my spare parts in my closet, I was left with a box full of fans, an anemometer and a very low pressure gauge that I never did use.

I pretty well burned myself out with thermal testing and water testing, but fans is something I havn't done much with. There's already some great fan testing out there with noise vs CFM on a radiator, but I havn't seen anyone try to measure a fan's full performance P/Q curve since Bill Adams was around many moons ago.

With that I decided I'd do a little something different. This first round, I'm tinkering with measuring a fan's P/Q curve. First just to see if I could do it, then to see if I can learn something about the different types of fan blades and thicknesses and the general performance behavior of fans in general.

So with that, this first round of testing will NOT include any sort of noise level testing, this is performance only.

First here is my testing rig I built, it's my second attempt after my first one failed because it was too restrictive. This one is an open box, anemometer permanently fixed at the inlet side (Laminar air flow works better), and my pressure gauge measuring at the inlet and outlet of the fan hub. To adjust restriction I simply move a foam box or my heatercore radiator different distances away from the end.

http://img3.imageshack.us/my.php?ima...stingpqrig.jpg

And this is what it looks like testing with the heatercore:


And this is what I have so far, but I still have several fans to test:

SUMMARY TEST RESULTS


And these are some of the things I've learned so far:

-Yate loons in general are "RPM Overachievers"... the curved D12SL12 run nearly at 1500 RPM which is what the medium speed spec is at.
-The curved bladed yate loons do perform a little better than the straight blade types.
-Fan efficiency is better the "Higher the RPM". In other words you get more CFM per RPM through a radiator the higher the RPM. Performance really falls on it's face with the ultra low speed fans. This is part of why some radiators can see up to 5X the performance difference simply by using different fans. Here is a graph of two of the higher speed fans showing this important generality. For example the San Ace can see 3X the efficiency at high speeds vs. low speeds. I always thought there would be some sort of optimum. In the case of fans, faster is always more efficient in addition to the increased power, it's more efficient AND powerful. I believe it has something to due with air and turbulence of slower speeds....you just can't slice through the wind when moving slowly as well.

-Static Pressure vs open Air Flow. They are about equally as important even looking at the pressure drop curve of a bonnie heatercore which is very dense and restrictive.
-Fairly minimal difference in performance per RPM. The San Ace is one of the stronger ones per RPM, but it also has more motor noise than some 25mm fans. Also not all 38mm fans are necessarily better. The 1K RPM Ultra Kaze didn't show much benefit from the 38mm thickness where it matters in the curve.
-Performance in general is primarily dependant on RPM as you would expect. I was hopeful I would be seeing some curves that are more rounded or have bigger humps in them to produce more flow for specific radiators, but I'm just not seeing too much of that. Some of the fans have a much more pronounced flat spot (Stall) especially the 38mm fans, but for the most part they have similar shapes with little bumps here and there with nothing dramatically distinquishing itself from the others.
-Sound will be key. In the end, I'm just not finding a whole lot here other than the fact that what's noted above, and the most distiguishing feature between fans will be noise. I've taken a liking to some of the noiseblocker fans myself, they do seem to have a better motor noise and smoothness to them compared to the others. The San Ace is also a pretty amazing fan regarding performance per RPM and it's ability to undervolt pretty well.

I will carry on until I run out then see what I can do about recording some noise levels some how.

Cheers!
Martin

INDIVIDUAL FAN TEST REPORTS, Test 1 (PQ Curve Varied Resistance, Test 2 Open Air Varied Voltage, Test 3 On Radiator Varied Voltage
I'll create links to my individual test reports here:

Noiseblocker XL-1
http://www.overclock.net/5890042-post14.html

Yate Loon D12SL12 (S) Straight Blade type
http://www.overclock.net/5890068-post15.html

Yate Loon D12SL12 (C) Curved Blade type
http://www.overclock.net/5890141-post16.html

Noise Blocker XL-2
http://www.overclock.net/5890177-post18.html

San Ace 109R1212H1011 TESTED @5V on PQ curve
http://www.overclock.net/5890216-post21.html

Scythe Ultra Kaze 1000
http://www.overclock.net/5890300-post26.html

Yate Loon D12SL12 S-LED Straight Blade LED UV Blue type
http://www.overclock.net/5890327-post28.html

Noise Blocker XLP
http://www.overclock.net/5890381-post30.html

Noise Blocker S1
http://www.overclock.net/5890783-post34.html

Noise Blocker S2
http://www.overclock.net/5893761-post38.html

Gentle Typhoon AP-15 (1850 RPM)
http://www.overclock.net/5895163-post41.html

Zalman ZM-F3 (1800 RPM)
http://www.overclock.net/5895769-post42.html

Noiseblocker S-3 (1800 RPM)
http://www.overclock.net/5896488-post46.html

Panaflo FBA12G12M
http://www.overclock.net/5897639-post50.html

Scythe S-Flex SSF21E (1200 RPM)
http://www.overclock.net/5948724-post67.html

Scythe S-Flex SSF21F (1600RPM)
http://www.overclock.net/5953019-post77.html

Scythe S-Flex SSF21G (1900RPM)
http://www.overclock.net/5955215-post79.html

Nanoxia FX-12 2000
http://www.overclock.net/5963171-post80.html

More to come...

__________________
Flow Rate Estimator | Radiator Sizing Estimators

Sullivan

So which one would you recommend for me to get, i need 4 120mm fans, those stats look nice.

I want some that are quite, push are, and cheap.

+rep for your work.

Martinm210

It's hard to know what people are willing to put up with when it comes to noise.

My personal preference would be this:

Cheap but good performance, fairly quiet = Curve bladed Yate Loon D12SL12s or D12SM12s. Get the mediums if you have a speed controller, then you can dial it in. Petra's carrys the curve bladed types, some others may as well, but you can see the back of the blade is curved and not straight.

More costly, but better sound = So far I like the Noiseblockers. The XL-2 is good quiet one with still decent performance. The S-2 may be another good option for a little lower RPM, havn't tested that one yet.

Performance all out, noise is not a problem = Any of the 38mm 2000+ RPM fans. I have some Ulra Kaze 3000, and they are extremely strong, but WAY TOO LOUD for me. Panaflo, Sunon, and Delta also make some really STRONG and very high quality fans. The San Ace is also a very high quality fan. When it comes to POWER, these high speed 38mm types are amazing.

Unfortunately it's always a give and take..you give up silence for more power, and you have to balance that yourself. In general I think you get better sound quality choosing the right fan speed and running them at a full 12V. Many of the fans exhibit a clicky motor sounds as they are undervolted, some worse than others.

I currently have 8 curved yate loon medium speeds on my dual quad rad setup and have them all turned down to about 1000RPM. Somewhere between 1000 and 1300RPM is all I can stand to hear for long periods of time. Unfortunately...they are all very weak in power at that point, but I can afford that with dual quad rads that are WAY oversized for the heat I'm dissipating.

__________________
Flow Rate Estimator | Radiator Sizing Estimators

Mr_Nibbles

Nice work, your tests are always very helpful.

__________________
"Never mind the milk, comrades!"

wastedtime

+rep martin. Thanks for the report

__________________
Never agrue with an idiot. They drag you down to their level and then beat you with their experience.

Complete - Gemini -Lian-Li PC-Q7 Mini-ITX

In Progress: ANT - NZXT Rogue SFF

Completed: Storm

Completed : Azure - A Rack Server Mod

Mmansueto

I love seeing tests from you Martin. Awesome info as usual.

SubstancenUnknown

yes +rep good work

ChielScape

back to testing marty?

__________________

MADMAX22

Very nice. Go heater core lol.

__________________
E6400@3856E4300@4158E6600@4510,Opti148@3102,opti170@2918,Xeon3220(B3)@3430(490fsb)
UD3P@8400600FSB

ira-k

Hey Martin, glad to see you back.

Nice set-up you built there, thats something thats really needed to. Fans are one of our main tweaks left anymore.

Are you going to try them on low fin per inch rads also? Really I guess whatever you found on the higher fin per inch would be about the same on low, except the sound level maybe.

Thanks for the testing! I was wondering when you'd get the bug again..

EDIT: I'm going to go ahead and put this with your rad calc's in the stickies.

__________________
...***Beware all ye Men born of Woman, beware the Hemogoblins***...
**E8600 @ 5751Mhz ***OCN Water Cooling Club And Picture Gallery ***150 Opty @ 3242Mhz **