Getting ready for some fan measurements using my propellor-comparison method of achieving equal (but unknown) airflows on two fans, and then measuring the noise level up close (*real* close) on both fans. This should provide noise/airflow figures of merit for each fan.
I have a sample or samples of five of the 120mm fans SPCR has reviewed, four from roundup 4 and one from roundup 2.
120mm fans that I own sample(s) of: their RPMs at 25CFM as measured by SPCR , my dBA measurement at that RPM and 0.25" from the center of the fan's hub, SPCR's dBA measurement at 1 meter, from SPCR Fan Roundup 2 or 4 (*n), and the fan identification:
590RPM 35.8dBA (<19, *2) Scythe SFlexE
820RPM 44.2dBA (20, *4) SilenX
880RPM 40.2dBA (20, *4) YL D12SM-12
950RPM 44.2dBA (28, *4) Enermax UC12EB
970RPM 46.4dBA (20, *4) GW NCB
(my ambient noise level <30dBA, SPCR's ~18dBA)
SPCR measured a high 28dBA noise level on the UC12EB, noting a single tone noise. My UC12EBs (two of them) were purchased when the fan was first introduced and don't have that single tone noise. SPCR got 25CFM from the SFlexE at *what* RPM
To these 5 fans, I'll add the Scythe SlipStream 800 and do some one-on-one airflow comparisons, noting the noise level at 0.25" from the hub on each. This should be interesting!
As I proofed the above words, the parts arrived. The $17 tachometer works, but in the interests of haste, I didn't use it for these first measurements. I'm gonna have to move the sensor propellor back - it tried to sync on the GW NCB RPM (but not other fans).
These measurements were made quickly but not carelessly. I've taken pics of the fixture, and will add them later as an edit. Here's the results I got for 4 fans:
As my standard fan I selected the GW NCB fan, at 970RPM. SPCR sez that's what it takes to get 25CFM. I measured 971RPM on the propellor - as I said, it was trying hard to sync (probably same pitch).
GW NCB 970RPM 971prop 46.4dBA (from above)
UC12EB 880RPM 971prop 41.8dBA (a sweet-sounding fan)
SFlexE 852RPM 971prop 42.3dBA (quieter than NCB!)
9blade 686RPM 971prop 40.5dBA Scythe SlipStream
My comments: aside from the 9-blade, the other 3 fan blades (by visual inspection) seemed to have very similar pitches. The SFlexE pitch was higher *near the hub*, but not at the edges of the fan, which is probably why its RPM is lower than the other two 7-blade fans.
How in the world SPCR got 25CFM out of the SFlexE at 590RPM I dunno. Either SPCR or I made a huge mistake. As I understand it, the only way for a 120mm fan to have that RPM advantage over another 120 is by having a drastically higher pitch... which the 9-blade fan does have and the SFlexE does not. Even the 9-blade needed almost another 100RPM beyond "590" to get the same sensor prop rotation rate.
Based on the highest RPM, the GW NCB fan has the highest measured noise and, presumably, the lowest pitch of these fans. I've always believed (and in one case, measured) that lower-pitch fans have an advantage in a typical PC case. My high regard for the NCB has always been for its performance in my cases.
Thus the caution: these are open-air measurements!
SPCR's fan noise measurements are very close to their room ambient noise, and perhaps below that for a fan or two. So differences of a couple-dBA get masked by the high ambient. I "close-miked" to avoid the room ambient problem. I had to block the sensor prop when making noise measurements; it was slightly unbalanced.
edit: added the following:
This fixture is made of white styrene .06" sheet. Note the small angle-brackets that locate all 25mm fans in the same spot, also styrene:
The $17 tachometer was not in use for these tests; I used my Extech stroboscope in the interests of getting done quickly. The fan is mounted on a 3/16" rubber strip and the entire fixture is supported by 9ea 3/8" square by 1/4" high "feet" made of very soft polyeurethane foam weatherstripping.