Testing methodology. Results
To examine fan speed control and power features, we used motherboard fan header that delivers speed control via PWM method (24KHz). The header can handle loads of up-to 2.5A. Fan speed management and RPM monitoring was organized via SppedFan utility. This allowed fan power voltage change within full range 0-12V with 1% stepping. A digital multimeter connected in parallel to fan showed voltage readings in real time. Another multimeter was used to collect data of current drawn. All the measurements were taken with 1V stepping, until the fans stopped sending RPM readings or/and stopped rotating.
At this point, the following data was collected:
Looking at the graph, we can note a non-linear fan speed change pattern, which is not very convenient sometimes and can also cause additional noises at lower RPMs. On the other hand, nonlinearity of power drawn curve is good, since fan consumes less power at lower speed rates than if he had linear dependence.
L.N.A and U.L.N.A adapters correspond to 9.00V and 6.35V respectively. RPM readings for adapters does not perfectly match numbers specified in official fan specification, but are very close to it: 879 vs.900 RPM and 659 vs 600 RPM, respectively. Note that Noctua NF-S12B FLX starting voltage is 5.27V, which is quite high considering users often use 4-pin Molex connectors as 5V power sources to keep fans quite. Well, it won’t work with these fans. The start-up current is 0.051A which is smaller that the nominal value (0.076A).
To evaluate fan efficiency, we use triple 140 water-cooling radiator with corresponding adapters to exclude losses of pressure and airflow. Thermal load is provided via 300W in-line heater. Sensors located on radiator collect temperatures of Air passing through it as well as water in/out temperatures. Fans were tested at three speed rates — 1200/900/600 — using supplied power reducers. For simple comparison we added San Ace 109R1212H1011 fan that is known for its high performance and share almost the same price with Noctua NF-S12B FLX.
- Pump: Swiftech MCP35x @ 12V + stock top
- Reservoir: 5 litter canister
- Radiator: Thermochill PA140.3
- PVC hose: inner diameter 12mm, outer diameter 16mm
- Barbs: 1/2 Fatboy, inner diameter 9.8mm
- Data acquisition devices: mCubed T-Balancer + Sensor Hub
- Heat carrier: distilled water
- Thermal load: 1x Hydor ETH-300 in-line heater
Test results show that Noctua NF-S12B FLX can be quite effective with WC radiators, however, mainly low static pressure had put it into second place when compared to San Ace fans. Regarding adapter use, note that when lowering fan speed by 25% (L.N.A) Delta growth is only about 20%, whereas another 25% speed drop (U.L.N.A) results in almost 75% Delta increase. This indicates how non-linear fan static pressure is in reference with RPMs.
We have also measured noise level and air flow for Noctua NF-S12B FLX installed on the rad. The results are tabulated below:
Noise wise, Noctua NF-S12B FLX is barely audible at full speed or with L.N.A. power reducer, however, low-frequency humming can be heard at very close distance to the fan. U.L.N.A. makes the fans dead silent, although overall efficiency is reduced considerable, but such performance should be enough for particular tasks.