Both fans behind the same dust filter and on the same 30 mm radiator, and what each needs to deliver the same airflow. Calculated from the datasheets with the CaseCalc model, not measured.
| DeepCool THETA 21 PWM 1700 | Gelid SILENT 9 PWM SCHWARZ | |
|---|---|---|
| Free air, datasheet | 41.9 CFM | 37.9 CFM |
| rpm range | 500 – 2400 | 900 – 2000 |
| Noise at max, datasheet | 30.2 dB(A) | 23.5 dB(A) |
| Blades | pressure-optimized | pressure-optimized |
| Behind dust filter, max rpm | 33 CFM | 30 CFM |
| On 30 mm radiator, max rpm | 29 CFM | 26 CFM |
| CFM delivered | THETA 21 PWM 1700 rpm | dB(A) | SILENT 9 PWM SCHWARZ rpm | dB(A) | Quieter |
|---|---|---|---|---|---|
| 10 | 734 | <15 | 900 min | <15 | both inaudible |
| 20 | 1469 | 19.7 | 1353 | 15.5 | SILENT 9 PWM SCHWARZ by 4.2 dB |
Across the overlap, the SILENT 9 PWM SCHWARZ is quieter by about 4.2 dB at the same delivered airflow behind a dust filter. That is outside the ±3 dB cross-brand uncertainty and should be audible. A fan pinned at its minimum rpm delivers more than the row asks; the comparison is then not at equal airflow.
Delivered airflow is CFMfree air · derate · rpm/rpmmax. The free-air figure is the
manufacturer's. The derate is the share of it that survives the mounting; it does not depend on rpm because fan
pressure and flow resistance both scale with rpm². Airflow is therefore linear in rpm.
| Mounting | Airflow blades | Pressure-optimized blades |
|---|---|---|
| Open grille | 0.88 (0.80–0.95) | 0.88 (0.80–0.95) |
| Dust filter | 0.70 (0.60–0.80) | 0.78 (0.70–0.86) |
| Radiator 30 mm | 0.60 (0.47–0.76) | 0.70 (0.58–0.83) |
| Solid or glass front (side slits) | 0.50 (0.40–0.62) | 0.58 (0.48–0.68) |
The bands are the plausible spread of filter mesh density, radiator fin density and grille geometry. They are not statistical confidence intervals from measurements; they are the range the same fan lands in across typical cases.
Noise is the aerodynamic term Lmax + 50 · log₁₀(rpm/rpmmax) added on an energy
basis to a constant 6 dB(A) motor and bearing floor, so the curve flattens at low rpm instead of running to zero.
The band uses slopes of 40 and 60 dB per decade, ±1.5 dB on the rating and a 3 to 9 dB(A) floor. Manufacturer dB(A)
figures are not measured the same way across brands; differences under 3 dB between brands mean nothing.
Static pressure, where the datasheet gives it, is drawn as a parabola from the shut-off pressure to free delivery and scaled by the fan laws: flow ∝ rpm, pressure ∝ rpm². It is a shape between two datasheet points, not a measured curve; real axial fans dip in the middle.
Nothing here is measured. The values are what the fan laws and typical case resistances predict from the datasheet. Use them to compare and to size; expect a real build to land inside the band, not on the line.