Loudspeakers

We couldn't quite match Revel's claimed 90dB sensitivity for the F228Be. While the simple response average came close at 89.5dB, the more representative pink noise and 'music' values (88.9dB and 88.7dB, respectively) suggest that 89dB is more realistic. Quite low impedance is used to help achieve this, the minimum modulus of 3.4ohm indicating a 4ohm nominal figure, not Revel's specified 8ohm. Impedance phase angles are large at low frequencies, resulting in a minimum EPDR (equivalent peak dissipation resistance) of 1.6ohm at 24Hz. While this isn't far off the 1.7ohm we typically find with modern high-sensitivity designs, further dips to 1.7ohm at 333Hz and 1.9ohm at 74Hz indicate that the F228Be is a tougher than average load to drive.

Forward frequency responses, measured at 1m on the tweeter axis, show an overall trend that's flat but with dips around 1.6kHz and 7kHz [see Graph 1, above]. Despite these, the response errors (300Hz-20kHz) are modest at ±3.1dB and ±2.8dB. Similarly, the pair matching error over the same frequency range is fine at ±0.9dB, improving to an excellent ±0.5dB if its narrow-band disparity at 1.8kHz is ignored. Diffraction-corrected nearfield measurement – the summed response of drivers and port – indicated a bass extension of 46Hz (–6dB re. 200Hz), an initial second-order decline in output below 70Hz becoming steeper below 30Hz. At the opposite end of the frequency range the response is maintained to over 40kHz (–6dB re. 10kHz) but what appears to be a 37kHz resonance frequency for the beryllium tweeter is a little disappointing given that it is possible to achieve a similar result using aluminium. The CSD waterfall [Graph 2] shows only low-level treble resonances. KH