I just completed testing both channels of the latest version of the smd version of the CFA3 boards, and here are some distortion measurements:
Test Setup: balanced 1KHz/6Vrms from Tektronix SG 505 oscillator,
to HeadAmp Gilmore CFA3 (balanced output),
to General Radio 1840-A output power meter
& to Tektronix AA501A THD analyzer (which has about .001 to .003% THD on its own).
2Vrms Output THD Test
(Load, THD)
300, .002%
200, .002%
150, .002%
100, .002%
50, .003%
32, .007%
16, .01%
6Vrms Output THD Test
(Load, THD)
300, .002%
200, .002%
150, .002%
100, .002%
50, .003%
32, .013%
16, .15%
20Vrms Output THD Test
(Load, THD)
300, .017%
200, .017%
150, .018%
100, .022%
50, .26% (12W)
32, .42% (18W)
Max Output Level THD Test
(Load, Watts, THD)
300, 2W, .03%
200, 3W, .03%
150, 4W, .04%
100, 6W, .05%
50, 12W, .26%
32, 18W, .42%
16, 20W (meter limit), <1% ZF, <.3% SS
One thing that surprised me was the 2 channels had virtually identical THD (even in ZF mode) out to 2-3 decimal places until getting down to ~16 ohm impedances where there started to be a little variation