The last days I have been doing some performance measurements on AMP6-BASIC. It is a standard component AMP6-BASIC, no modifications done.

THD tests

First a 1KHz signal while varying the power output.

AMP6_BASIC_thd_1khz1.PNG

The THD is typically around 0.02% and starts to go up when the output reaches 7.5Vrms (+/-10.6V peaks) which is 14W into 4 ohms. At this level, the current is around 1.5A per channel. At 9V out, the output would be 20W per channel and the distortion around 1%.

Here are THD+N data for 14V supply, 4 and 8 ohm load. The output was 4VRMS, which is equivalent to 4Wrms at 4 ohms and 2Wrms at 8 ohms. The difference between 4 ohms and 8 ohms is surprisingly small.

AMP6_BASIC_thdn_4VRMS.PNG
The FFT looks like this for a 1KHz signal. The noise floor is typically -100dB and the main harmonics at around -80dB

AMP6_BASIC_thd_1khz.PNG

There is a drop of THD+N at around 7KHz the plot below shows why. The third harmonic is at 21KHz for a 7KHz fundamental. When the frequency is increased a bit more, the 21KHz peak moves to the right in the diagram, falls outside, over, the measurement range (and audible range) and therefore the THD drops by around 10dB. This is the reason THD diagrams often drop at high frequency.

AMP6_BASIC_thd_7khz.PNG

The measurements where done at 48KHz, with a delta-sigma type ADC which brick-wall filters the signal at around half the sampling rate.

The signal source was a EMU 0202 external sound card which has a typical THD+N of around 0.002% The meter itself, a NI PXI-4461 has a typical self THD+N well below 0.001%

MAX POWER test



I also did some maximum power tests, at the edge of frying the chip.
  • 2 hours at full load, sinusidoal and pink noise signal, well into clipping, up to 5% THD, 16.5V supply voltage at this torture power output levels, constantly above 2A per channel. No damage, no parts runing hot. This should be a higher average power level than ever possible with music.
  • I still do NOT recommend using more than 14.5V supply! Why?
    • My test was with a very stable 10A bench power supply. The voltage was very stable. Probably much better than most smaller supplies. A less stable supply may have had voltage spikes, killing the chip.
    • There are manufacturing tolerances. I may have been lucky.
    • I used a large heat sink. Keeping parts cool is vital for survival.
    • I will test to higher voltages later, but need to swap the bulk capacitor first...