- Written by John E. Johnson, Jr.
- Published on 27 October 2008
On the Bench
First, the SSP. Except where specified, I used the CD input, configured for XLR (balanced analog audio). I set the input as CD Direct, which bypasses all signal processing attributes including Audyssey Room EQ. Analog input signals were supplied by our Audio Precision SYS-2722 analyzer.
At 1 kHz and 2 volts output, THD+N was a very low 0.19%.
For comparison, I used a Philips SACD test disc, played from a Denon 2930CI DVD player with Denon Link, so that the signal was sent to the processor as a DSD bitstream, and processed as DSD all the way through, not converted to PCM at some point. For this, I used a different input on the SSP and set it to DSD Direct. Here is a 1 kHz sine wave from the SACD test disc, recorded at 0 dB. It produced an output of 2.67 volts with THD+N at 0.034%. However, notice the noise differences. With the SACD connection, the 60 Hz and its harmonics are not present. In other words, no hum that so often gets into systems with various components connected together.
Back to analog with the CD input again, here is a 10 kHz sine wave at 2 volts output. THD+N was 0.02%.
Our IMD test, using 60 Hz and 2 kHz, 2 volts output, resulted in 0.021% IMD. Notice the AC noise at 120 Hz and above.
Now, compare the graph above with an IMD test taken from the SACD test disc. It uses 60 Hz and 7 kHz because that is an SMPTE standard. I use 2 kHz instead of 7 kHz because I can see the side peaks next to the fundamental more easily. Again, the SACD signal was passed to the SSP via the Denon Link as a pure DSD digital signal. The IMD graph above was through the CD input as an analog signal, CD Direct, no extra signal processing. See how the AC harmonics are not there in the SACD test signal. IMD is slightly less as well, although by a negligible amount.
THD+N vs. Frequency indicated a constant amount of distortion through the audible band, which is good. This means a neutral sound.
THD+N vs. Output shows that somewhere close to 3 volts RMS is the maximum output. An average listening power output with reasonable volume would be about 500 mV on any one channel of the SSP, and a typical power amplifier has an input sensitivity of about 1.5 volts to drive it to full output.
I measured the frequency response with the CD input configured as CD Stereo. This let me turn Audyssey processing on or off (not available when the input is set to CD Direct). It is flat to 20 kHz at 1 volt but the high frequencies begin rolling off above 7 kHz at 2 volts output. I suspect this was deliberate to protect speakers and your ears if you tend to crank things up too much in action movies. The blue graph line shows the response with Audyssey processing turned on. I had run the Audyssey setup with 5.1 speakers being electrostatic, and the resulting graph shows the EQ that has been applied (this is the left front channel). You can see where it rolls off in the bass to crossover to the subwoofer. Audyssey uses filters rather than EQ bands, and their website states that hundreds of points in the spectrum are measured and corrected. The resulting response is a result of the speaker's natural frequency response and also changes induced by the room. I have to say that I love the changes that Audyssey made. It all sounded so much more natural. This is really a terrific feature.
Now to the power amplifier.
At 1 kHz and 20 volts output into 8 ohms, THD+N was 0.005%.
And at 4 ohms, it was 0.01%.
IMD, with 60 Hz and 2 kHz sine wave inputs, 20 volts output into 8 ohms, was 0.006%.
At 4 ohms, IMD rose to 0.011%. With IMD measurements, the calculations are based on the peaks 250 Hz on either side of the fundamental (2 kHz). However, there are other areas that have obvious IMD peaks (4 kHz, 6 kHz, 8 kHz, etc.) This is why it is important to show graphs rather than simply report the IMD number.
For THD+N vs. Frequency, shown below at 20 volts output into either 8 or 4 ohms, the distortion rises above 300 Hz. This will make for a forward sounding midrange and treble.
THD+N vs. Power output indicate that the POA-A1HDCI outputs 175 watts RMS (one channel) before going into a sharp rise to clip (1% THD+N) at 215 watts, while at 4 ohms, it outputs 230 watts RMS before the sharp rise to clipping at 320 watts. Each amplifier module has its own power supply capacitors, so I just measured the output with one channel being driven.