- Written by John E. Johnson, Jr.
- Published on 24 November 2011
The MD-75 Integrated Amplifier On the Bench
All distortion measurements were within an 80 kHz bandwidth. Both channels were driven, with an 8 ohm load. I did not run tests at the 4 ohm tap since this simply limits the voltage, and I did not run tests on the power amplifier by itself (there is a set of RCA pre-in jacks to use just the power amplifier section).
Each distortion graph will begin with the MD-75 in triode Mode, followed by the same measurement with the amplifier in ultra-linear mode.
At 5 volts and a 1 kHz sine wave, distortion was less than 0.1% in triode mode.
In ultra-linear mode, distortion dropped to less than 0.05%. In both cases, the second order harmonic was predominant. Also, in ultra-linear mode, the difference between the height of the second order harmonic and third was greater.
In triode mode, 1 kHz and 15 volts output, distortion was less than 0.5%, which is a reasonable number for a tube amplifier.
In ultra-linear mode, the 15 volt distortion measurement was not significantly different than it was in triode mode.
In triode mode, maximum output before clipping was 40 watts, but in ultra-linear mode, it was 75 watts. So, I am only showing the 20 volt (50 watts into 8 ohms) measurement in ultra-linear mode. The result was less than 0.5% THD+N, a good number.
Looking at the spectrum when using 19 kHz and 20 kHz sine waves as the test signal is always interesting. Here is the graph with 5 volts output in triode mode. The B-A (20 kHz minus 19 kHz) peak at 1 kHz was 38 dB below the fundamentals. Notice also the IM peaks surrounding the fundamental peaks.
In ultra-linear mode, the B-A peak was 55 dB below the fundamental (the lower the peak, the better). So here, we see an advantage in the ultra-linear mode. Also, there are 8 IM peaks on either side of the fundamental peaks in triode mode, but 7 peaks in ultra-linear mode. These are very small details, but can add up to discernible differences in the sound.
At 20 volts (50 watts), ultra-linear mode, the spectrum is filled with IM peaks. This is a point when the midrange would begin to start sounding congested. However, most of the time, when listening to music, we are only using a few watts.
Using 60 Hz and 7 kHz as test signals, there was less than 0.4% IMD in triode mode at 5 volts output.
In ultra-linear mode, IMD dropped to less than 0.2%, which is excellent.
At 15 volts output, IMD was 3.4% in triode mode, which would result in some congestion in the midrange sound.
However, in ultra-linear mode, MID was 1.28%, which is much more reasonable. (I know what you solid state guys are thinking, but you have to put on a different hat when dealing with tube amps).
Even at 20 volts, the IMD in ultra-linear mode was less than it was in triode mode at 15 volts.
The frequency response in triode mode was down almost 2 dB at 20 kHz. This produces a very smooth sound that tube-o-philes love.
In ultra-linear mode, the response was the same, down 2 dB at 20 kHz.
THD+N vs. Frequency for triode mode is shown below. It's reasonably flat from 20 Hz up to 10 kHz, then rises up to 30 kHz before declining.
In general, ultra-linear mode showed less distortion in the audible band for the various output voltages tested.
Power output in triode mode indicated a sharp knee at 35 watts, and clipping (1% THD+N) at 40 watts.
The sharp knee in ultra-linear mode was at 70 watts, and clipping at 75 watts, exactly on spec.