Multiple distortion measurements were taken before I discovered the Power Step feature of the LAUD 2.2 software. (Read the manual first, it helps !) This allows to perform a automated measurement of harmonic distortion versus SPL. The software steps the output to the amplifier down while trying to keep the signal to noise ratio constant by stepping the input sensitivity on the microphone up. So proper setting of microphone level and DSP card output level is advised here.
The measurements below were performed nearfield with the microphone positioned 1.5 inch from the ribbon. I felt that this was the best way to minimize the interaction signal/reflected sound at any given power level. This measurement really asks for an anechoic room and lots of power, maybe later ;-)
The two curves below show SPL level as measured @ the microphone (1.5" from the ribbon) at the top of the display. There are five dashed lines in total. The top of the display is referenced at 120 dB SPL. Remember, this is what the microphone sees at this particular distance at an unspecified amplifier level.. The solid lines in the bottom half of the display are the measured THD levels. The frequency axis of the display runs from 50 Hz to 20 kHz. The sample rate poses restrictions on the number of harmonics that can be measured at any given frequency. The second harmonic falls away at 10 kHz, the third at approximately 7 kHz etc. (Nyquist again.)
The distortion level can be seen peeking at different frequencies. Since I performed this test without notch filters installed, a really sharp peak can be seen at 5.5 kHz. This peak doesn't show up with notch filter installed. Neither does it show up at lower drive levels as witnessed in the lower trace. The bump @ 125 Hz in the SPL curves is in my opinion caused by the low ratio "width of the driver/distance microphone-to-driver".
Although the SPL produced by the driver is higher, the overall harmonic distortion is lower over the workable frequency range. (From 200 Hz on. The distortion peaks at different frequencies. Interesting to see is the fact that even without the notch filter there's only a small dip at the 7 kHz resonance peek. We see a similar bump in the SPL display at 200 Hz, probably because of the near-field measurement.
Without performing some "real" high power tests in an anechoic room it is hard to formulate any final "conclusions". It does however seem to me that the extended low of the RD-75's comes with an increased harmonic distortion level in that extended low. I assume that this is caused by the difference in the materials used and more specifically the tensioning of the ribbons. A higher tension dampens resonances better and provides a more critically damped roll-off at the extremes. This comparison will definitely be repeated at higher power levels with both ribbons equalised to verify if this line of reasoning can be sustained.