Interesting work Kier - many thanks. I think with my ancient cloth ears I'd struggle to hear any differences and any differences I did hear would likely be due to the performances being different, as others have said.
I think to compare them "formally" side-by-side I'd do the following:
- Use the performance with the reference sinewave (as you have done) and look at the /spectrum/ of what comes out of both Montage's. It's likely not enough to just do an FFT of an oscilloscope trace, as 'scopes are only about 10-12 bits and you need more than that. You'd likely need a spectrum analyser (expensive) or a really good data acquisition board with good ADCs on it and a low noise floor, and then post-process the data afterwards. I'd be looking for harmonics of the sinewave, indicating distortion. You could then calculate things like SNR, THD, etc, and compare the two Montage's.
- Then do the same, but sweep that sinewave to compare the frequency response of both Montage's. Out of interest, you could also see if the noise and distortion levels (SNR, THD) change with frequency.
- Then setup a performance with two sinewaves close in frequency and look for a a 3rd peak in the spectrum between the two, indicative of 3rd-order intermodulation distortion (due to nonlinearity)
I'd bet my car on their being a difference between the two. And I'd bet my house on there being significant unit to unit variation of the same type of Montage as well. The question is, which is greater!?
I think to compare them "formally" side-by-side I'd do the following:
- Use the performance with the reference sinewave (as you have done) and look at the /spectrum/ of what comes out of both Montage's. It's likely not enough to just do an FFT of an oscilloscope trace, as 'scopes are only about 10-12 bits and you need more than that. You'd likely need a spectrum analyser (expensive) or a really good data acquisition board with good ADCs on it and a low noise floor, and then post-process the data afterwards. I'd be looking for harmonics of the sinewave, indicating distortion. You could then calculate things like SNR, THD, etc, and compare the two Montage's.
- Then do the same, but sweep that sinewave to compare the frequency response of both Montage's. Out of interest, you could also see if the noise and distortion levels (SNR, THD) change with frequency.
- Then setup a performance with two sinewaves close in frequency and look for a a 3rd peak in the spectrum between the two, indicative of 3rd-order intermodulation distortion (due to nonlinearity)
I'd bet my car on their being a difference between the two. And I'd bet my house on there being significant unit to unit variation of the same type of Montage as well. The question is, which is greater!?
Statistics: Posted by Davelet — Thu Oct 03, 2024 8:29 pm