The role of the op amp in the composite implementation including TDA7293 is extensively discussed here - https://www.kaltecs.com/tda7293-composite-amplifier/
As far as I remember the guy says that "below certain THD+noise level there is no difference in the sound no matter what op amp is used". Actually the same is stated by Douglas Self in Audio Power Amplifier Design Handbook.
So bellow certain THD+nose level (< -118.0dB) there is no such thing as "sound of the op amp".
I only did the tests that I could do on a normal oscilloscope. I don't have a good sound card, so I ordered quantum asylum. All necessary performance measurements will be made and shared. The measurement data I can provide you right now are on the first page.
Attachments
Definitely you right! After a lot of LM3886 applications, i feel more performans and musical sound….
I can't quite agree with such a statement.
According to my experience when both TDA7293 and LM3886 are implemented in a ultra low distortion composite amp architecture they sound almost identical. For example amps properly build using
Kaltec's Ultra-low distortion kits - https://www.kaltecs.com/60w-ultra-low-distortion-kit/
or
Modulus kits - https://neurochrome.com/products/modulus-86
sounds very similar.
7 out of 10 listeners can't spot a difference in a short time (2 min) blind test. Both amps are really sonically neutral.
The difference is perceivable after long term listening (20 min) though, as TDA7293 based kits showed much less listening fatigue. 6 out of 10 listeners noted that.
My listening experiences, the analytical and cold sound of modulus, its weaker state than the dynamism we call torque, turned me off from listening to music... Among the applications I made with LM3886, I got the best sound with ne5532. For me, the LM3886 and the designs made with it were good on paper, but not good for me when listening. The jfet balanced input composite feedback tda7293 application is definitely more successful than its competitors in listening tests… measurements will be shared very soon…
Quantum asylum system is statistically reliable if you're measuring THD+noise > -113dB. For lower THD+noise the error becomes too high and obscures the results. For the fraction of the price I build analog measuring system (THD+noise < -135dB) using as a final step an oscilloscope with FFT.
The THD+noise performance is widely accepted as a reliable objective mark for evaluating the merit of a given amplifier. If you disagree with this, then what is your point of reference?
As I mentioned before both amplifiers (Kaltec's and Modulus) are utterly sonically neutral. This is a subjective opinion based on comparing the experience in the Walt Disney Concert Hall and listening to both amps. Eight of the participants in the blind tests confirmed this as well.
.
You can use any scope with FFT and 8-bit resolution or more. The heavy lifting is done in the analog part. I'll start a new thread soon, describing the project.