Well, Bruno's article being a technical articles I was assuming that the criterium was a reproduction as free from artifacts and coloration and harmonics and PIM and SID etc as possible. With that criterium, his article stands until other technical arguments refute it. Which, I noted, have not been offered by any of the supposedly heavy hitters here.
I do appreciate that such a criterium is different from, say, listener preference. By the same token, listener preference is just that and cannot be argued with.
jan didden
Now, all kinds of EMI interferences are the highest issue in audio, especially HF interference coupled through power rails.
Fast complementary-differential JFET input circuits with high slew rate and high OLG handle this much better compared to 20V/us single-differential input BJT opamps. That is one of the main reasons why topologies like John Curl's sound so good and are so successful. No THD opamp datasheet plots are going to reveal this.
Fast complementary-differential JFET input circuits with high slew rate and high OLG handle this much better compared to 20V/us single-differential input BJT opamps. That is one of the main reasons why topologies like John Curl's sound so good and are so successful. No THD opamp datasheet plots are going to reveal this.
Because there is a negligible dynamic saturation. Dynamic saturation is an issue even for signals that arrive through supply rails.
Run the bjt opamp heavily into class a. There will be very little garbage on the rails. Modern opamps (VFA) also have extremely high PSRR. One device I am thinking of is > 90db on both plus and minus rails at 20kHz.
Last edited:
Or build it like RF equipment, so RF only goes where it is supposed to go. In the case of audio equipment this is nowhere.
Pavel,
Although I agree with the issue, there are still folks who make products that perform badly enough, that that is not an issue! I know you assume we are talking about well designed modern products, but somebody will go build the ultimate line filters and still hook it up to poor product and then complain you are wrong.
ES
Jan,
Let me finish my questions.
If I have a second amplifier that has the function Vo = 1000 x ( Vin + .05 Vin exp 2) and set the gain to 10 what is the output?
Now if I set amplifier #2 before amplifier #1 each open loop and set the overall gain to 100 what is the output?
That is the gist of my possible article for Vol. 2, but it seems no one understands it!
Yes Ed, I understand. There are at least 3 input paths for any amplifier. I tried to emphasize that some topologies are inherently more immune (if properly designed) than other. The complete lack of dynamic saturation issue is a perfect starting point.
You are referring to the "concept amp" which is still in the pipeline for
release this fall. The prototype at that time was 0fb, but after reviewing the
results of our work with new devices this past year, the subjective
performance bar has been raised, and we are still working out the details.
😎
You keep asking us to do messy algebra, but have not told us why. Therefore it seems premature to conclude that we don't understand. Tell us what it is that you believe you understand, and we (allegedly) do not. We have already established that there will be an infinite series. The significance of the higher order terms depends entirely on the signal level, which you have not told us.simon7000 said:
Listener perception is everything where I come from. No distortion
analyzer has ever purchased one of my amplifiers.
I recall Bruno's article as being a nice overview of his thinking - obviously
I must have missed a proof as profound as you describe. As I have the
disadvantage of being at the coast, you will have to wait until I get back
for my response, unless you would care to email me another copy of the
piece.
😎
Right on, Nelson. His 'history' could use a little correction as well. IF he had asked, one of us, either Walt or me, could have brought him back to the right track. There is nothing new here. PIM is what I look for, not TIM, we solved the TIM problem over 35 years ago.
Thanks for the info. Always look forward to your new designs.
If I weren't a complementary-symmetry kind of guy, it would be interesting to play around with the EPC devices. These are lateral parts with extremely high transconductance -- the best of both worlds. Too bad they can't use their process to make a P-channel part...
Damn cheapskate, tire-kicking distortion analyzers! When I see one come into my store, I throw it right out! I'm not going to waste my time with those looky-loos...
For signal bursts like 16MHz or 32MHz that come through PSU? You had no issues like this 35 yrs ago.
- Status
- Not open for further replies.