Yo should really read Bruno´s article on the F word to better understand ´slow´ and ´fast´ in this context.
Especially the chapter Open loop bandwidth as a measure of speed starting on page 7.
https://linearaudio.net/sites/linearaudio.net/files/volume1bp.pdf
Jan
Last edited:
Delay under 100nsec, preferable under 50nsec, In-OUT with a voltage step on input (but under lever of cliping output)....let's say.
For ME an amplifiers shoud have low pas filtre at the input and output with a bandwidth not more over audio, let's say 100khz BUT the internal speed high, low delay and capable of full power bandwidth of over 5Mhz.
On Sara amplifier (you can find it here on forum) I measured over 1Mhz full power bandwidth but I want more because I want sr of over 250V/usec not "only" 80V/usec like în Sara.
On Sara amplifier (you can find it here on forum) I measured over 1Mhz full power bandwidth but I want more because I want sr of over 250V/usec not "only" 80V/usec like în Sara.
> Most audio circuits have an open loop bandwidth which is much smaller than audio bandwidth. They work fine.
The gain may be rolling-off by 100Hz or lower.
However the DC gain may be a million. 100Hz high-cut, we still have gain of 10,000 at 10KHz. Say closed-loop gain is 20. We have 500:1 of NFB at 10KHz, 250:1 at 20KHz. Any error is about 0.5% of the naked amplifier. If the "error" is 10% THD, this becomes 0.04% THD. While not zero, it is very-very low.
Many authors suggest that the LF corner of the open-loop gain is not very important. What matters is that you have a gob of gain all over the audio band. And in practice, we specifically set the open-loop audio gain (with compensation) and let the DC and LF gain be what it is.
OTOH, the inevitable declining NFB with rising freq does lead to rising THD, which "may" be more offensive than a flat THD plot. (This path is well-trodden by low-NFB fans.)
The gain may be rolling-off by 100Hz or lower.
However the DC gain may be a million. 100Hz high-cut, we still have gain of 10,000 at 10KHz. Say closed-loop gain is 20. We have 500:1 of NFB at 10KHz, 250:1 at 20KHz. Any error is about 0.5% of the naked amplifier. If the "error" is 10% THD, this becomes 0.04% THD. While not zero, it is very-very low.
Many authors suggest that the LF corner of the open-loop gain is not very important. What matters is that you have a gob of gain all over the audio band. And in practice, we specifically set the open-loop audio gain (with compensation) and let the DC and LF gain be what it is.
OTOH, the inevitable declining NFB with rising freq does lead to rising THD, which "may" be more offensive than a flat THD plot. (This path is well-trodden by low-NFB fans.)
There are 2 types of Anti-NFB fans:
1.) Who does not know how to use NFB properly. They spread the myth that any NFB is bad, except may be some "low feedback" that they are not afraid of.
2.) Who know nothing, and trust #1. I call them "Cargo-Culturists"
Also, there are 2 types of FB fans, who:
1) don't understand that not only THD matters, but also how sound different misbehave-conditions. They tend to apply as much as possible of global only feedback.
2) who understands how feedback works, and pays attention to minimise all audible distortions, searching for optimal results. They (we, actually) prefer nested feedbacks, that can get both low IMD before clipping like "as much as possible global feedback", but as less as possible audible distortions when an amp grossly misbehaves, like "no feedback at all" amps.
1.) Who does not know how to use NFB properly. They spread the myth that any NFB is bad, except may be some "low feedback" that they are not afraid of.
2.) Who know nothing, and trust #1. I call them "Cargo-Culturists"
Also, there are 2 types of FB fans, who:
1) don't understand that not only THD matters, but also how sound different misbehave-conditions. They tend to apply as much as possible of global only feedback.
2) who understands how feedback works, and pays attention to minimise all audible distortions, searching for optimal results. They (we, actually) prefer nested feedbacks, that can get both low IMD before clipping like "as much as possible global feedback", but as less as possible audible distortions when an amp grossly misbehaves, like "no feedback at all" amps.
Use CFA topology. Dadod, Bonsai, Lazy Cat made some here and I did, but I do not share it here.
And WHO on Earth would use an Op Amp, ANY Op Amp, open loop?
I've used them as a comparator ;-)
Jan
I wondered if that might be the case. You may need to review your understanding of feedback.sesebe said:
So you accept that your preference for wide bandwidth and high slew rate is not related to actual performance, but is merely cosmetic? That is fine, but newbies reading your remarks might be confused by them.
Yes.Wavebourn said:
I would not say it is cosmetic but more a taste problem and personal preference based on audio experiences.
I liked the sound of a big SR amplifier with compared to my Yamaha that has an SR of 100V / usec or close.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.