It is posted in:
Forums/Amplifiers/solid state
not in:
Forums/source and line/Analog Line Level
😕
Forums/Amplifiers/solid state
not in:
Forums/source and line/Analog Line Level
😕
I think this topic is great for exploring how we reason things out. Class D raises a dichotomy, doesn't it? Between extreme NFB applied at 20kHz (good) and its switching design with bandwidth one-tenth to one-hundredth that of a linear amp (bad). The latter has considerably less publicity and yet the pursuit of bandwidth has been the focus of NFB design since I can remember.
Is there a false premise here? You decide.
Is there a false premise here? You decide.
Hello
I use global feedback but I design my amp to have a negative global feedback arround 30 db. So I don't do amp with high open loop voltage gain.
I agree with Hugh (aksa), it's the distortions spectrum that are important, not the lowest distortion numbers.
Bye
Gaetan
I use global feedback but I design my amp to have a negative global feedback arround 30 db. So I don't do amp with high open loop voltage gain.
I agree with Hugh (aksa), it's the distortions spectrum that are important, not the lowest distortion numbers.
Bye
Gaetan
Armed with this insight, all we need is a break to figure out how to achieve low spectral distortion.
Yes. My 1 year arrival anniversary was last Friday.
I’m back home after many decades. 🥰
Less GNFB (loop gain) make components variance more impact to the performance. Maybe need components matching and precision components.
sser2,
insightfully put, except for the fact that motional feedback is indeed highly efficient in reducing cone resonances at low (sub-bass) frequencies. It is not just a matter of diaphragm rigidity. You should urgently revise your current stance.[/QUOTE]
What should I revise? I see no daylight between our positions. I posted that MFB is utopia at frequencies above the lowest. I never said that it is bad in subwoofers.
I wholeheartedly agree with you that NFB is not necessary in audio amplifiers. Folks who peddle unquestional benefits of NFB live in the wonderland of theoretical infinite gain bandwidth circuits allowing infinite amounts of NFB.
insightfully put, except for the fact that motional feedback is indeed highly efficient in reducing cone resonances at low (sub-bass) frequencies. It is not just a matter of diaphragm rigidity. You should urgently revise your current stance.[/QUOTE]
What should I revise? I see no daylight between our positions. I posted that MFB is utopia at frequencies above the lowest. I never said that it is bad in subwoofers.
I wholeheartedly agree with you that NFB is not necessary in audio amplifiers. Folks who peddle unquestional benefits of NFB live in the wonderland of theoretical infinite gain bandwidth circuits allowing infinite amounts of NFB.
I haven't read this thread, just used the search function, but it appears that no-one has mentioned Daugherty and Greiner's flawed article from 1966 yet: "Some design objectives for audio power amplifiers", IEEE Transactions on Audio and Electroacoustics, vol. AU-14, number 1, March 1966, pages 43...48. See Some design objectives for audio power amplifiers - IEEE Journals & Magazine for the article or https://linearaudio.net/sites/linearaudio.net/files/volume1ltemvdg.pdf for a short, freely accessible summary and explanation what is flawed about it.
Excellent example of local negative feedback. Can you also apply global negative feedback?
> so please put up a power amp circuit ...
The difficulty with a power amp not using any NFB is not distortion.
It is output impedance, or damping factor.
It is very difficult, if not impossible to achieve low Zout without using very high bias.
Hence, a small amount of NFB helps with that.
There are at least two power follower known to not use any NFB, namely the F4, and the Beast.
FIRST WATT
http://www.firstwatt.com/pdf/art_beast.pdf
> hijacking the thread with your marketing
Well, you asked me to share yourself, "Care to share? I guess not.."
So I obliged.
And since I am not offering anything right now, there is no marketing need. 😉
Cheers,
Patrick
The difficulty with a power amp not using any NFB is not distortion.
It is output impedance, or damping factor.
It is very difficult, if not impossible to achieve low Zout without using very high bias.
Hence, a small amount of NFB helps with that.
There are at least two power follower known to not use any NFB, namely the F4, and the Beast.
FIRST WATT
http://www.firstwatt.com/pdf/art_beast.pdf
> hijacking the thread with your marketing
Well, you asked me to share yourself, "Care to share? I guess not.."
So I obliged.
And since I am not offering anything right now, there is no marketing need. 😉
Cheers,
Patrick
‘Infinite gain bandwidth’ and ‘infinite amounts of NFB’ ?
This is so disingenuous. Really it is.
I can’t speak for other designers here, but I’ve spent many hours on LTspice investigating feedback of all kinds to develop my knowledge practically and plenty of reading on the subject to get to the point where I can use feedback confidently in the stuff I design and build.
To use feedback correctly, you have to invest a few thousand hours in learning about the subject. If you think you can slap an amp together and simply put feedback around it to clean it up, you’re in for a surprise. No competent amplifier designers ever do this. And that’s exactly the issue. The anti- feedback crowd don’t want to invest the time and effort so the default position is to dis it because they don’t understand it or won’t make the effort.
How about positive feedback? Shouldn't it be the antithesis of NFB, increasing distortion and deteriorating all things that NFB improves. Meanwhile, John Miller (of Miller capacitance fame) patented positive feedback inside global NFB loop that numerically negates NFB. The amp has less distortion than equivalent circuit without PFB. I know a commercial implementation of this design, the sound is far superior to the NFB-only topology.
Another sign of lacking understanding. The point is that the (local) pfb is used to increase the gain, which you then can translate to loop gain for the feedback to make the amp more linear.
With the advent of cheap, high gain (solid state) devices, it is now very easy to pack enough gain so that this type of tactics are not required.
https://linearaudio.nl/sites/linearaudio.net/files/miller combined fb electronics march 1950.pdf
🙄
Jan
Last edited: