Negative feedback in audio
HTML-код
- Опубликовано: 23 июл 2024
- What is negative feedback? How does it work and what does it do the sound quality of audio, especially in a high end system? Learn the inner working of negative feedback in language you'll understand.
I am getting close to publishing my memoir! It's called 99% True and it is chock full of adventures, debauchery, struggles, heartwarming stories, triumphs and failures, great belly laughs, and a peek inside the high-end audio industry you've never known before.
I plan a few surprises for early adopters, so go to www.paulmcgowan.com and add your name to the list of interested readers. There's an entire gallery of never before seen photos too. 
Наука
![Sean Rii, Karyon, Sharzkii - Taungule [My Love] (Official Music Video)](http://i.ytimg.com/vi/eJoMuypbSzQ/mqdefault.jpg)
There's something addictive in the way you explain the subject.
I'm sure you would've been an excellent teacher. Thanks Paul for sharing the knowledge.
You teach way better than my EE profesor. Seriously lol
Paul you would have been a great electronics teacher
You definitely know how to explain. Thank you for sharing. It would be fantastic if you continued this explanation by explaining more about the components of an amplifier. Have you thought about explaining a complete project, like Douglas Self for example?
Very helpful explanation. Thanks Paul, I think I finally understand this now.
Awesome! Thank you! I love this.
Thanks for your knowledge.
Good Information Paul. I learned something new today.
Best of Paul: 0:09 3:16 5:06 5:14 5:35
Ha Ha Ha Ha Ha!!! Completely agree! Paul is an excellent teacher, story teller and a clown! Love every one of those moments! :) :) :)
Love the white board and explaining the physics of electronics...love it.
I've often wondered what a negative feedback is and what it does... thanks for the question and your excellent explanation Paul. Hope there's a part 2 on this subject where you can expand a little bit more without going into a full course on this topic.
Thank a lot for going to your white board for today's explanation of today's question.... PLEASE do it more often.... And have a good day..
This is a great video; thank you. Originally, I though feed back was to amplify the signal twice.
Actually you can tell by implementing a separate digital model in a separate part with MCU. That is calibration data, some ADCs and sensors to adjust the working of the amplifier.
Do you add backdoors to your system software? If so an you detail why?
Thanks for explaining the NFL for the layman. I never quite understood how this works in Class AB amps.
Paul, this tutorial was great. Best way in my opinion, to make yourself new customers. Like Nelson Pass approach, in witch he ''gives'' to the DIY community and, tutorials too. Gives me (us), confidence in your products.
The gain on a non inverting OP amp circuit is the feedback resistance divided by the input resistance plus 1.
Great info. We are still waiting in the "Damping factor video"
Hi Paul looking forward to the op-amp video's !*_*! i would like to learn as much possible , want to start making effect pedals ,, I have a Rockman XPR and that is shire genius full with op-amp don't know how close i can come to understanding the many ways what op-amp can give in effect design . Thx in advance and greetings
Tom Scholz and his Rockman gear were a big inspiration for me to become an electrical engineer back when I was a kid! The "effect" sounds typically don't come from the op-amps themselves, but from how you use them with other components. (unless you just overdrive them for fuzz). The Rockman signal chain goes something like this - Input stage, AGC compression stage, noise gate stage, clipping stage (silicon diode soft-clipping in the opamp's feedback loop OR LED hardclipping on the opamp's output, cabinet simulation filter stage, then the BBD delays he uses for Chrus, Reverb, etc before the output. I agree, his methods were definitely genius, though, looking at the schematics of the Rockman gear now... I see certain big design no-no's that he was able to get away with because of how forgiving and relatively slow the TL072 op-amps are. It's no wonder when folks try to replace them with modern "hifi" op-amps they have major problems.
Hi 0x07AF that's nice to hear and super interesting to me
,,Hope to get your level of understanding how the components interact with each other !*_*! ,
,What motivate me now for gtr effects making, is i get the idea that most of the effects that are around seam to be made without musician input ,,,
Tom his approach seams different , and looking into the inside of the XPR i got that idea like how you explained .
Bought my XPR when it was new and blow my mine up, four months ago ,with volt on the effect send .
There is no Schematics around from the XPR (and i would like to try if its possible to make one !)
so i bought (luckily) an other one a Rockman XPRa (very expansive ,, now my original XPR is sort of partly fixed , 3 IC's are replaced , one of them was a volt divider CPU style..
But it sounds a bit like its over shouting itself , (compared with the XPRa )and the level from the echo /Revb does nothing ,, its always full
So there is still a mistake in it ,
but having them both , so i can see the different components that make from a XPR a XPRa .
Would you like to connect with me 0x07AF
Thanks Paul. You should do more presentations at the white board. Takes me back to my graduate days being taught control system theory by a very enthusiastic lecturer, Dr 'Doc' Aughtie. We all liked him enormously for his knowledge and presentation skills . You would have made a fine teacher I think. Amplifier design theory presentations would be wonderful, as long as they don't detract from finishing that book...
Excelent!
Thankyou sir
So a real time comparison is impossible so the leading edge of the output signal will be compromised? Will that translate into an veiled attack of the musical notes to some degree?
Thanks
Omg this guy made me laugh so hard
Lol i need Paul to prob my amp lol
Hm, so negative does not mean it is bad as I thought but it means something different ,right? It comes from fact that it is attached to negative lug in input of the amp ? THank you for correcting me :) Best
Please make an affordable PS Audio amplifier kit!
Thanks for the video!
It would be sweet! Paul said PS Audio used to make some kits, but sadly they didn't sell well at all. :-(
Yeah, saw it.
Paul, If there was a kit - maybe you can resell it?
no R&D needed
Great stuff, what would you call someone who designs and builds amplifiers? Electrical Engineer ??
In the real world it would be an electrical engineer but my observation over decades of time is that in the High End Audio business you can find people that are not engineers of any kid designing and selling amps, speakers and cables. I think that this is one of the reasons that there is so much hype, misinformation and bull in the High End Audio industry. I have personally met many of the "stars" of high end audio over the years and I was surprised at how many really do not know and understand the details of the inner workings of their products, which leads me to believe that many of them do not really do any of the critical work on their own products, which again could explain why there is so much hype and not so much truth in the High End Audio business. Of course I did meet a few folks that really did know their stuff and who seemed to be able to actually be able to do the research, experimentation, design, and build of their products, but there seemed to be very few of them.
In this video, there was no explanation (that I noticed) of why negative feedback (in any form) is secretly (or not so secretly) hated in High End Audio. The use of negative feedback does stabilize amplifiers, that is true, but is not mandatory its just an inexpensive and easy way to do so. You can also get amazingly low levels of distortion by pumping in more and more various types of negative feedback, which seems to be how some amplifier manufacturers are able to claim that their amps produce something like 0.0009% THD, which is certainly possible. What is nonsense is that you need to have distortion levels that low, when speakers commonly have distortion levels of somewhere around 15%. Obviously the amplifier's distortion is such a tiny fraction of the speaker's distortion that it blends together and becomes inseparable and irrelevant. Even if an amplifier had 2% distortion or even 5% distortion, it seems logical that you could not hear it over the speakers' 15% distortion. The importance here is that negative feedback alters the output signal so that it no longer is an exact (as exact as it can be) duplicate of the input signal, at a higher power level. With more negative feedback, the subtlety and nuance that customers pay enormous amounts of money for is eroded or erased, so why bother paying all that money for them? It seems that the people who market High End Audio products are better at selling them than the manufactures are at actually building them.
An interesting experiment is to listen to an amplifier that claimed to have no negative feedback at all, like the classic Rappaprt AMP-1, and compare it to a conventional amplifier that has plenty of negative feedback. I have done exactly that on many occasions and was surprised that I could hear a clear difference, the AMP-1 sounded far more natural and realistic where all the other amps sounded like they all had a lot less detail (in different ways, but they all sounded like they had muted details).Unfortunately Rappaport had to shut down his company before I had he chance to save up enough to buy a pair, so that I could continue my experiments in comparing them with amplifiers that had a lot of negative feedback. The tradeoff that Andy Rappaport made was to bias his amp so that it was mostly Class A and so it gave off tremendous amounts of heat, huge amounts of heat. Standing next to them I was shocked at how much heat a 75 watt amp could give off, and it was huge, having massive heat sinks which obviously seemed to be to dissipate all that heat.
So the question that this video totally sidesteps is "What tradeoff in sound quality does negative feedback impact us with", "Does various level of negative impact create different sonic signatures", "What is the sound quality of negative feedback". These questions are the ones that someone shopping for high end audio gear might want to know the answers to.
It is not the feedback itself which causes audible issues. It is the reliance on feedback to bring distortion levels down, in cases where the amplifier operating open-loop (without feedback) would have high distortion.
The public is being deceived to a great extent, concerning power amplifier distortion, because amplifiers are always tested with non-reactive loads when it comes to spec'ing them. It is only when reactive loads are used, that one sees an amplifier's true distortion character.
For example, well designed class AB amplifiers will do a fine job of minimizing polarity-crossover notch distortion when the switching occurs near 0V, but often not nearly as well with a reactive load which forces the crossover to occur well away from 0V. Harmonic distortion from switching at the crossover point is a highly audible type, especially when the musical signal is spectrally simple and does not have much high-frequency content. In general, it's a mistake to consider any certain distortion level as the threshold of audibility, as it depends entirely on the spectral distribution of the distortion, as well as the character of the musical signal.
Reactive loads can also cause issues for amplifiers which do not have generous stability margins (determined primarily by the compensation and feedback design), resulting in spurious frequency response when driving real speaker loads.
Both of the above, are important reasons why various power amps can sound so different when driving the same speakers. This is a very complex topic, and engineers will normally resort to numerical simulation, to perform the analyses that are needed.
Thank you, well said!
>5:15 DONE AHHHHHHHHHHHRGHHAAAA???? OR? LOL you're a really good teacher love your videos and i am learning new things on your channel. keep up the good work AWESOME content
what did you put in your granola today?
Show some uses for op amps inverting non inverting
So, I am now confused!
Accepting the explanation you offer Paul, does it mean amplifiers that claim 'no negative feedback ' (there's several in high end audio), are clipping and "out of control " as you put it?
I was led to believe, amplifiers with negative feedback, generally speaking, take the 'life' out of the music. Is that claim wrong?
Amplifiers that claim to have no negative feedback are probably ones that have no global negative feedback. Global means the feedback encompasses the entire amplifier circuit (like Paul showed). Amplifiers with no global negative feedback would have to be using lots of local negative feedback. Local feedback just means feedback is around parts of the circuit inside that make up the amplifier. In truth, even a global feedback amp will use some local feedback to help reduce distortion.
When an amplifier is designed, it should be designed to minimize distortion in each section before global negative feedback is applied to get the distortion extremely low. Relying solely on global negative feedback to fix the problems is a poor design in my opinion.
Global negative feedback lowers distortion, lowers noise and makes the damping factor high. Stealing the life from the music is audio fool nonsense although some speakers may sound "better" with low damping factors. Also some people may like the sound of added low order harmonics such as with a single ended tube amplifier. Personally, I don't want the amplifier messing with the sound.
Excellent! Very informative response to my remark, thank You!
What confused me was the fact that Paul didn't differentiate between 'global' and 'local' feedback. Several high end manufacturers advertise "no negative feedback" as an asset in the design of their product, leading many to believe that "negative feedback" is something best avoided.
Your comment about low order harmonics (mostly second order), is indeed one of the attractions for tube lovers, especially single ended designs.
Each to each own!
Leporello Gatos - If you're interested in learning about audio circuits in practical detail, the fellow JohnAudioTech who replied to your comment here makes some of the best RUclips tutorials on the subject, and he's made a lot of them. Highly recommend!
EDIT: Uncle Doug is another great channel for mostly tube circuits and amplifiers. JohnAudioTech does some tube stuff, but mostly solid state.
Thanks for that OxO7AF!
I will follow your suggestion and check these two RUclips channels. It's always interesting to hear / see various techniques applied to audio circuit design.
LeRoy, Ny. Ah, Frost Ridge campground. Old fashion ski hill.
We like mini electronic courses from Mr. Paul
I'm civil engineer , but if the lecture is Paul, I don't mind to take another degree in electronics ...
Oh to be a fly on the wall in that room...
"It was my understanding there would be no math."
I prefer you looking on a scope in my system
So why do zero feedback amps work and sound good if they're so out of control?
JB Berg it's because the specific stage of amplification has a low open loop gain. In those cases you can get away without using negative feedback. Although, you might lose some of the other benefits of negative feedback besides gain control.
0x07AF Ok, thanks, but what are those other benefits, lower distortion? In guitar amps low nfb means more break-up, which is good, but my Densen transistor hifi amps have zero feedback and I can’t detect any distortion. Also prefer my tube hifi amp in the 0db feedback setting. 6 and 10 db sound muffled and the volume drops too much.
Noise reduction, bandwidth control, stability over temperature (less fluctuation over temperature, device ageing, etc.), stability in general, output load compensation (dampening factor) which keeps the output level correct instead of sagging with higher load currents, and as you said, lower distortion... to say, 'low negative feedback in a guitar amp causes more break-up which is good'... a properly designed guitar amplifier would use negative feedback to give the user more *predictable and repeatable* control over when the break-up happens. If you have a system that sounds bad when negative feedback options are enabled, I'm guessing the feedback scheme wasn't very well designed. Depending on stage configuration, individual components like transistors and tubes provide elements of negative feedback between their own nodes... ie, when you drive a current into the base of a BJT, the resistance on the emitter is fed back "felt" by the base in how it affects how much current is fed into the base for a certain gain. Often times, you'll think your amplifier has no feedback, when in fact, it's probably happening in your circuit in ways you're not aware of. There is local (stage-by-stage) feedback, and global (final output fed all the way back to the input of the amplifier). You're probably not using global feedback, but I'm betting you've got forms of local feedback in your amplifier stages which are keeping it from being noisy or oscillating.
0x07AF So that’s why a hot amp sounds better :) The Densens are an always-on design, sound sterile when just turned on. In a guitar amp, break-up and feedback are controlled at the guitar by its volume pot and with playing dynamics. Some early Fender amps have no feedback and they can be fuzzy sounding. Marshalls vary depending on age and model. Less nfb has more power amp distortion and softer bass.
Not sure is my tube hifi has any local feeback, it’s a pretty simple desgin.
The 300B vacuum tube is probably the last audio amplificaton device built to work without any need for negative feedback.
Aahaha! sloa011.pdf of www.ti.com appears again. Try to read it. Texas Instruments is very charity.
Cheers from Indonesia
A = (R1+R2)/R1, not R2/R1 so it your examples 11 and 21. Sorry to be to be a pain.
Airgead The example was considering R2>>R1. Then R2/R1is a good approximation :)
Folks, Read Billy Bob's reply below. It's very good info except I believe he is wrong in terms of speaker distortion. Modern speakers are much better than that I believe. Also, I am not sure what the number is today, but humans 30 to 40 years ago could not hear less than .5% THD in electronics.
Negative feedback is the work of the devil. I insist on 2 things, no global negative feedback and pure class A. Simple really.
f-bombs make me feel at home! :D
This time lag is not audible. A class A amp without negative feedback does not sound any better. Negative feedback has only advantages. It is the best invention in amp construction apart from Clasa D up to date. It makes amps efficient and effective at the same time. No reason not to implement it technically speaking.
I think a lot of folks confuse time lag with phase. When the phase is shifted enough that feedback is turning positive I've heard so many variations of, "that damn time lag turned my amplifier into an oscillator again!! Feedback sucks!!" lol
"Negative feedback has only advantages."
Not so. Back in the late 60s and early 70s amp designers were pursuing lower and lower levels of Total Harmonic Distortion (THD) with copious amounts of negative feedback. Then Jim Sugden released his A21 amp with ten or a hundred times as much THD as most high end amps. The A21 was obviously better sounding even to those with cloth ears. Negative feedback is good, but more than necessary is deleterious to accuracy.
I don't think there is disadvantage of negative feedback.