What is wrong with negative feedback?

Negative feedback falls into the same category as damping factor both which alot of people dont understand including myself to a point,my counterpoint amp has a damping factor of i think 70 while the great or my bias is NOT GREAT digital amps go on and on about the high amounts of damping factor they trump on their stats.,My Counterpoint has plenty of bass ,it just has to be on the recording in the first place.

Kirkus you know who Atmasphere is right?

If a person cares to look in any book covering filter theory, they'll find gain/phase graphs that illustrate propagation or group delay. For lowpass filters, which is what most amplifers are classified as, low frequencies have little or even no delay while higher frequencies have more, such as the nominal 45 degree phase lag at the -3db point. A phase lag corresponds to a delay.

The model you are proposing relies on propagation time being mutable, which it certainly is not.

Atmasphere, forgive me if I'm being a snot . . . but I think you need to brush up on some basic electrical theory. Pole/zero networks do indeed have different delays based on frequency. If you don't believe me, try constructing a simple R-C lowpass network with, say, a .47uF capacitor and a 750 ohm resistor. Compare the "Propegation Delay" between input to output, using SINEWAVES, at 10KHz and 20KHz. For the former, you will find it to be about 24uS, for the latter about 12uS. For both, the phase shift is about 90 degrees. Or you can do it in SPICE in just a few minutes.

Again, some basics here. A real-world amplifier circuit contains mechanisms that produce both frequency-dependent and frequency-independent delays. In a typical well-designed Miller-compensated amplifier, the goal is to choose the compensation capacitor so that the frequency-independent delay is completely swamped by the frequency-dependent delay of a first-order slope, yielding a phase margin of 90 degrees at all frequencies above unity gain.

Here's the conceptual error with your square-wave timing test. If we assume that it's indeed a perfect square-wave on input, and the circuit in question doesn't have infinate bandwidth . . . then the output square-wave will have a longer rise time and more rounded leading edge than the input. So we set up our scope, and use the markers to decide where to measure on the x-axis. For the input side, it's easy to locate the marker because the rise-time is infinately short. But on the output, it's comparatively slopey and rounded . . . so when you look at the output and place the marker, the exact placement across the slope determines for which frequency you're measuring the delay. If you just place the marker where it "looks about right", then you're simply meauring the delay of "kinda one of those frequencies" . . . one of an infinate number contained in the perfect squarewave on the input.

But really the time-honored method is to use X/Y mode on your scope to compare the phase as you vary the frequency of a sinewave. You can then CALCULATE the precise delay for any frequency, based on phase. And no, there won't be just one number.

Kirkus, my technique for measuring propagation delay is simple: compare the input to output while using a squarewave source. Observe the difference in time between the rising input waveform and the rising output waveform. That's the delay time. I have yet to see an amplifier where I could not see that on the 'scope.

Since negative feedback only exists if the open-loop (feedback-free) gain is above unity, and since the open-loop response falls off at 6dB/octave . . . the input/output phase response must be 90 degrees or less. So if we're going to talk about "transit time", how would you define that?

It really seems to me that something is glossed over here. In this model phase and time become the same, and is inadequate to explain the behavior of an amplifier that has wide (+200KHz) open loop bandwidth. In such amplifiers the model below falls apart:

Since we know that comparing the phase at the input the output will give us 90 degrees, the "transit time" at 100KHz will be 2500 nanoseconds. At 200KHz, it will be 1250 nanoseconds. At 20KHz, it will be 25000 nanoseconds. So it seems that talking about "transit time", or "propegation delay"[sic], or "delayed feedback", or whatever . . . is a wholly inadequate way of understanding what's going on. Rather, classical Control Theory uses phase relationships to analyze feedback.

Propagation Delay does not alter with frequency anywhere near the audio band, and at those frequencies the delay time is easily measurable. In fact, we can see that at low frequencies feedback works pretty well, but as frequency increases, the feedback is progressively inadequate due to the fixed propagation delay of the circuit having a larger effect as the waveform time decreases. This introduces a time-domain distortion- ringing and odd-ordered harmonic enhancement. It is this phenomena that requires networks in many amplifier designs to prevent negative feedback from becoming positive feedback due to the phase at very high frequencies that are out-of-band but can cause the amp to go into oscillation if not addressed. The model you are proposing relies on propagation time being mutable, which it certainly is not. I'm with Spectron on this one. Sounds to me like control theory is being misapplied here.

Distortion has the property of masking detail in addition to adding loudness cues, so if you can get rid of distortion you get greater transparency and greater smoothness at the same time, provided your techniques for getting rid of distortion don't enhance the 5th, 7th and 9th harmonics. IOW real reductions in distortion have real, immediate sonic benefits that anyone can hear: extreme detail accompanied by smoothness are the hallmarks to look for.

Absolutely true. And there is absolutely no design technique or topology (tubes, solid-state, Class A operation, balanced push-pull, local or global negative feedback, etc.) that can guarantee meaningful improvements in audible distortion. It of course comes down to the proper implementation of a wide variety of techniques.

" Odd ordered harmonics are exacerbated by noise problems in the ground and the power supply..."

Fully agree here with Atmasphere. The more regulated (and noiseless) power supply the better sound quality will be. One can assert that the quality of the power amplifier is not in its signal path so much as in its power supplies. And in many (but not all) cases I would agree with it.

Simon

Atmasphere - concept is beautiful. Tube class A balanced operation without output transformer. The only problem I can see is that this design requires a lot of tubes and each one has about 2.5A heater current - a lot of wasted power. On the other hand any class A has as low as 12.5% efficiency.

Have you ever investigated ultra high vacuum tubes. Military division of Tesla made them before communism fell and Stereophile posted great review of amp built with them. Such tubes can deliver large currents.

The typical transit time of linear amplifiers is about 2000-3000 nanoseconds, which is too slow for effective implementation of global feedback and error correction.

I think this description nicely highlights so many of the conceptual and terminological errors that audiophiles and audiophile equipment designers have about negative feedback.

Looking generically at a solid-state feedback amplifier, their frequency response before feedback is defined by a single "Miller-compensation" capacitor at the voltage-amplifier stage. It is generally flat from DC to some frequency (i.e. 1kHz), and then rolls off at at 6db/octave all the way to the point to where the gain falls below unity, which may be something like 2MHz. While the gain and the frequencies may vary, virtually every common audio opamp has a frequency response that can be described like this. Again, we're talking about it WITHOUT feedback.

Since negative feedback only exists if the open-loop (feedback-free) gain is above unity, and since the open-loop response falls off at 6dB/octave . . . the input/output phase response must be 90 degrees or less. So if we're going to talk about "transit time", how would you define that? Since we know that comparing the phase at the input the output will give us 90 degrees, the "transit time" at 100KHz will be 2500 nanoseconds. At 200KHz, it will be 1250 nanoseconds. At 20KHz, it will be 25000 nanoseconds. So it seems that talking about "transit time", or "propegation delay", or "delayed feedback", or whatever . . . is a wholly inadequate way of understanding what's going on. Rather, classical Control Theory uses phase relationships to analyze feedback.

And classical Control Theory is wholly adequate to understand the circuit behavior when feedback is applied. Musical information isn't "time smeared" from "delayed feedback", it's simply that part of the amplifier circuit operates in quadrature for a huge chunk of the frequency range (in the case of our generic SS amplifier). Just like the filter slope of the very simplest first-order speaker crossover. And this phase relationship doesn't change whether or not feedback is applied (because it's defined by the Miller capacitor) . . . the feedback simply corrects the phase response at the output.

This lagging results in ringing artifacts and enhances ODD-order harmonics which are particularly annoying to the human hearing so even the smallest amounts of these distortions are highly noticeable.

Ringing when feedback is applied is indicative of an open-loop response that is something other than a simple 6dB/octave slope, and this may be due to factors both in the circuit itself and the load it's driving. And this is indeed something that commonly can occur in the real world. But this phenomenon is wholly analyzable with classical Control Theory, and a careful analysis of the amplifier's stability. Further, this type of analysis virtually always reveals the specific mechanisms responsible for the subjective complaints associated with negative feedback.

There are good sounding components using feedback and no feedback, which is simply more proof you need to listen to the component, because the component really is an extension of the skills and philosophy of the designer, and there are good skilled designers employing both methods.

Precisely.

If you can increase the speed of the amp (decrease the propagation delay) theoretically you could speed it up to the point that the odd-ordered enhancement is pushed well outside of the audio band.

Our amps are also pretty fast- 600V/usec is a typical risetime, and we only have one stage of gain. But IME this is still not fast enough, so we resort to other means of getting rid of distortion: class A operation coupled with fully differential balanced operation, which cancels even ordered harmonics not just at the output, but throughout the amplifier. This leaves us with the 3rd harmonic, which is controlled by using only one stage of gain. Odd ordered harmonics are exacerbated by noise problems in the ground and the power supply, so we use star grounding (a lot easier since most of the grounds are balanced) and separate power supplies for the driver and output sections, which also reduces IM distortion.

Distortion has the property of masking detail in addition to adding loudness cues, so if you can get rid of distortion you get greater transparency and greater smoothness at the same time, provided your techniques for getting rid of distortion don't enhance the 5th, 7th and 9th harmonics. IOW real reductions in distortion have real, immediate sonic benefits that anyone can hear: extreme detail accompanied by smoothness are the hallmarks to look for.

Best thread EVER.

ULTRAFAST NEGATIVE FEEDBACK
==================================
When an amplifier has difficulty in delivering required voltage and current many forms of distortions will occur. For example, in transistor amplifiers the increased current drawn by speakers will cause a small voltage drop across the source--i.e., the amplifier itself--which will heavily contribute to the unpleasant so-called "transistor sound.” Many transistor amplifiers use global negative feedback to reduce distortions and widen the bandwidth.

The crucial factor in negative feedback is transit time, the amount of time it takes from when an error is detected at the input until it is corrected at the output. For example, a typical transistor power amplifier has three primary sections: a low-noise high-gain differential input stage, feeding a differential-to-single-ended conversion driving a high-current output stage. Each of these three stages is designed for low distortion and noise, but those attributes typically come at the sacrifice of speed.

The typical transit time of linear amplifiers is about 2000-3000 nanoseconds, which is too slow for effective implementation of global feedback and error correction. This lagging results in ringing artifacts and enhances ODD-order harmonics which are particularly annoying to the human hearing so even the smallest amounts of these distortions are highly noticeable. Long delays in feedback also introduces transient and phase discrepancies, susceptibility to transient overload and vulnerability to disturbances at the output such as reactive speaker interactions.

In contrast, many switching amplifiers don't use low-distortion circuits. Instead, they use much faster digital logic circuits. For example, the Spectron Musician III transit time is much less then 200 nanoseconds. Such an ultra-short transit time allows the amplifier to correct for many small errors; and the control loop can follow the input much more accurately. These characteristics result in a more detailed, transparent sound with less noise and louder yet cleaner musical reproduction.

Roger Modjeski has some interesting design philosophies in general. Topics like feedback certainly stir the hornets nest within him. Just don't get him going on cables. At that point it becomes a swarm.

I have to say that the RM-10 manual is one of the best audio reads I've experienced. The amp itself is excellent too, even with 14db of negative feedback.

Mapman - Moderation is the word. Instead of feedback or no feedback we can settle for moderate feedback. Sound coloration comes from dynamic nonlinearity of the system with the feedback. Amplifier itself is far from being first order low-pass filter and feedback creates loss of stability - hence dynamic nonlinearity. Part of the problem is speaker - being complex load. That would suggest to me that things are really complicated and listening instead of reading will bring better results. Speaker choice and synergy with amplifier appears to be very important.

"There are good sounding components using feedback and no feedback, which is simply more proof you need to listen to the component, because the component really is an extension of the skills and philosophy of the designer, and there are good skilled designers employing both methods."

I think this is the bottom line practically for most.

The caveat is you cannot listen to a single component alone, only a system with all the required components (source, amp, speaker/room) together. A single listen can only tell you what each component is capable of, not how good or bad each piece is or sounds. To sort it all out requires doing your homework and listening to as many combos as you can over time.

I have read Roger Modjeski address this issue, and his opinion seems to be that negative feedback got a bad name as a result of designers who had no formal training in electronics misusing and not being competent to properly implement negative feedback. In the right hands, it is a good thing. I think it would sort of be like a chisel in the hands of an artist versus a non-artist. The artist makes great things with the chisel, and the non-artist cuts his hands.

I listened in a group setting to a Berning amplifier with adjustable feedback: No feedback, low feedback, and high feedback. We all agreed that the low feedback setting was the best, followed closely by no feedback, and distantly by high feedback. The latter setting was the only setting I would consider unlistenable. The no feedback setting was a little too soft and mellow for my tastes, but I could see someone enjoying that type of sound, and I would not kick it out of bed. For me the low feedback setting provided the best of both worlds - detail, controlled bass, without being irritating.

There are good sounding components using feedback and no feedback, which is simply more proof you need to listen to the component, because the component really is an extension of the skills and philosophy of the designer, and there are good skilled designers employing both methods.

Loop feedback in any form is supposed to reduce distortion. It is questionable whether it increases bandwidth, and in some models (see the link I provided) it reduces 'output impedance'. You'll see why I use the quotes if you look at the link.

The *big* problem is that loop feedback, in the process of doing all this stuff, exacts a penalty. This comes from the fact that any circuit that can amplify is doing so at speeds that are easily measured on rather pedestrian test equipment. This time is called Propagation Delay- the time it takes for the signal to propagate from input to output.

Now feedback is created by taking the some of the signal from the output, and applying it to an earlier portion of the circuit, which has a propagation delay. So you can see that the feedback signal is arriving ever so slightly too late to do its job right. The fact that it is too late causes the amplifier to become less stable as frequency increases, and there can be inharmonic noise created at the point where the feedback is returned.

This causes feedback to inject a low level harmonic distortion noise floor composed of harmonics up as high as the 81st harmonic into the output of the amp, and it has two audible artifacts.

The ear uses naturally-occurring odd-ordered harmonics to figure out how loud the sounds are. They are the 5th, 7th and 9th harmonics and they get enhanced (distorted) by feedback by a small amount. However, because these are loudness cues to the human ear this small amount **is easily audible** and audiophiles use the terms 'hard' 'harsh', 'bright', 'brittle', 'chalky', 'clinical' and so on to describe this distortion. Keep in mind that this is the case when the distortion of these harmonics may only be 100ths of a percent!! This is why two amps can measure the same frequency response on the bench but one will be bright and the other is not.

The 2nd problem is that the harmonic noise floor, through another hearing principle called 'masking', will block the ear's natural ability to hear into the noise floor of the playback system (the ear can hear 20 db into a natural noise floor like tape hiss or the wind blowing). Any information below the noise floor is not heard by the ear or not detected as easily. Since ambient soundstage information exists at low level, one of the more obvious effects of feedback is to foreshorten the soundstage depth and width.

Amplifiers in particular that use no feedback tend to have a different voltage response in dealing with the loudspeaker and the designer of the speaker has to accommodate this behavior. IMO, a speaker that requires an amplifier with feedback, due to the issues above, will never sound like real music. Speakers that *are* friendly to zero feedback amps at least have a chance.

see
http://www.atma-sphere.com/papers/paradigm_paper2.html for more information

Negative feedback extends bandwidth, lowers Harmonic and Intermodulation Distortions and lowers output impedance.

Unfortunately, if not used wisely, is increasing TIM - Transient Intermodulation. In time domain it will show as just small overshoot on fast changing signal like square wave. In frequency domain it shows as exaggerated odd harmonics that our ears are very sensitive to (especially higher order - responsible for perception of loudness). In really bad case it can momentary saturate output transistors that will stop responding for a short time since charge is trapped at the output transistor junctions. Our brain fills small gaps like that but it will make us tired. Whole thing (overshoot) happens because of limited bandwidth that is causing delay thru the amp. Delayed signal when summed (in opposite phase) with input signal that is changing rapidly is coming too late and amp for a moment has much higher gain. Class A amps don't require a lot of global feedback and gain (without feedback) is often as low as 200 but class AB has gains reaching 4000.

How amp should be designed? I would pick the most linear transistors I can find. I would use a lot of local feedbacks. I would measure bandwidth without global feedback and would limit bandwidth of the input stage to that bandwidth (necessary condition). Harmonic distortion would be probably 5-10%. I would use just enough feedback to get distortion below 1%. That would be great sounding amp that nobody would buy because of poor spects (distortion, bandwidth).

No feedback (or low feedback) design might sound more alive because distortion gives this effect (like distorted vs clean guitar) but mostly it would sound pleasant and not tiring instead of sounding brightly Hi-Fiish.

TIM was discovered in 70s. Before that designers went crazy with negative feedback - still claiming that it has to be sounding better than tubes. Logic says that if you see numbers like THD=0.000001% something else has to give. I believe that spects are pretty much useless since amp with greatest spects might sound the worse. People often use amps exact power doubling with 4 ohm load vs. 8 ohm load as a sign of great amplifier. I'm not so sure. It will show that power supply is strong but it will also show that a lot of negative feedback is used (since power supply is most likely unregulated).

The main "problem" is that that it's widely misunderstood by both those circuit designers that use it, and those that eschew it. It's also very much out of vogue these days. What negative feedback is, in essence, is the technique of trading circuit gain for circuit bandwidth and linearity.

But the vast majority of audio circuits use some form of negative feedback, regardless of whether or not they're advertised as "zero feedback". It's interesting to notice that many who shun feedback also prefer triodes . . . as most triode circuits have a good dose of negative feedback (based on the tube's internal characteristics). In fact, when a given circuit or active device (tube or transistor) displays the combination of less gain and improved linearity, it's likely there's some kind of negative feedback mechanism that's making it that way.

The common audiophile response is then "well that's LOCAL feedback, which is good!" . . . this usually is explained by lack of "delays" and such. But actually in many cases (i.e. typical solid-state amp), most of the "delay" (really a phase lag, NOT a pure delay) is in one stage (the Miller-compensated voltage amp), and most of the nonlinearity is in another (the output stage), so the stability consequences of global feedback are usually very similar to that of just local feedback around the voltage amp . . . but the global feedback arrangement of course works so much better. Incidentally, this is the main cause for higher-order/frequency distortion products in solid-state amps that use feedback - the distortion rises with frequency because the Miller-compensation technique shifts feedback from global to local as frequency increases, and takes the output stage out of the feedback loop. So the problem is really not with the feedback, but more the lack of it . . . that is, it's not available equally at all the frequencies that need it.

And it's also a misconception that local feedback results in automatically better stability . . . instability in cathode/emitter follower circuits from certain source impedances is a very well-documented condition. In fact, this is the whole reason for the invention of tetrodes and pentodes . . . triodes exhibit poor stability at the limits of their gain and bandwidth, as a result of their internal feedback mechanism. The addition of the screen-grid mostly eliminates the feedback, thus the increase in gain, and decrease in output impedance and linearity.

Oh and as for positive feedback . . . "bootstrapping" networks are extremely common in all kinds of analog circuitry . . . does that count?

I think you may see a response from Atmasphere (Ralph} as he in my opinion is one of the top amplifier designers in the industry;if he replies pay attention to his response.

It makes the music sound less alive. I have a Mesa Baron amp, which has adjustable feedback levels, so it is easy to hear the difference. Increasing the feedback improves the specs, cleans up the bass a little, and robs the mids and highs of life and air. This matters less with heavily processed music (eg., pop or some rock,) which can even benefit (see "cleans up the bass" above) but it is fatal to acoustic or classical music.

Like this?

http://www.wolcottaudio.com/WA_presence.htm

Appreciate Wolcott's integrity.

What about feed Forward ?

There's open loop (global) and closed loop (local). Few designs can be stable into low impedance loads without some feedback and, if you read carefully, some advertise "no global feedback" and sometimes forget a word. In the case of op-amps, it's usually already there.

Easy to find all the negative things about negative feedback. I'm still waiting for someone to advertise positive feedback as being better :)

I think you will find the answer to your question if you read about amplifiers by Soulution. Apparently, their research showed that in conventional amplifiers, the negative feedback that was applied was slightly out-of-time with the signal it was trying to correct, therefore throwing off the coherence of the original signal. I remember years ago hearing a similar criticism of the servo systems in Velodyne subwoofers. Anyway, they supposedly found a way to corrrect that problem, and I think that it is their position that negative feedback is not a bad thing when it is implemented in the way that they do it. The only bad thing is the price of their amplifiers...whew!

Negative feedback has been used in amps for a long time but the use got out of hand in the 1970s when the "spec wars" got going big-time. Amp designers found they could get very low distortion numbers which looked good in magazine ads.

The catch is these spectacularly low distortion numbers were measured with a static signal. Music is a dynamic, constantly changing signal. After a while, people discovered that the correlation between a very low distortion number and how the amp actually sounded in use was a bit more tenuous than first thought.

Amp design is a balance of competing factors; feedback is only one aspect of that equation. The key is to find the best overall balance for the amp in question. Using zero-feedback. when it is the subject of tunnel vision focus that ignores other parameters, is not a sole guarantee of performance any more than the mindless pursuit of super-low distortion numbers was 30 or 40 years ago.