Ah, Terry,
I'm too bloody old for this caper now....... and you are too young and fit. A strategic withdrawal over lattes would be a much better option.
Besides, if I talk to you long enough, and buy a LOT of lattes, will you tell me your audio secrets? Inquiring, lazy minds want to know......
Cheers,
Hugh
I'm too bloody old for this caper now....... and you are too young and fit. A strategic withdrawal over lattes would be a much better option.
Besides, if I talk to you long enough, and buy a LOT of lattes, will you tell me your audio secrets? Inquiring, lazy minds want to know......
Cheers,
Hugh
Graham Maynard said:
This is fabrication, as is your claim that you did not make it up. No competent author would ever write such nonsense as a definition of linearity. The definition of linearity was stated correctly by Rodolfo in this post . It's a fundamental concept, not only to engineering, but to mathematics as well. As to whether your claim that linearity is the same thing as "continuous proportionality" is correct or not, it is so vague that it says essentially nothing, so the issue of correctness can't even be addressed.
I mean a better amp audibly
Actually there are many interesting thing(s) in this article.
Let's start with human ear. I see in Fletcher-Munson curve, our ear is not having flat sensitivity for frequencies from 20hz-20khz. The response is curved, and most sensitive around 1-5khz. And, the curve differs for various sound level, getting more linear at louder sound, and getting more curved at fainter sound. This means our ear is very sensitive, especially in faint sound ambient.
This means our ear is different than a microphone, right? Measurement microphone has no Fletcher Munson curve, takes all frequency amplitude equally. Our ear can recognize pleasant or not pleasant reproduction.
The Hiraga article (which is translated by Janneman) started with this :
He's talking about what is known as "Back EMF" here?
Ok, the point is clear. I think he means blue light and green light produces yellow light.
He is right. I hear "pinknoise", it is not 1 tone, but I hear it as many tones, from bass to treble all in one. I see white light as white, cannot know it has 7 colors inside white light.
In the chapter of "Harmonic Sounds, Dissonant Sounds", he mentioned the works of Zarlin (Zarlin scale), Herrmann von Helmhotz (On the sensation of Tone), Fletcher, Zwicker, S.S. Stevens, Steinberg. Wagel and Lane analyze amplifier harmonic distortion and its subjective influence.
I imagine harmonics of audio amplifier is like water pillow. If we push the lower order ones, the higher orders will rise. If we let the lower order ones not pressed, the high orders will now shown. Maybe in measurement it can be "considered" to have levels of not-noticeable, but our ear is sensitive.
lumanauw
My posting re your question below was incomplete:
I want to ask a question. What do you think of feedback system (input differential) that is built with non-linear devices (transistors), will it be able to do good job in managing non-linear distortions (like the so called speaker nonlinearities with all its components)? Or, they do the opposite, makes worsen things?
You have to accept the fact that there are NO linear amplifying devices. Some are more, some are less, but none is linear.
To manage speaker distortion, one shall go to current drive - that is not the same as a moderatelly low damping amplifier.
But then it requires a dedicated speaker/amp combo.
The current drive amp can be made with tubes or transistors - it's just a matter of personal preference, as in this site we have tubes and solid state forums.
And chip amps, and class D amps.
Now, not to use NFB would mean that the circuit can be just as good as the non linear components are - independent of the type..
And, going back to my first reply, Hiraga's test has shown that there are good and bad amplifiers, and that the high NFB one wasnt the bad boy of the group.
My posting re your question below was incomplete:
I want to ask a question. What do you think of feedback system (input differential) that is built with non-linear devices (transistors), will it be able to do good job in managing non-linear distortions (like the so called speaker nonlinearities with all its components)? Or, they do the opposite, makes worsen things?
You have to accept the fact that there are NO linear amplifying devices. Some are more, some are less, but none is linear.
To manage speaker distortion, one shall go to current drive - that is not the same as a moderatelly low damping amplifier.
But then it requires a dedicated speaker/amp combo.
The current drive amp can be made with tubes or transistors - it's just a matter of personal preference, as in this site we have tubes and solid state forums.
And chip amps, and class D amps.
Now, not to use NFB would mean that the circuit can be just as good as the non linear components are - independent of the type..
And, going back to my first reply, Hiraga's test has shown that there are good and bad amplifiers, and that the high NFB one wasnt the bad boy of the group.
lumanauw
We both were posting at the same time..
From my viewpoint, the worse distortion is intermodulation - becuse for the ear blue and green is not yellow, if it is it sounds bad...
Yes, Hiraga's "Counter Electro Magnetic Force (CEMF) is the same as 'Back EMF".
Re the rising level of the harmonics, he is right again.
The manufacturers have chosen a THD number that was easy to measure, against a proposal by BBC's Shorter (a looong time ago) that would give much meaningful numbers (more weight for the higher order harmonics).
That's why, for me at least, THD is an empty number.
Now, how do we solve the rising harmonic problem?
There are two schools, each one calling 'his truth' the absolute truth:
- One is the no feedback school, in which one accepts some low order (mostly second and third) harmonics in exchange for no higher order harmonics.
But such a circuit will have intermodulation, so it will produce yellow sound...
- the other is the feedback school, in which harmonics have to be of such a low level that even our sensitive ear cannot detect it.
If the design is right, this goal can be achieved, and by consequence intermodulation will not be present also.
Which is the right one? It depends very much on one's personal taste, sytem, listening habits, etc.
That's why these two schools are always in different camps since the just tube times - triodes X tetrodes.
We both were posting at the same time..
From my viewpoint, the worse distortion is intermodulation - becuse for the ear blue and green is not yellow, if it is it sounds bad...
Yes, Hiraga's "Counter Electro Magnetic Force (CEMF) is the same as 'Back EMF".
Re the rising level of the harmonics, he is right again.
The manufacturers have chosen a THD number that was easy to measure, against a proposal by BBC's Shorter (a looong time ago) that would give much meaningful numbers (more weight for the higher order harmonics).
That's why, for me at least, THD is an empty number.
Now, how do we solve the rising harmonic problem?
There are two schools, each one calling 'his truth' the absolute truth:
- One is the no feedback school, in which one accepts some low order (mostly second and third) harmonics in exchange for no higher order harmonics.
But such a circuit will have intermodulation, so it will produce yellow sound...
- the other is the feedback school, in which harmonics have to be of such a low level that even our sensitive ear cannot detect it.
If the design is right, this goal can be achieved, and by consequence intermodulation will not be present also.
Which is the right one? It depends very much on one's personal taste, sytem, listening habits, etc.
That's why these two schools are always in different camps since the just tube times - triodes X tetrodes.
Hi, AndyC,
I think non feedback power amp stuck in output stage. Maybe classA can be better, this means it is not good making a non-feedback with low biased classAB output stage. You simulate output stages alot Is it possible to get linear performance of output stage without feedback?
Hi, Jorge,
Is the artifacts (what is left over) of feedback process is really not-detectable by our ear? I feel that the artifacts of feedback process is mosly high order one, the one which our ear is most sensitive to.
Some designs are making this artifact are to be said "below noise". Not noticeable by our ear? What makes pleasant and non pleasant (fatiguing) sound reproduction?
Talking about linear device. I think that it is not easy to make a nice sounding power amp. Why? Because our ear is not a LINEAR device at all, like Fletcher Munson curve pointed.
I think non feedback power amp stuck in output stage. Maybe classA can be better, this means it is not good making a non-feedback with low biased classAB output stage. You simulate output stages alot
Is it possible to get linear performance of output stage without feedback? Hi, Jorge,
I'm not quite capturing this. Intermodulation is problem with feedback amp or non-feedback amp?
Is the artifacts (what is left over) of feedback process is really not-detectable by our ear? I feel that the artifacts of feedback process is mosly high order one, the one which our ear is most sensitive to.
Some designs are making this artifact are to be said "below noise". Not noticeable by our ear? What makes pleasant and non pleasant (fatiguing) sound reproduction?
Talking about linear device. I think that it is not easy to make a nice sounding power amp. Why? Because our ear is not a LINEAR device at all, like Fletcher Munson curve pointed.
Graham Maynard said:
Graham,
I am a bit at a loss now. I realise there is much in your post that I can relate to. Yes, FC measurements say *something* about the amp. But there are many, established measurement types that do what FC measurements do. For example, the old AP S1 does a fully automated ISO-31 test which is a standardised (ISO) test with 31 different freqs simultaneously, from 20Hz to 20kHz. The analyses shows ALL THD and IM products from all possible combinations of freqs and levels. It also does DIM tests where one uses a combination of square wave with sine wave to test small signal performance across the full amplitude spectrum. Indeed, your FC signal will do some similar things.
But I now think you case actually is that amps that handle dynamic, high level signals well, sound well, ie the JLH. That is a valuable result in itself. What is very confusing is all that talk about FCD, output-inductor-originating-uncorrected-errors. There are none! It's just a freq response non-linearity.
So, WHAT exactly IS your case?
Jan Didden
From pcavtech.com:
"What is Linear Distortion?
Linear Distortion, is a form of signal processing error that creates no new frequencies beyond those that are present in the input signal. Linear Distortion changes the relationship of the size and relative timing of the frequencies that compose the input signal.
Nonlinear distortion differs from linear distortion in that does create new tones at the output. The changes in relative timing of various frequencies due to linear distortion are different from the changes in timing due to Jitter or FM distortion. FM distortion changes the timing within a tone, while linear distortion changes only the timing of the entire tone.
Linear and Nonlinear distortion are directly based on frequencies present in the input signal. If there is no input, then there will no errors at the output, due linear and nonlinear distortion."
Jan Didden
"What is Linear Distortion?
Linear Distortion, is a form of signal processing error that creates no new frequencies beyond those that are present in the input signal. Linear Distortion changes the relationship of the size and relative timing of the frequencies that compose the input signal.
Nonlinear distortion differs from linear distortion in that does create new tones at the output. The changes in relative timing of various frequencies due to linear distortion are different from the changes in timing due to Jitter or FM distortion. FM distortion changes the timing within a tone, while linear distortion changes only the timing of the entire tone.
Linear and Nonlinear distortion are directly based on frequencies present in the input signal. If there is no input, then there will no errors at the output, due linear and nonlinear distortion."
Jan Didden
It's always good to be clear what you are talking about...
From mastersofaudio.com:
"Virtually everything that affects an audio signal as it passes through the system is a form of distortion, although some forms are known by more specialized names. Two main categories exist: non-linear and linear distortion.
The non-linear variety is what usually carries the name distortion. Most familiar, perhaps, is "harmonic" distortion, where the equipment generates spurious signals related to the wanted signal harmonically. Since most music is rich in harmonics, this sort of distortion is often masked by the program material, unless it is high enough to change the character of the instruments themselves. Also contributing new sounds is intermodulation distortion (IM), which generates frequencies at the numerical sum and difference of two sounds in the program material. Because these are unlikely to be harmonically related to the music, they are less likely to be masked.
In linear distortion, nothing is added, but the relationships between different parts of the signal are altered in some fashion. The most prevalent of these is rarely termed distortion; when an audio component deals with some frequencies more efficiently than others -- emphasizes them -- we normally talk of frequency response irregularities (or non-linearities) rather than distortion. But it is a form of distortion nonetheless.
By the same token, some types of audio equipment tend to delay certain frequencies with respect to others. This is known as "phase shift" or "group delay," and is most often a physical attribute of speakers, although it has always been a measurable factor in the performance of filters, crossover networks, equalizers, and the like. "
Jan Didden
From mastersofaudio.com:
"Virtually everything that affects an audio signal as it passes through the system is a form of distortion, although some forms are known by more specialized names. Two main categories exist: non-linear and linear distortion.
The non-linear variety is what usually carries the name distortion. Most familiar, perhaps, is "harmonic" distortion, where the equipment generates spurious signals related to the wanted signal harmonically. Since most music is rich in harmonics, this sort of distortion is often masked by the program material, unless it is high enough to change the character of the instruments themselves. Also contributing new sounds is intermodulation distortion (IM), which generates frequencies at the numerical sum and difference of two sounds in the program material. Because these are unlikely to be harmonically related to the music, they are less likely to be masked.
In linear distortion, nothing is added, but the relationships between different parts of the signal are altered in some fashion. The most prevalent of these is rarely termed distortion; when an audio component deals with some frequencies more efficiently than others -- emphasizes them -- we normally talk of frequency response irregularities (or non-linearities) rather than distortion. But it is a form of distortion nonetheless.
By the same token, some types of audio equipment tend to delay certain frequencies with respect to others. This is known as "phase shift" or "group delay," and is most often a physical attribute of speakers, although it has always been a measurable factor in the performance of filters, crossover networks, equalizers, and the like. "
Jan Didden
Hi Andy,
Fabrication ! Here we go again ! You calling me a liar ????
You know very well how I reported a definition of linearity.
Allow me to repeat it.
Linearity;- where output *continuously* varies in *direct proportion* to input.
I am totally fed up with being picked on for the tiniest wordage contraction as is necessary for me to compact writing when replying to posts. This 'rubbishing attitude' helps no one, and if you disagree with that definition then where do we go from here ?
( I did not jump on Jan's recent slip, which he corrected a couple of posts back, because I took the trouble to look for the message he thought he was writing !)
Hi Lumanauw,
Post#583 - Very nicely expanded indeed !
And at those high harmonic frequencies it takes but the tiniest of reactively induced circuit delays to generate minute levels of amplitude error in time, whether they are caused by 'linear' circuitry ( as measured in time isolation ) or not.
Time - frequency - amplitude: All are inextricably linked by conservation of energy.
When testers examine an amplitude response to a steady sinewave in time isolation, even when the 'distortion' is said to be below noise level, they ignore conservation of energy in relation to a whole composite waveform, which brings us right back to the load ( loudspeaker ) modified circuit/path delays and their effect upon amplifier and cable amplitude in real time; ie. time from t=0 !
Hi Hugh,
I'll go to Carlos' thread later for that one.
That circuit simulates with a first sine cycle thd figure of <0.01% from t=0 to t=100uS at 10kHz whilst outputting 25W into 8 ohms, and it is still mostly 2nd plus 3rd harmonic in spite of the energising spectrum and the non-allowance for group delay.
Even with the input filter bypassed on other circuits and after years of looking I have not yet seen this low a FCD figure before (some are truly atrocious but their designers don't know !), though whether the eventual sound reproduction is preferable will be influenced by many and other inter-related factors.
Hi Jan,
You've just posted, and I can see that I'm going to have to study carefully to understand the wisdom therein.
Cheers ........... Graham.
Fabrication ! Here we go again ! You calling me a liar ????
You know very well how I reported a definition of linearity.
Allow me to repeat it.
Linearity;- where output *continuously* varies in *direct proportion* to input.
I am totally fed up with being picked on for the tiniest wordage contraction as is necessary for me to compact writing when replying to posts. This 'rubbishing attitude' helps no one, and if you disagree with that definition then where do we go from here ?
( I did not jump on Jan's recent slip, which he corrected a couple of posts back, because I took the trouble to look for the message he thought he was writing !)
Hi Lumanauw,
Post#583 - Very nicely expanded indeed !
And at those high harmonic frequencies it takes but the tiniest of reactively induced circuit delays to generate minute levels of amplitude error in time, whether they are caused by 'linear' circuitry ( as measured in time isolation ) or not.
Time - frequency - amplitude: All are inextricably linked by conservation of energy.
When testers examine an amplitude response to a steady sinewave in time isolation, even when the 'distortion' is said to be below noise level, they ignore conservation of energy in relation to a whole composite waveform, which brings us right back to the load ( loudspeaker ) modified circuit/path delays and their effect upon amplifier and cable amplitude in real time; ie. time from t=0 !
Hi Hugh,
I'll go to Carlos' thread later for that one.
That circuit simulates with a first sine cycle thd figure of <0.01% from t=0 to t=100uS at 10kHz whilst outputting 25W into 8 ohms, and it is still mostly 2nd plus 3rd harmonic in spite of the energising spectrum and the non-allowance for group delay.
Even with the input filter bypassed on other circuits and after years of looking I have not yet seen this low a FCD figure before (some are truly atrocious but their designers don't know !), though whether the eventual sound reproduction is preferable will be influenced by many and other inter-related factors.
Hi Jan,
You've just posted, and I can see that I'm going to have to study carefully to understand the wisdom therein.
Cheers ........... Graham.
lumanauw said:
Hi, Jorge,
I'm not quite capturing this. Intermodulation is problem with feedback amp or non-feedback amp?
Is the artifacts (what is left over) of feedback process is really not-detectable by our ear? I feel that the artifacts of feedback process is mosly high order one, the one which our ear is most sensitive to.
Some designs are making this artifact are to be said "below noise". Not noticeable by our ear? What makes pleasant and non pleasant (fatiguing) sound reproduction?
Talking about linear device. I think that it is not easy to make a nice sounding power amp. Why? Because our ear is not a LINEAR device at all, like Fletcher Munson curve pointed.
Hello, lumanauw
Any non linear component will generate both harmonic and intermodulation components at the same time.
In the simplest of all circuits, a tube, having a more rounded transfer curve will generate less higher order harmonics than a transistor, that has a steeper transfer curve. And both will generate intermodulation.
This is before any global feedback is applied.
Now, since transistors are cheap and use little power, there's no reason to use them in the simplest circuits, and by using more refined circuits it's possible to design amplifiers (usually class A) that do not need global feedback and are considered decent sounding
That's where the problem begins - there is not an universally accepted definition of a decent sounding amplifier.
Some people's ears are more sensitive to harmonics, others to intermodulation, and the subjective aspect in audio is very strong.
Not to forget the marketing department of manufacturers...
The question then is - if it's possible to design amplifiers without global feedback why to use it?
Because people's hearing is not the same, some like it one way, some another!
Now, re amplifiers with global feedback - there are lots of very highly regarded amps with global feedback, quite more than the feedback less ones.
A few manufacturers even have both types of amps - pick the one you like better!
The higher order harmonics problem has been taken care through good design.
Out of curiosity, the late master Baxandall wrote an article in which he showed that little global feedback is worse than no feedback. But then more feedback was the solution.
And I don't remember any JLH amp without feedback.
Re fatigue - in my personal experience, it's due to hidden ressonances in the signal (what in loudspeaker parlance is a ringing).
It's more likelly to be present in a badly designed global feedback amp than in a feedback less one.
Re Fletcher Munson curves: don't place it high on your scale. The amps that had Fletcher Munson compensation in the volume control were never highly regarded.
The ear is even more complex than one may think!
Regards,
Jorge
Graham Maynard said:
Repeating an incorrect definition of linearity over and over again can easily be explained by a lack of understanding. But brazenly misrepresenting this definition as a textbook definition goes beyond simple misunderstanding.
If you don't like people calling your bluff, the solution is to not bluff in the first place.
Rodolfo (re:#570)
I did not claim it didn't. But excellent - may we conclude that any system capable of oscillatory behavior by definition has a nonlinear component?
May we claim that, therefore, any (feedback or not) amplifier-load system capable of oscillatory behavior is a nonlinear system?
May I assume that most real life amplifiers may exhibit oscillatory behavior and are therefore nonlinear systems?
Jan,
The feedback makes a difference *with time lag dt*. Or in other words, for a sine wave input, phase shift. Therefore, the feedback will not correct any signal to perfection. This is what I meant. Rodolfo stated the same in post #570. The better (faster) the feedback, the lesser the distortion, but the higher (in order, not necessarily amplitude) the harmonics eventually created (for a sine wave). The question is whether the deviation is significant for our purposes.
In essence, strictly speaking, the feedback does not correct the initial error - it always comes a little too late. The approximation is very good but it is an approximation. You can clad that in formulas and plots of phase and group delay but that would just change the wording. A nonlinear fix for an initial nonlinearity. It works very well in engineering as a whole - but for a perfect correction of a nonlinearity, the control system would have to be anticipatory of the expected error, not post fact (feedback) based.
This is all sophistry anyway. The proof is in the pudding. Different approaches my yield satisfactory results.
I did not claim it didn't. But excellent - may we conclude that any system capable of oscillatory behavior by definition has a nonlinear component?
May we claim that, therefore, any (feedback or not) amplifier-load system capable of oscillatory behavior is a nonlinear system?
May I assume that most real life amplifiers may exhibit oscillatory behavior and are therefore nonlinear systems?
Jan,
The feedback makes a difference *with time lag dt*. Or in other words, for a sine wave input, phase shift. Therefore, the feedback will not correct any signal to perfection. This is what I meant. Rodolfo stated the same in post #570. The better (faster) the feedback, the lesser the distortion, but the higher (in order, not necessarily amplitude) the harmonics eventually created (for a sine wave). The question is whether the deviation is significant for our purposes.
In essence, strictly speaking, the feedback does not correct the initial error - it always comes a little too late. The approximation is very good but it is an approximation. You can clad that in formulas and plots of phase and group delay but that would just change the wording. A nonlinear fix for an initial nonlinearity. It works very well in engineering as a whole - but for a perfect correction of a nonlinearity, the control system would have to be anticipatory of the expected error, not post fact (feedback) based.
This is all sophistry anyway. The proof is in the pudding. Different approaches my yield satisfactory results.
Hi, Jorge,
It came out, and from 2 people whom I respect.
Offcourse I don't intend at all to make an amp (or signal processor) that do not have linear amplification from 20hz to 20khz. They has to have flat frequency response. I never put passive EQ's or notch filter anywhere in amp design. Making audio gear with frequency response like Fletcher Munson curve will be wrong from any point of view (measurement or auditioning). Like PMA, I think if we hear live performance, there are no Fletcher Munson curve there, all flat natural response, without knowing what kind of balance our ear likes.
I don't know if anyone does this.
I'm a Diy'er with "not enough" measuring equipment. I only got scopes, signal generator, speakers/headphones. I cannot view high order harmonics or IM.
So, what's the point of bringing Fletcher Munson curve? This is my "theory". If an amp has high order harmonics, it will be spreaded within 20hz-20khz. Maybe in 5khz it consist of stacking of 10th order from 25hz, or plus 7th order from 200hz, who knows?
In making amps, offcourse I will auditioned it full range, I hear the tonal balance and how the sound from 20hz to 20khz.
But I give attention more for the range that our ear is most sensitive, like 1khz-5khz. Because in this area our ear is most sensitive, it will be easier to know wheter a certain design has high order harmonics (which will be not pleasant). It will be difficult to notice this in bass area, for example.
So I hear more in 1khz-5khz (midrange and upper midrange) to know wheter one design has more or less high order harmonics than other design. And I do find out, one design is more pleasant /non fatiquing than others in 1khz-5khz.
Maybe this is "stone age" way to know how one design differs from the other. And I follow by asking "why is that?"
Hi, PMA,
Why is that after I wrote that post that mentioned Fletcher Munson curve, I got a feeling that it will be misinterpreted.
It came out, and from 2 people whom I respect. Offcourse I don't intend at all to make an amp (or signal processor) that do not have linear amplification from 20hz to 20khz. They has to have flat frequency response. I never put passive EQ's or notch filter anywhere in amp design. Making audio gear with frequency response like Fletcher Munson curve will be wrong from any point of view (measurement or auditioning). Like PMA, I think if we hear live performance, there are no Fletcher Munson curve there, all flat natural response, without knowing what kind of balance our ear likes.
I don't know if anyone does this.
I'm a Diy'er with "not enough" measuring equipment. I only got scopes, signal generator, speakers/headphones. I cannot view high order harmonics or IM.
So, what's the point of bringing Fletcher Munson curve? This is my "theory". If an amp has high order harmonics, it will be spreaded within 20hz-20khz. Maybe in 5khz it consist of stacking of 10th order from 25hz, or plus 7th order from 200hz, who knows?
In making amps, offcourse I will auditioned it full range, I hear the tonal balance and how the sound from 20hz to 20khz.
But I give attention more for the range that our ear is most sensitive, like 1khz-5khz. Because in this area our ear is most sensitive, it will be easier to know wheter a certain design has high order harmonics (which will be not pleasant). It will be difficult to notice this in bass area, for example.
So I hear more in 1khz-5khz (midrange and upper midrange) to know wheter one design has more or less high order harmonics than other design. And I do find out, one design is more pleasant /non fatiquing than others in 1khz-5khz.
Maybe this is "stone age" way to know how one design differs from the other. And I follow by asking "why is that?"
You are right. Too many factors involved in this. It's not just using or not using feedback. The answer to make good sounding power amp is to take care of ALL the details (like PRR and John Curl said).
Aha, so there is something in feedback amp that don't exist in non feedback type.
Hello, lumanauw
Right, I misinterpreted your post...It seems I'm doing it a lot, latelly
You're right about the most sensitive area of hearring. Maybe that's one of the resons 1kHz was choosen as the main reference frequency - the lower order harmonics are in the ear's most sensitive range?
About measuring equipment - have you considered using a PC sound card as generator/spectrum analyser? That's what I use.
Re feedback again - a feedback and a non feedback amp can share exactly the same thing - bad design
The biggest mistake a designer can make (usually due to cost or marketing pressures) is to use feedback as a band aid for poor design.
This is a positive aspect of non feedback amps - as there is no band aid, any design mistakes will be there upon first listening.
And high feedback amps do distort horribly upon clipping - so they have to be powerfull enough not to clip in the normal listening environment.
Feedback shall be used as 'icing in the cake' , a final (and necessary in my opinion) step in a good project.
A good amp can be listened without feedback and will sound good. Then, one applies feedback to make it excelent.
My conceptual project (posted in another thread) has less IM distortion without feedback than many commercial amps with feedback...
Right, I misinterpreted your post...It seems I'm doing it a lot, latelly
You're right about the most sensitive area of hearring. Maybe that's one of the resons 1kHz was choosen as the main reference frequency - the lower order harmonics are in the ear's most sensitive range?
About measuring equipment - have you considered using a PC sound card as generator/spectrum analyser? That's what I use.
Re feedback again - a feedback and a non feedback amp can share exactly the same thing - bad design
The biggest mistake a designer can make (usually due to cost or marketing pressures) is to use feedback as a band aid for poor design.
This is a positive aspect of non feedback amps - as there is no band aid, any design mistakes will be there upon first listening.
And high feedback amps do distort horribly upon clipping - so they have to be powerfull enough not to clip in the normal listening environment.
Feedback shall be used as 'icing in the cake' , a final (and necessary in my opinion) step in a good project.
A good amp can be listened without feedback and will sound good. Then, one applies feedback to make it excelent.
My conceptual project (posted in another thread) has less IM distortion without feedback than many commercial amps with feedback...
MBK said:
MBK,
Kindly, read my post again. You are concluding any system capable of oscillatory behavior must be nonlinear and this is wrong.
In fact, just the opposite is true. For a system to be capable of oscillation, its transfer must have at least a couple of conjugate poles within the right half of the s plane. What this means is even the slightest disturbance with signal components coinciding with the poles (usually thermal noise) is enough to excite an increasing amplitude oscillation. The rate of increase of this oscillation is faster the farther from the imaginary axis the poles are located.
A ideal linear system will give rise to an oscillation whose amplitude approaches infinite with time tending to infinite, thus requiring infinite power.
In the real world, there is no such thing as infinite amplitude and infinite power. What happens is the system at some amplitude threshold ceases to be linear, typically when the peak amplitude reaches near power rails voltages.
So what we have is a linear system that behaves as such restricted obviously (so much so we never even mention it) to a certain signal maximum range, we discount it ceases to behave linearly beyond it.
In summary, a linear system can oscillate. The nonlinearity required in an oscillator is simply a consequence of the imposed limits as to how large the signal can be, but has nothing to do with what makes it oscillate.
Rodolfo
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.