You can define "works" any way you want. I define it as getting the job done. Without feedback my hands would be flapping about all over the place, probably not even hitting the keyboard. With feedback I can type at a fair pace.Kuei Yang Wang said:
The averaging into straght line is, again, exactly why I advocate good damping in an amplifier. Cut out the ringing and those little jitters average out pretty much perfectly into a straight line.
Konnichiwa,
Hmmm. You found that you preferred that by traditional criterias performed poorely in areas where amplifiers are traditionally caharterised, but for which many times a poor if any correlation between these parameters and percieved "good sound" has been illustrated.
Therefore you may have merely preferred the absence of some forms of distortion not easily qualified using traditional measurements.
The only way you could have illustrated what you claim to have done would be to utilise a traditionally designed high NFB Amplifier (eg an Amplifier you REALLY DISLIKE) and make you to "like it" simply by adding distortion and output impedance. Did you do that?
Sayonara
SY said:
Hmmm. You found that you preferred that by traditional criterias performed poorely in areas where amplifiers are traditionally caharterised, but for which many times a poor if any correlation between these parameters and percieved "good sound" has been illustrated.
Therefore you may have merely preferred the absence of some forms of distortion not easily qualified using traditional measurements.
The only way you could have illustrated what you claim to have done would be to utilise a traditionally designed high NFB Amplifier (eg an Amplifier you REALLY DISLIKE) and make you to "like it" simply by adding distortion and output impedance. Did you do that?
Sayonara
Thorsten, read again. I was comparing output to input. The amp I "liked" better (the open loop amp) changed the sound. The one I didn't like as much (the feedback amp) had an output that was indistinguishable by ear from the input.
To Jeff and tube dude's points, it does raise some philosophical issues!
To Jeff and tube dude's points, it does raise some philosophical issues!
take a simple emitter follower , feed it with a square wave ...tadaaaa -overshoot= NFBserengetiplains said:
the same effect with global feedback , the only difference is the global feedback is slower .
i dont know what is this , i know a method where u compare the input to the output with some good instrumental amp or a differential amp to see the "difference"SY said:
Konnichiwa,
Which implies the presence of a third amplifier or device to allow the comparison between low level line feed and power output, which in turn becomes the bottleneck.
Through a third amplifier.
You added in sufficiently additional variables to invalidate the test if your aim was to decide if you preferred the fact that the Amplifier you liked added certain types of distortion and frequency response aberations, or if you actually preferred the absence of something which was masked by eiuther source or third amplifying device.
A valid test for the assertation "I like the open loop amp better because it adds distortion and alters the speakers frequency response and not because it omits a non-classified type of distortions that is present in Amp's I don't like!" is possible only by taking an Amp you do not like and adding the open loop's amp's distortion and other "problems" and by then comparing the two amplifiers in a preference test.
If you find yourself unable to distinguish or prefer one to the other; then what you liked was the added distortion and frequency response alteration and it may be wise to start building amplifiers which add these in.
If you still prefer the open loop Amp however then what you preferred was not the added distortion and frequency response alteration (which you merely tolerated as "harmless") but the absence of proposed factor X.
It strikes me that your experiment by the very experimental setup was designed to return a specific result which appeared toi answer the Question "Is the proposition of factor X true or false!" when in fact the experimental setup garanteed a reliable return of "False" regardless of the actual (objective) situation.
Experimental design is really crucial and it is easy to be decieved by others unless one QUESTIONS EVERYTHING (eg. behaves like a good sceptic should).
I have before experimented with hybrid Amp's that go a long way to that end, meaning they add a lot of THD in classic triode fashion AND they have a high output impedance (by solid state standards anyway). In most classically characterised terms the Hybrid Amp behaved closely like a SE Valve Amp. Did it sound like one? Hell NO!
It sounded pretty decent though, but that is another story.
It led me however to conclude that to attempt to gain "SE Tube Sound" by using a "well designed" solid state Amplifier and introducing the "imperfections" that in the traditional sense charaterise SE Valve Amplifiers was not possible. Which sucked, as such an "invention" would have been worth quite a bit of money in the commercial context.
Just add the apropriate marketing BS to it and start marketing $ 500 Integrated Amplifiers that give "True High End SE Sound". It might even work just by making it anyway and patenting/marketing it suitably, even if the premises on which the principle builds is false, but I am not interested in deception.
So, you may wish to either redo your experiments thusly that third influences are removed or fully disclose the experimental setup you used to lay to rest concerns you may have engaed in an expensive and labourious excercise in self deception.
Sayonara
SY said:
Which implies the presence of a third amplifier or device to allow the comparison between low level line feed and power output, which in turn becomes the bottleneck.
SY said:
Through a third amplifier.
You added in sufficiently additional variables to invalidate the test if your aim was to decide if you preferred the fact that the Amplifier you liked added certain types of distortion and frequency response aberations, or if you actually preferred the absence of something which was masked by eiuther source or third amplifying device.
A valid test for the assertation "I like the open loop amp better because it adds distortion and alters the speakers frequency response and not because it omits a non-classified type of distortions that is present in Amp's I don't like!" is possible only by taking an Amp you do not like and adding the open loop's amp's distortion and other "problems" and by then comparing the two amplifiers in a preference test.
If you find yourself unable to distinguish or prefer one to the other; then what you liked was the added distortion and frequency response alteration and it may be wise to start building amplifiers which add these in.
If you still prefer the open loop Amp however then what you preferred was not the added distortion and frequency response alteration (which you merely tolerated as "harmless") but the absence of proposed factor X.
It strikes me that your experiment by the very experimental setup was designed to return a specific result which appeared toi answer the Question "Is the proposition of factor X true or false!" when in fact the experimental setup garanteed a reliable return of "False" regardless of the actual (objective) situation.
Experimental design is really crucial and it is easy to be decieved by others unless one QUESTIONS EVERYTHING (eg. behaves like a good sceptic should).
I have before experimented with hybrid Amp's that go a long way to that end, meaning they add a lot of THD in classic triode fashion AND they have a high output impedance (by solid state standards anyway). In most classically characterised terms the Hybrid Amp behaved closely like a SE Valve Amp. Did it sound like one? Hell NO!
It sounded pretty decent though, but that is another story.
It led me however to conclude that to attempt to gain "SE Tube Sound" by using a "well designed" solid state Amplifier and introducing the "imperfections" that in the traditional sense charaterise SE Valve Amplifiers was not possible. Which sucked, as such an "invention" would have been worth quite a bit of money in the commercial context.
Just add the apropriate marketing BS to it and start marketing $ 500 Integrated Amplifiers that give "True High End SE Sound". It might even work just by making it anyway and patenting/marketing it suitably, even if the premises on which the principle builds is false, but I am not interested in deception.
So, you may wish to either redo your experiments thusly that third influences are removed or fully disclose the experimental setup you used to lay to rest concerns you may have engaed in an expensive and labourious excercise in self deception.
Sayonara
"Which implies the presence of a third amplifier or device to allow the comparison between low level line feed and power output, which in turn becomes the bottleneck."
Good point. VERY good point.
The only way to do a comparison between the source and the output of the amplifier is to insert another amplifier.
This method may seem to work if you attenuate the output of the test amplifiers to match the source and then feed them through an ABC switch to a third amplifier.
Unfortunately, the resistive load on the amplifiers under test will invalidate the test. In the real world, power amplifiers feed loudspeakers ... not purely resistive loads.
After all is said and done, there is a time delay through every amplifier. Inverse feedback can improve the performance in some areas while degrading the performance in other areas.
Only the listener can decide which trade-off he (or she) prefers.
Good point. VERY good point.
The only way to do a comparison between the source and the output of the amplifier is to insert another amplifier.
This method may seem to work if you attenuate the output of the test amplifiers to match the source and then feed them through an ABC switch to a third amplifier.
Unfortunately, the resistive load on the amplifiers under test will invalidate the test. In the real world, power amplifiers feed loudspeakers ... not purely resistive loads.
After all is said and done, there is a time delay through every amplifier. Inverse feedback can improve the performance in some areas while degrading the performance in other areas.
Only the listener can decide which trade-off he (or she) prefers.
Frank, two things to consider. First, why impose a resistive load as a limitation? It's easy to cob together a load that looks like a loudspeaker. You can argue that it's an easier load than a loudspeaker (I'd disagree) but even so, if an amp can't drive that without changing the sound, then it certainly won't do better with a more demanding load.
That horrible "third amp" (second, really- you only test one amp at a time) was apparently good enough to show that the non-feedback amp was distorting.
That horrible "third amp" (second, really- you only test one amp at a time) was apparently good enough to show that the non-feedback amp was distorting.
Does a non-feedback power amp will always produce output signal different than its input (always the case)? Or there is an amp without feedback that has very good linearity compared to input signal?
Reading about feedback is interesting. If I read from the first of its invention (for telephone use), I get it as a "good" thing. It can make the telephone sound better for miles and miles of cables, makes the tube amps lower distortion, from 1-2% to 0.1%. But read it further, the same feedback becomes "not preferable" in currently called "Hi-End" power amps, because the same feedback produces delay caused distortion, and makes high order distortion. In this later situation, I cannot see "the solution is to use moderate/low feedback", but the option is "use feedback (global feedback) or none at all"
If I stop at the middle of the reading (the part where the feedback is a "good" thing), I wonder if feedback can be applied to speakers also. I read that the most distortive component in audio reproduction is speaker (tweeter, midrange, etc). Speakers are working in "open loop" mode, without feedback at all, so the distorton can achieve 2-5%.
Here I can found Velodyne subwoofer, where the manufacturer claims it is a subwoofer with feedback, so the distortion is lower than ordinary subwoofer.
Has the speaker designers come up with a Tweeter or midrange with feedback? That distorted only 0.1% (maybe).
Or feedback is only applicable in subwoofer only (the frequency is not too fast) with current technology?
Reading about feedback is interesting. If I read from the first of its invention (for telephone use), I get it as a "good" thing. It can make the telephone sound better for miles and miles of cables, makes the tube amps lower distortion, from 1-2% to 0.1%. But read it further, the same feedback becomes "not preferable" in currently called "Hi-End" power amps, because the same feedback produces delay caused distortion, and makes high order distortion. In this later situation, I cannot see "the solution is to use moderate/low feedback", but the option is "use feedback (global feedback) or none at all"
If I stop at the middle of the reading (the part where the feedback is a "good" thing), I wonder if feedback can be applied to speakers also. I read that the most distortive component in audio reproduction is speaker (tweeter, midrange, etc). Speakers are working in "open loop" mode, without feedback at all, so the distorton can achieve 2-5%.
Here I can found Velodyne subwoofer, where the manufacturer claims it is a subwoofer with feedback, so the distortion is lower than ordinary subwoofer.
Has the speaker designers come up with a Tweeter or midrange with feedback? That distorted only 0.1% (maybe).
Or feedback is only applicable in subwoofer only (the frequency is not too fast) with current technology?
lumanauw said:
All amplifiers will produce a different signal at its output to its input regardless of whether they have feedback or not. It is distortion and we cannot eliminate it completely - though we can reduce it to very low levels, possibly so low that it is immeasurable. Zero distortion implies infinite levels of negative feedback, which implies infinite open loop gain (which is impossible).
Historically, negative feedback was shunned because it threw away gain, which was difficult to come by in early days. Then valves with their output transformers limited the amount practicable without making an unstable amp. Only with the (OTL) transistor amp has very large amounts of feedback been practicable.
lumanauw said:
Bringing the loudspeaker into the feedback loop would be good for distoriton figures, but is not particularly easy to achieve in practicality. I think Mr. Evil and SY have tried this. I personally have never bothered. In any case, something which I have read (though have no idea of its truthfulness or otherwise) is that loudspeakers, being mechanical devices produce distortion which less offensive to the ear - but then we're heading down the road of psychoacoustics....
Kuei Yang Wang said:
Also depends on your point of view. Tube designs tend to be single isolated amplification stages, while SS designs tend to be a mess of gain (mainly due to situations which alone are highly unstable, variable constant current devices sitting between fixed current sources and sinks) with NFB which does three things: first, and most importantly, it stabilizes the circuit; two, it reduces gain to a specific value which is useful, and three, a convienient coincidence, it reduces distortion and output impedance to negligible values.
But I digress; this is another argument for another day, stabbing at the very heart of thermionic philosophy and the tendancy towards (relatively) high distortion, and low NFB.
Ah, so feedback, specific - local, one might say - to exactly one one tube, is not local? What odd use of English...
Now I'm very curious, what you DO consider local?
The only thing more electrically "local" must be taking place inside the very tube, but "triodes don't have NFB".
Looped NFB need need reduce gain to unity, that much is clear. (Cathode followers are merely the "full NFB" arrangement of, for instance, CFB output topologies, which have greater than unity gain.)
But are not the more external forms of shunt FB or global FB also controlling electron flow, albeit more indirectly?
Ah, but he did, though not specifically. He laid the groundwork which states how electric fields are ordered inside such a device. Indeed, the very same electric fields produced by the plate of a triode are equal to those produced by the screen grid on the volume contained within. A screen is just a "leaky" plate.
Well my friend... you are welcome to deny that Mr. Maxwell's laws apply, or that electrons don't actually exist (and other quantum theory), or that the sun orbits around the Earth, but there is one interesting thing all the accepted theories have going for them -- they ALL work, in 100% of observed situations.
Tim
Konnichiwa,
Yes, but was it "good enough" to show reliably that the proposed "factor X" was absent?
If your test weas to show that you could distinguish the the Non-NFB Amplifier as distorted compared to the other your test illustrated this amiably.
If you wanted to find out if you preferred the Non-NFB because it introduced distortion or if you preferred it becaused it FAILED to introduce specific types of distortion not covered by traditional amplifier specifications and DESPITE it's other audible colorations, your test was not capable of answering that question.
Therefore, on strict and pure logic, if you take the results of your test as proof that you subjectively prefer distorted amplifers to non-distorted ones you are falling prey to self decption.
I repeat, the test was so aligned that it would reliably yield a certain result, irregardless of real conditions. In that it matches many of the test the DBT/ABX crowd constantly are tubthumping and welkringing about and it has about as much value.
I am disappointed, I would have expected more from you.
Sayonara
SY said:
Yes, but was it "good enough" to show reliably that the proposed "factor X" was absent?
If your test weas to show that you could distinguish the the Non-NFB Amplifier as distorted compared to the other your test illustrated this amiably.
If you wanted to find out if you preferred the Non-NFB because it introduced distortion or if you preferred it becaused it FAILED to introduce specific types of distortion not covered by traditional amplifier specifications and DESPITE it's other audible colorations, your test was not capable of answering that question.
Therefore, on strict and pure logic, if you take the results of your test as proof that you subjectively prefer distorted amplifers to non-distorted ones you are falling prey to self decption.
I repeat, the test was so aligned that it would reliably yield a certain result, irregardless of real conditions. In that it matches many of the test the DBT/ABX crowd constantly are tubthumping and welkringing about and it has about as much value.
I am disappointed, I would have expected more from you.
Sayonara
Konnichiwa,
To quote sporting life from Porgy & Bess: "It ain't neccesarily so!".
Let me propose an amplifier to you.
It has a heavily degenerated cascoded input stage (may be differential, but need not be) where the gain is set to an exact value that makes sense in the scheme of things, the degenerative device is a resistor, the load are resistors.
It has a second gain stage where again the gain is heavily degenerated to give exactly the gain needed with the same design features as above.
It finally uses compound feedback pair output devices in full looped feedback follower mode in Class A.
You will find the resulting Amplifier if competently desiygned to give results in terms of output impedance, distortion and so on that match the same devices applied in a conventional feedback amplifier.
It will show however a few areas where it is invariably superior to the high feedback Amplifier.
1) No or minimal noisefloor modulation with signal.
2) Massive tolerance to "out of band" signals.
3) Instant and gracefull clipping recovery.
4) Absolute and reliable freedom from any transient distortion modes, known (TIM/SIM/PIM) and postulated (Factor X).
Therfore, in my view the competently designed feedback free (or low feedback/short loop type, where the feedback loop terminates at the gainstage, not the input stage) is preferbale to the NFB Type amplifier and in fact represents competent design, while the NFB Amp represenst a design that is either done by a lazy and incompetent designer or by one who simply follows orthodox design approaches without questioning them or understanding the operational principles and implications.
That of course is JUST ME, my personal opinion..
The way you present it one might be misled to believe that looped inverse feedback is neccesary or required to achieve any of the above, which is patently untrue.
Hmm, have you ever wondered how much distortion your speakers and your ears produce as well what sort of spectrum such distortion would have?
So, applying looped inverse feedback around a single stage is "local feedback"? If you wish to use this definition, be my guest. However, if that is so, then I understand COMPLETELY, COMPREHENSIVELY and ENTIERLY how to apply and use local feedback. I thought you said I did not and I asked you to illustrate a mode of "local feedback" I failed to understand.
I am well familiar with both looped inverse feedback (applied to one, two or as many stages as one likes) and degeneration as design tools.
I don't consider "local" looped inverse feedback to be materially different from "global" looped inverse feedback. Therefore I consider the term "local feedback" an oxymoron, used by people with muddy, unclear thinking and a barely tenuous grasp on electronic principles. Modern language, be it technical or colloquial is really fubar and replete with such cases.
Triodes have neither "build in degeneration" nor "build in looped inverse feedback". That is fact. Triodes, but the nature of their operation behave in certain areas analog to a different device with an applied inverse feedback loop, however in other areas they do not.
One may therefore on the evidence conclude that Triodes behave similar to a different with looped inverse feedback applied, but to conclude that "triodes have internal NFB" is a bridge too far. It is a statement dictated by the desire to to people who prefer to use a circuit without looped inverse feedback using triodes that in fact do use such, however as such the statement is motivated by the desire to claim a patently untrue thing to be true, for ideological reasons and is not the least rooted in reliable fact.
I take it you meant to write "Looped NFB needs not to reduce gain to unity". I was refering to the case of the pseudo triode. If we take the screengrid as inverting input in the traditional sense, then connecting it to the anode without intervening voltage devider to set a again above unity should reduce the circuits gain to unity. The fact that it does not illustrates that what happens inside thermionic emission devices is DIFFERENT from traditional external inverse looped feedback.
My point. One (LIF) is an external, deliberate methode, the other (Triode) is merely a device charateristic of the triode, according the various equations by, as you so rightly note Maxwell among others.
Let us be absolutely clear here.
Maxwell NEVER, EVER at any point siad "A triode is a Pentode with internal inverse looped feedback!". Nor did he ever say anything to that effect using a slightly differnet wording.
Do we agree on that?
If so, then your claim that he did in an earlier post was an untrue statement made knowingly and deliberatly (more colloquially known as lie) in an attempt to invoke academic athority as substitute for actual knowledge.
I have no beef with Maxwell, his laws or how they work in thermionic devices.
I do have beef with you putting so to speak "wods into Mr. Maxwells mouth", moreover words that such an intelligent and discerning gentleman as Mr. Maxwell would have never said, knowing them to be untrue.
Have you ever seen one? Can you provide actual, irrufutable prrof that Electrons really are singular particles with an actual eXistenZ in the sense of being absolutely there, indivisible and not composed of smaller particles?
Once Atoms where held to actually exist. We now know they do not exist. They have zero reality. They where merely an illusion of perception at one stage of our enquiry into the nature of the universe. I suspect "Electrons" will be eventually (if they have not been already - which some would note) be shown to belong into the class of "observational error" as the Atom, Neutron and Positron.
I was not aware that sun orbits around the earth, in fact I find such a statement rather in line with "Triodes have internal NFB", in other words a patently and transparently false statement.
I believe I leave this argument at this stage, chalking up a "point not prooven" to your assertations of "Triodes have internal NFB" and "Kuei Yang Wang does not understand local feedback" and indeed to your assertation that there is such a thing as "local feedback" in a sense that makes it distinct from other forms of feedback (inverse or normal).
You may consider to re-word your key statement fo contention to:
"Triodes behave IN CERTAIN ASPECTS alanog or similar to other devices with looped inverse feedback applied, however in other aspects they do not."
This covers reliably and exactly the observation and theory and has two interesting things going for it -- it works, in 100% of observed situations and is in accordance with accepted electronic and physical theory, which is more than can be said for your assertation and I actually agree with it.
Sayonara
Sch3mat1c said:
To quote sporting life from Porgy & Bess: "It ain't neccesarily so!".
Let me propose an amplifier to you.
It has a heavily degenerated cascoded input stage (may be differential, but need not be) where the gain is set to an exact value that makes sense in the scheme of things, the degenerative device is a resistor, the load are resistors.
It has a second gain stage where again the gain is heavily degenerated to give exactly the gain needed with the same design features as above.
It finally uses compound feedback pair output devices in full looped feedback follower mode in Class A.
You will find the resulting Amplifier if competently desiygned to give results in terms of output impedance, distortion and so on that match the same devices applied in a conventional feedback amplifier.
It will show however a few areas where it is invariably superior to the high feedback Amplifier.
1) No or minimal noisefloor modulation with signal.
2) Massive tolerance to "out of band" signals.
3) Instant and gracefull clipping recovery.
4) Absolute and reliable freedom from any transient distortion modes, known (TIM/SIM/PIM) and postulated (Factor X).
Therfore, in my view the competently designed feedback free (or low feedback/short loop type, where the feedback loop terminates at the gainstage, not the input stage) is preferbale to the NFB Type amplifier and in fact represents competent design, while the NFB Amp represenst a design that is either done by a lazy and incompetent designer or by one who simply follows orthodox design approaches without questioning them or understanding the operational principles and implications.
That of course is JUST ME, my personal opinion..
Sch3mat1c said:
The way you present it one might be misled to believe that looped inverse feedback is neccesary or required to achieve any of the above, which is patently untrue.
Sch3mat1c said:
Hmm, have you ever wondered how much distortion your speakers and your ears produce as well what sort of spectrum such distortion would have?
Sch3mat1c said:
So, applying looped inverse feedback around a single stage is "local feedback"? If you wish to use this definition, be my guest. However, if that is so, then I understand COMPLETELY, COMPREHENSIVELY and ENTIERLY how to apply and use local feedback. I thought you said I did not and I asked you to illustrate a mode of "local feedback" I failed to understand.
I am well familiar with both looped inverse feedback (applied to one, two or as many stages as one likes) and degeneration as design tools.
Sch3mat1c said:
I don't consider "local" looped inverse feedback to be materially different from "global" looped inverse feedback. Therefore I consider the term "local feedback" an oxymoron, used by people with muddy, unclear thinking and a barely tenuous grasp on electronic principles. Modern language, be it technical or colloquial is really fubar and replete with such cases.
Sch3mat1c said:
Triodes have neither "build in degeneration" nor "build in looped inverse feedback". That is fact. Triodes, but the nature of their operation behave in certain areas analog to a different device with an applied inverse feedback loop, however in other areas they do not.
One may therefore on the evidence conclude that Triodes behave similar to a different with looped inverse feedback applied, but to conclude that "triodes have internal NFB" is a bridge too far. It is a statement dictated by the desire to to people who prefer to use a circuit without looped inverse feedback using triodes that in fact do use such, however as such the statement is motivated by the desire to claim a patently untrue thing to be true, for ideological reasons and is not the least rooted in reliable fact.
Sch3mat1c said:
I take it you meant to write "Looped NFB needs not to reduce gain to unity". I was refering to the case of the pseudo triode. If we take the screengrid as inverting input in the traditional sense, then connecting it to the anode without intervening voltage devider to set a again above unity should reduce the circuits gain to unity. The fact that it does not illustrates that what happens inside thermionic emission devices is DIFFERENT from traditional external inverse looped feedback.
Sch3mat1c said:
My point. One (LIF) is an external, deliberate methode, the other (Triode) is merely a device charateristic of the triode, according the various equations by, as you so rightly note Maxwell among others.
Sch3mat1c said:
Let us be absolutely clear here.
Maxwell NEVER, EVER at any point siad "A triode is a Pentode with internal inverse looped feedback!". Nor did he ever say anything to that effect using a slightly differnet wording.
Do we agree on that?
If so, then your claim that he did in an earlier post was an untrue statement made knowingly and deliberatly (more colloquially known as lie) in an attempt to invoke academic athority as substitute for actual knowledge.
Sch3mat1c said:
I have no beef with Maxwell, his laws or how they work in thermionic devices.
I do have beef with you putting so to speak "wods into Mr. Maxwells mouth", moreover words that such an intelligent and discerning gentleman as Mr. Maxwell would have never said, knowing them to be untrue.
Sch3mat1c said:
Have you ever seen one? Can you provide actual, irrufutable prrof that Electrons really are singular particles with an actual eXistenZ in the sense of being absolutely there, indivisible and not composed of smaller particles?
Once Atoms where held to actually exist. We now know they do not exist. They have zero reality. They where merely an illusion of perception at one stage of our enquiry into the nature of the universe. I suspect "Electrons" will be eventually (if they have not been already - which some would note) be shown to belong into the class of "observational error" as the Atom, Neutron and Positron.
Sch3mat1c said:
I was not aware that sun orbits around the earth, in fact I find such a statement rather in line with "Triodes have internal NFB", in other words a patently and transparently false statement.
I believe I leave this argument at this stage, chalking up a "point not prooven" to your assertations of "Triodes have internal NFB" and "Kuei Yang Wang does not understand local feedback" and indeed to your assertation that there is such a thing as "local feedback" in a sense that makes it distinct from other forms of feedback (inverse or normal).
You may consider to re-word your key statement fo contention to:
"Triodes behave IN CERTAIN ASPECTS alanog or similar to other devices with looped inverse feedback applied, however in other aspects they do not."
This covers reliably and exactly the observation and theory and has two interesting things going for it -- it works, in 100% of observed situations and is in accordance with accepted electronic and physical theory, which is more than can be said for your assertation and I actually agree with it.
Sayonara
Konnichiwa,
All of this depends upon definitions. You can build a very linear non-feedback system. You can build feedback systems with very good transient performance.
In fact, you can do anything you like and attain any result you desire, if make sure your result is supportable within the general physical framework of this universe.
It is non-trivail to enclose Speaker drivers in feedback loops I played with this in the 1980's and found that I could make it work and that I could get a similar LF performance and low distortion out of a pair of long Throw NFB Applied 8" Woofers that I got from what is best described Klipsch Kornwall Clones with EV Drivers.
Yes, in germany Silbersand has done that.
I personally used current drive in my own experimental active speakers back in the 1980's. That is a pair of 8" Drivers using NFB and a 5" "fullrange driver" as midrange using an "apperiodic" enclosure and current feed and a 1" Suprynol Dome tweeter with current feed.
All this made for a pretty low distortion speaker with no measureable compression in the operational range and with a wide bandwidth. The bad news was that I still preferred my oldfashioned Horn Midrange & Horntweeter plus 15" Woofer passive crossover speaker and my studio EL34 Push-Pull Amplifiers (with quite a lot of NFB BTW) to the new solid state active gizmo filled speaker.
That said the active speaker sounded quite good, quite impressive and was a lot more living friendly. So the big EV Speakers went into the project studio and the active speaker staid in the living room.
Sayonara
lumanauw said:
All of this depends upon definitions. You can build a very linear non-feedback system. You can build feedback systems with very good transient performance.
In fact, you can do anything you like and attain any result you desire, if make sure your result is supportable within the general physical framework of this universe.
lumanauw said:
It is non-trivail to enclose Speaker drivers in feedback loops I played with this in the 1980's and found that I could make it work and that I could get a similar LF performance and low distortion out of a pair of long Throw NFB Applied 8" Woofers that I got from what is best described Klipsch Kornwall Clones with EV Drivers.
lumanauw said:
Yes, in germany Silbersand has done that.
I personally used current drive in my own experimental active speakers back in the 1980's. That is a pair of 8" Drivers using NFB and a 5" "fullrange driver" as midrange using an "apperiodic" enclosure and current feed and a 1" Suprynol Dome tweeter with current feed.
All this made for a pretty low distortion speaker with no measureable compression in the operational range and with a wide bandwidth. The bad news was that I still preferred my oldfashioned Horn Midrange & Horntweeter plus 15" Woofer passive crossover speaker and my studio EL34 Push-Pull Amplifiers (with quite a lot of NFB BTW) to the new solid state active gizmo filled speaker.
That said the active speaker sounded quite good, quite impressive and was a lot more living friendly. So the big EV Speakers went into the project studio and the active speaker staid in the living room.
Sayonara
As Circlotron said, the "not wanted" thing about feedback happens because amp need a little time from input to output. The feedback signal is giving wrong information, because what happened in the output is "history" than what input needs (correction signal) at a certain point of time. This happens again and again like in the hall of mirror, creating high order distortion.
Some said, that minimal stages power amp sounds better, and could this because it needs less time from input to output?
In Error Correction (EC) discussion, we can (just can, but don't do it) create "negative output impedance", where the output signal could be rising when headed to load, not decreasing like in ordinary amp without EC.
I know that time is irreversible, cannot make "negative time". But has someone came up with idea how to handle this "time propagation needed" so that feedback can gives exact signal to input stage? Or it cannot be done, so we are left only with "don't use feedback / non global feedback" type of design?
Some said, that minimal stages power amp sounds better, and could this because it needs less time from input to output?
In Error Correction (EC) discussion, we can (just can, but don't do it) create "negative output impedance", where the output signal could be rising when headed to load, not decreasing like in ordinary amp without EC.
I know that time is irreversible, cannot make "negative time". But has someone came up with idea how to handle this "time propagation needed" so that feedback can gives exact signal to input stage? Or it cannot be done, so we are left only with "don't use feedback / non global feedback" type of design?
"By definition of your terms, it can't be done. Feedback is feed "back." Back is back is back is back."
Feedback only injects a signal sample into an earlier amplification stage. It does not stop time from moving forward.
As an example, you can go back to your hometown. You can't go back to your hometown as it was when you were younger.
After all is said and done, inverse feedback is a valuable option.
It's not a perfect solution.
Feedback only injects a signal sample into an earlier amplification stage. It does not stop time from moving forward.
As an example, you can go back to your hometown. You can't go back to your hometown as it was when you were younger.
After all is said and done, inverse feedback is a valuable option.
It's not a perfect solution.
Konnichiwa,
Well, you may think of inverse looped feedback as a magic trick (speaking of stage magic as in making something appear what it is not - not magick as in actual alteration of reality).
With a magic trick, as long the hand is sufficiently faster then the eye, the "trick" works. Make it too slow, the trick does not work.
Therefore the structure inside the feedback loop must be VERY FAST. There are a number of approaches that can all be used to make loops faster.
One option is for example to take the miller compenation of the amplifiers frequency response from the output stage instead from the Voltage amplifier output. This makes for a "short" and "fast" loop. Simmilar aproaches apply to valve Amplifiers.
Most people are struck by how my Pentode/Tetrode Push-Pull Amplifier with around 26db negative feedback sounds so much alike to my 300B SE Amplifier(s). They sound much more alike than they sound different when compared to traditionally implemented Amp's.
It illustrates how much is in the implementation. One Amp has zero NFB and well over 1% THD at 1 Watt output, plus it class A single ended triode. The other has around 26db loop feedback and around 0.1% THD at 1Watt and is class AB push pull tetrode or pentode. They should sound drastically different, but they sound very close....
Sayonara
lumanauw said:
Well, you may think of inverse looped feedback as a magic trick (speaking of stage magic as in making something appear what it is not - not magick as in actual alteration of reality).
With a magic trick, as long the hand is sufficiently faster then the eye, the "trick" works. Make it too slow, the trick does not work.
Therefore the structure inside the feedback loop must be VERY FAST. There are a number of approaches that can all be used to make loops faster.
One option is for example to take the miller compenation of the amplifiers frequency response from the output stage instead from the Voltage amplifier output. This makes for a "short" and "fast" loop. Simmilar aproaches apply to valve Amplifiers.
Most people are struck by how my Pentode/Tetrode Push-Pull Amplifier with around 26db negative feedback sounds so much alike to my 300B SE Amplifier(s). They sound much more alike than they sound different when compared to traditionally implemented Amp's.
It illustrates how much is in the implementation. One Amp has zero NFB and well over 1% THD at 1 Watt output, plus it class A single ended triode. The other has around 26db loop feedback and around 0.1% THD at 1Watt and is class AB push pull tetrode or pentode. They should sound drastically different, but they sound very close....
Sayonara
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.