Hi Michael,
I'm going through Post#128. Discussion over ?
An EMF is a source of electrical energy that causes current to flow in a circuit.
A charged capacitor can be a source of EMF, and a capacitor can be charged by EMF.
We need to be able to 'think' EMF to have a regard for its nature and the way it provides for current flow through sometimes resistive and sometimes reactive circuits.
Thus you should not advise everyone that the term 'EMF' ought not be used just because you would rather stick to talking about 'reactive loads'.
I can see your point, but a reactive load is one where the current and voltage are not in phase; and yet loudspeaker current and voltage are in phase at some frequencies, such that the load cannot then be described as being reactive. It is even possible for loudspeaker system back-EMF to be completely in or out of phase at frequencies when the loudspeaker is not reactive, it then has a modified but wholly real current draw.
You also discuss what happens to an amplifier subject to back-EMF, and then conclude that there is no real problem.
But what happens when capacitive devices and stabilisation components are enclosed within a global NFB loop ?
Answer;- the amplifier's output terminal becomes inductive = reactive; ie. not passive.
And the greater the open loop gain, the lower the frequency the stabilisation components act, and thus the lower the audio frequency at which the output characteristic becomes inductive, though the lower the resistive component too.
Amplifier output terminal current flow is modified by loudspeaker system generated back-EMF. Loudspeaker sytem back-EMF comprises a complex mix of crossover component back-EMFs and driver/cabinet induced back-EMFs.
Changes in loudspeaker current flow can develop a leading error voltage across the amplifier's internal inductance, and this error voltage is unavoidably in series with the NFB loop controlled output.
The error voltage is directly related to changes in output terminal current flow, not input, and it increases with frequency. It is modified during output cycles as complex induced back-EMF changes follow drive.
That leading error voltage affects the nature of high frequency reproduction; often it produces a falsely bright sound that might actually impress, but it is most definitely not accurate. Some designers attempt to reduce this tendency by fitting an input filter, but that still cannot prevent the loudspeaker induced output error from arising, and the more reactive/efficient the loudspeaker, the more pronounced the error.
I illustrated a simulated outcome in Post#175 of Lumanauw's thread. The sudden increase in error voltage due to leading load current flow can fractionally 'reverse drive' an output stage through a portion of its bias potential before the global loop can minimise the error.
Post#130. Yes, and if we can manage to build more stable amplifiers we can make sure that they have reduced internal output inductance.
Cheers .......... Graham.
(Hopefully this feedback not causing instability.)
I'm going through Post#128. Discussion over ?
An EMF is a source of electrical energy that causes current to flow in a circuit.
A charged capacitor can be a source of EMF, and a capacitor can be charged by EMF.
We need to be able to 'think' EMF to have a regard for its nature and the way it provides for current flow through sometimes resistive and sometimes reactive circuits.
Thus you should not advise everyone that the term 'EMF' ought not be used just because you would rather stick to talking about 'reactive loads'.
I can see your point, but a reactive load is one where the current and voltage are not in phase; and yet loudspeaker current and voltage are in phase at some frequencies, such that the load cannot then be described as being reactive. It is even possible for loudspeaker system back-EMF to be completely in or out of phase at frequencies when the loudspeaker is not reactive, it then has a modified but wholly real current draw.
You also discuss what happens to an amplifier subject to back-EMF, and then conclude that there is no real problem.
But what happens when capacitive devices and stabilisation components are enclosed within a global NFB loop ?
Answer;- the amplifier's output terminal becomes inductive = reactive; ie. not passive.
And the greater the open loop gain, the lower the frequency the stabilisation components act, and thus the lower the audio frequency at which the output characteristic becomes inductive, though the lower the resistive component too.
Amplifier output terminal current flow is modified by loudspeaker system generated back-EMF. Loudspeaker sytem back-EMF comprises a complex mix of crossover component back-EMFs and driver/cabinet induced back-EMFs.
Changes in loudspeaker current flow can develop a leading error voltage across the amplifier's internal inductance, and this error voltage is unavoidably in series with the NFB loop controlled output.
The error voltage is directly related to changes in output terminal current flow, not input, and it increases with frequency. It is modified during output cycles as complex induced back-EMF changes follow drive.
That leading error voltage affects the nature of high frequency reproduction; often it produces a falsely bright sound that might actually impress, but it is most definitely not accurate. Some designers attempt to reduce this tendency by fitting an input filter, but that still cannot prevent the loudspeaker induced output error from arising, and the more reactive/efficient the loudspeaker, the more pronounced the error.
I illustrated a simulated outcome in Post#175 of Lumanauw's thread. The sudden increase in error voltage due to leading load current flow can fractionally 'reverse drive' an output stage through a portion of its bias potential before the global loop can minimise the error.
Post#130. Yes, and if we can manage to build more stable amplifiers we can make sure that they have reduced internal output inductance.
Cheers .......... Graham.
(Hopefully this feedback not causing instability.)
Graham Maynard said:
Hello, Graham
My post re transmission lines was related to your posting above.
You've cited the cables...
I have not found yet your post re the 0.22 ohms resistor.
Hi Jorge,
Thanks.
The cable is part of the loudspeaker system, but not necessarily a transmission line.
Cheers ...... Graham.
Thanks.
The cable is part of the loudspeaker system, but not necessarily a transmission line.
Cheers ...... Graham.
Hi Jorge,
And many thanks for your question; it has triggered from me this response.
If your question had not been posed in such a simple way I could not have come up with this explanation of my understanding.
This is what I wanted to discuss with Jan, but I was having real difficulty in getting the wordflow. Here goes;-
Music waveform from an amplifier starts to generate current flow through a loudspeaker system.
The system comprises cable, crossover inductors, capacitors, drivers, air spring/vent etc.
The loudspeaker system starts to generate back-EMF that is related to the first 90 degrees or leading edge of input, as seen by each crossover section/driver.
All of the inductors start to generate magnetic fields as current flow begins, and throughput is further waveform fraction delayed as capacitors develop voltage charge.
Then the audio signal changes the amplifier's output waveform that the loudspeaker has already started developing fields/voltages/back-EMFs for.
The embrionic back-EMFs that should have led to a combined steady sinewave with the leading/lagging current flow we might study in time isolation, are no longer appropriate.
By conservation of energy these real-time generated fields/voltages/back-EMFs cannot simply disappear when the waveform changes.
The NFB loop generated amplifier output impedance is so low that their energies become trapped within the cable/loudspeaker system.
These energies now modify (ring against) on-going amplifier output within the cable/loudspeaker system and modify the new audio waveform by developing error voltages across reactive components.
This happens on an on-going basis.
Amplifier output resistance or series output resistance will damp the ringing, without amplifier resistance or the series resistance reactively generating any new phase shifted voltage error.
(This is where amplifier output inductance can cause additional phase shifted voltage error wrt the input referenced and NFB loop controlled output node.)
Loudspeaker cables generally have low resistance, but they do have a reactive capacitance/inductance characteristic.
They thus develop a fractionally different and real-time reactively modified voltage at the loudspeaker end, when compared to an ultra-low impedance amplifier driven end, as they conduct complex audio currents to a complex and reacting loudspeaker load.
_______________________________________
Cue experiments. I am not in a position to try these.
Does anyone have a proper real-world equivalent loudspeaker load comprising only RL and C that won't make a noise.
Feed this dummy loudspeaker via a cable from a low impedance output solid-state amplifier; connect an isolated monitoring amplifier input across one of the conductor's ends.
I suggest you will hear dynamically (non-steady sinewave) induced error and it will be considerable at crossover frequencies where the components develop greatest fields and voltages, also at frequencies where the combined dummy system generates fully opposing back-EMF wrt on-going amplifier output, and that those error voltages will increase at hf due cable series inductance.
Then substitue a tube amplifier or try a series 0.22 ohm resistor in series with the solid state amplifier (but keep it external to the monitoring amplifier input) and the I suggest the reduction in system Q will very much reduce 'cable induced' waveform error.
Drive a dummy resistor load and the error will fall to extremely low level.
This is the first time I have attempted to write anything like this, so it might well need correction/modification.
Cheers ........ Graham.
And many thanks for your question; it has triggered from me this response.
If your question had not been posed in such a simple way I could not have come up with this explanation of my understanding.
This is what I wanted to discuss with Jan, but I was having real difficulty in getting the wordflow. Here goes;-
Music waveform from an amplifier starts to generate current flow through a loudspeaker system.
The system comprises cable, crossover inductors, capacitors, drivers, air spring/vent etc.
The loudspeaker system starts to generate back-EMF that is related to the first 90 degrees or leading edge of input, as seen by each crossover section/driver.
All of the inductors start to generate magnetic fields as current flow begins, and throughput is further waveform fraction delayed as capacitors develop voltage charge.
Then the audio signal changes the amplifier's output waveform that the loudspeaker has already started developing fields/voltages/back-EMFs for.
The embrionic back-EMFs that should have led to a combined steady sinewave with the leading/lagging current flow we might study in time isolation, are no longer appropriate.
By conservation of energy these real-time generated fields/voltages/back-EMFs cannot simply disappear when the waveform changes.
The NFB loop generated amplifier output impedance is so low that their energies become trapped within the cable/loudspeaker system.
These energies now modify (ring against) on-going amplifier output within the cable/loudspeaker system and modify the new audio waveform by developing error voltages across reactive components.
This happens on an on-going basis.
Amplifier output resistance or series output resistance will damp the ringing, without amplifier resistance or the series resistance reactively generating any new phase shifted voltage error.
(This is where amplifier output inductance can cause additional phase shifted voltage error wrt the input referenced and NFB loop controlled output node.)
Loudspeaker cables generally have low resistance, but they do have a reactive capacitance/inductance characteristic.
They thus develop a fractionally different and real-time reactively modified voltage at the loudspeaker end, when compared to an ultra-low impedance amplifier driven end, as they conduct complex audio currents to a complex and reacting loudspeaker load.
_______________________________________
Cue experiments. I am not in a position to try these.
Does anyone have a proper real-world equivalent loudspeaker load comprising only RL and C that won't make a noise.
Feed this dummy loudspeaker via a cable from a low impedance output solid-state amplifier; connect an isolated monitoring amplifier input across one of the conductor's ends.
I suggest you will hear dynamically (non-steady sinewave) induced error and it will be considerable at crossover frequencies where the components develop greatest fields and voltages, also at frequencies where the combined dummy system generates fully opposing back-EMF wrt on-going amplifier output, and that those error voltages will increase at hf due cable series inductance.
Then substitue a tube amplifier or try a series 0.22 ohm resistor in series with the solid state amplifier (but keep it external to the monitoring amplifier input) and the I suggest the reduction in system Q will very much reduce 'cable induced' waveform error.
Drive a dummy resistor load and the error will fall to extremely low level.
This is the first time I have attempted to write anything like this, so it might well need correction/modification.
Cheers ........ Graham.
in case someone's forgotten 20kHz sine has over 10 km vawelength in a cable and 15 km in vacuum.
Sorry Graham, but your theory has been exposed many times in the past on diverse forms - it's known as Moto Perpetuum or Moto Continuo.
At the very moment music start to generate current, losses takes its tolls.
Be it wiring resistence, inductor resistence, capacitors dielectric losses and resistence, low speaker efficiencies, hysteresis, box losses, etc, energy is lost, usually in the form of heath.
About delays due to capacitor and inductors charging and opposing the current flow - that's exactly how they are expected to behave, and when you combine their respective properties there is what is called ressonance, and the ratio of supplied and stored energy is known as circuit "Q" - quality factor.
BTW, ressonance is not an electrical property - it happens in mechanical (non eletricity driven objects) and even individual atoms (that's the base for an atomic clock)
So, to see your back EMF in action without noises, one doesn't needs a speaker - a simple L-R-C circiut will produce the same results.
Besides, although you refuse to see the facts, the tiny electrical generator part of the so called back EMF of speakers (yes, there is some) will not contribute to distortions.
And when the out of phase currents produced by ressonance interact with a generator - be it an AC mechanical power generator (have you considered electric motors are moving inductors?), a tube amp or a SS amp, the generator may oppose this out of phase currents and produce damping - that BTW may be perfectly a mechanical only effect, like suspension losses in a speaker, or the shocks in your car, but always based in losses.
And the 'high dampng factor' of SS amps is produced by generating an almost equal but opposite polarity wave - so there is kind of a controlled short circuit, that as is well known generates heath - losses, the same that happens in a SE tube amp - plate losses. Just the quantity is different.
Wrapping up, your pet theory doesn't holds based on basic physics principles.
Is there a difference in sound produced by tube and SS amps?
You bet.
Each one has it's 'sound signature' , as red wine has 'taste signature'..
The fact that you like one kind doesn't means the other is wrong - and even less justify vapor theories to explain why.
Regards,
Jorge
At the very moment music start to generate current, losses takes its tolls.
Be it wiring resistence, inductor resistence, capacitors dielectric losses and resistence, low speaker efficiencies, hysteresis, box losses, etc, energy is lost, usually in the form of heath.
About delays due to capacitor and inductors charging and opposing the current flow - that's exactly how they are expected to behave, and when you combine their respective properties there is what is called ressonance, and the ratio of supplied and stored energy is known as circuit "Q" - quality factor.
BTW, ressonance is not an electrical property - it happens in mechanical (non eletricity driven objects) and even individual atoms (that's the base for an atomic clock)
So, to see your back EMF in action without noises, one doesn't needs a speaker - a simple L-R-C circiut will produce the same results.
Besides, although you refuse to see the facts, the tiny electrical generator part of the so called back EMF of speakers (yes, there is some) will not contribute to distortions.
And when the out of phase currents produced by ressonance interact with a generator - be it an AC mechanical power generator (have you considered electric motors are moving inductors?), a tube amp or a SS amp, the generator may oppose this out of phase currents and produce damping - that BTW may be perfectly a mechanical only effect, like suspension losses in a speaker, or the shocks in your car, but always based in losses.
And the 'high dampng factor' of SS amps is produced by generating an almost equal but opposite polarity wave - so there is kind of a controlled short circuit, that as is well known generates heath - losses, the same that happens in a SE tube amp - plate losses. Just the quantity is different.
Wrapping up, your pet theory doesn't holds based on basic physics principles.
Is there a difference in sound produced by tube and SS amps?
You bet.
Each one has it's 'sound signature' , as red wine has 'taste signature'..
The fact that you like one kind doesn't means the other is wrong - and even less justify vapor theories to explain why.
Regards,
Jorge
Hi Darkfenriz,
Please look at my words. Then your comment.
Where do I talk about wavelength, and where does wavelength come into it ???
Cheers .......... Graham.
Please look at my words. Then your comment.
Where do I talk about wavelength, and where does wavelength come into it ???
Cheers .......... Graham.
Sorry Graham, don't feel offended.
I just wanted to suggest that transmision lines are considered for cable of more than 1/10 vawelength. So I get angry when I read about transmision lines (in electric, not acoustic sense) in audio forum.
I just wanted to suggest that transmision lines are considered for cable of more than 1/10 vawelength. So I get angry when I read about transmision lines (in electric, not acoustic sense) in audio forum.
Hi Darkfentiz,
I am not in any way offended.
I haven't talked about transmission lines either, only cable capacitance, inductance and resistance.
The opening line of Jorge's response means that I am not even going to read it until later.
Cheers ......... Graham.
I am not in any way offended.
I haven't talked about transmission lines either, only cable capacitance, inductance and resistance.
The opening line of Jorge's response means that I am not even going to read it until later.
Cheers ......... Graham.
Hi all, This matter is a lot of nonelementary and is clear that have not answered there simple.
The Maynard's posts are a lot of particular, and personally doesn't always share his conclusions or his analysis, but this depends much from the personal back ground. But not me exalt the "classics" conclusions typical of this environment: " the sound signature...".
"Philosophy and tradition" against "creative analysis". who will win?
Ciao
Mauro
The Maynard's posts are a lot of particular, and personally doesn't always share his conclusions or his analysis, but this depends much from the personal back ground. But not me exalt the "classics" conclusions typical of this environment: " the sound signature...".
"Philosophy and tradition" against "creative analysis". who will win?
Ciao
Mauro
If the concern is only with objective issues, then it just becomes a matter of determining what's true and what's not. The trouble I have with Graham's posts is that it's nearly impossible to determine whether they are true or false because they consist mostly of technical non-sequiturs. He has criticised his detractors for guessing what he's saying, but in the case of technically incoherent posts there is little choice but to guess.
Attempts at getting clarification seem to result in more of the same thing, causing an endless cycle of confusion and misinformation.
This is too bad, because his class A/AB/B amplifier is very cool.
Attempts at getting clarification seem to result in more of the same thing, causing an endless cycle of confusion and misinformation.
This is too bad, because his class A/AB/B amplifier is very cool.
Hi Jorge,
I'm going though your Post#147
Moto Perpetuum - Moto Continuo.
If this is perpetual motion, well of course there is no such thing,
If you are applying this to what I have just written then please tell me where the perpetual motion supposedly arises.
I always start at t=0, and the energy is always dissipated: Nothing perpetual there !
Yes, I agree with you that there are always losses.
Yes, I agree with you L,C resonance and Q, etc. (but lets keep it to audio )
Now you say I "refuse to see the facts".
What facts ?
Please state what it is I refuse to see.
You say back-EMF will not contribute to distortions.
I believe you are wrong to make such a claim.
A ported loudspeaker is a mechanical equivalent to electrical back-EMF.
A ported loudspeaker examined with steady sinewave in time isolation produces useful results; but a ported loudspeaker driven by constantly changing assymmetrical audio waveform introduces resonant error, thus the resultant output is a distorted copy of the original.
Next para, I'm now reading third time and wonder what you are saying !
Okay a generator might damp resonance. Yes.
Next para, yes, a solid state amplifier presents like a NFB loop generated short circuit. Yes, tube amp plate losses - different damping.
Next line, here we go with pet theory again. What pet theory ?
Theory is fundamental, and I am not trying to change anything.
Gee is that it, a personal jibe accusing me of saying that something is wrong because I prefer something else.
Is there something wrong with with my Post#145 or not ?
That is what counts here.
If yes, then what ?
You say what I write does not hold up to basic principles.
What did you find to be wrong ?
______________________________________________
Pointers;-
A tweeter in series with a capacitor is a series tuned circuit.
A bass crossover section is often an inductor in series with a capacitor, the driver being connected across the capacitor.
The 'Q' of a series tuned circuit is maximum when the driving (amplifier output) impedance is minimum.
Those circuits can be energised, and still have resonant energy after the audio has moved on.
The mid point of those series tuned circuits cannot be fully damped because to do so would prevent reasonable loudspeaker operation.
A solid state amplifier with ultra low output impedance maintains series resonant section 'Q', where a tube amplifier does not.
You say the damping quantity is different, but it is so different, and generally more than two orders different, as to be totally different.
The low output impedance of a voltage driving solid-state NFB amplifier locks energy into the cable/crossover/loudspeaker system, and energy remains *after* the audio has moved on; just as happens with a ported loudspeaker.
This loudspeaker system trapped energy distorts reproduction. Period.
It is not the amplifier that generates the different sound, but the way it is being used.
This would happen with current drive if there were any parallel tuned loudspeaker system circuits, but it would not happen with a non-global feedback designs as long as it did not have too low an output impedance.
The difference in sound with a SS NFB amp driving some loudspeakers when compared to a tube amp is so great as to be much worse than the 1% (or more) distortion possible with a SE tube amplifier, but it is not the SS amplifier that is distorting.
Hi Andy,
Incoherent posts ?
Its the easiest thing in the world for a reader to blame a writer if they do not comprehend, but the writer can do nothing about it if the reader merely comments about the writer, and not the text itself !!!
Please understand one thing though, in the absence of other similar contributions at least I am genuinely trying.
The endless confusion is actually everybody's problem, though some don't seem to realise it.
Maybe I should just give up........
Cheers .......... Graham.
I'm going though your Post#147
Moto Perpetuum - Moto Continuo.
If this is perpetual motion, well of course there is no such thing,
If you are applying this to what I have just written then please tell me where the perpetual motion supposedly arises.
I always start at t=0, and the energy is always dissipated: Nothing perpetual there !
Yes, I agree with you that there are always losses.
Yes, I agree with you L,C resonance and Q, etc. (but lets keep it to audio )
Now you say I "refuse to see the facts".
What facts ?
Please state what it is I refuse to see.
You say back-EMF will not contribute to distortions.
I believe you are wrong to make such a claim.
A ported loudspeaker is a mechanical equivalent to electrical back-EMF.
A ported loudspeaker examined with steady sinewave in time isolation produces useful results; but a ported loudspeaker driven by constantly changing assymmetrical audio waveform introduces resonant error, thus the resultant output is a distorted copy of the original.
Next para, I'm now reading third time and wonder what you are saying !
Okay a generator might damp resonance. Yes.
Next para, yes, a solid state amplifier presents like a NFB loop generated short circuit. Yes, tube amp plate losses - different damping.
Next line, here we go with pet theory again. What pet theory ?
Theory is fundamental, and I am not trying to change anything.
Gee is that it, a personal jibe accusing me of saying that something is wrong because I prefer something else.
Is there something wrong with with my Post#145 or not ?
That is what counts here.
If yes, then what ?
You say what I write does not hold up to basic principles.
What did you find to be wrong ?
______________________________________________
Pointers;-
A tweeter in series with a capacitor is a series tuned circuit.
A bass crossover section is often an inductor in series with a capacitor, the driver being connected across the capacitor.
The 'Q' of a series tuned circuit is maximum when the driving (amplifier output) impedance is minimum.
Those circuits can be energised, and still have resonant energy after the audio has moved on.
The mid point of those series tuned circuits cannot be fully damped because to do so would prevent reasonable loudspeaker operation.
A solid state amplifier with ultra low output impedance maintains series resonant section 'Q', where a tube amplifier does not.
You say the damping quantity is different, but it is so different, and generally more than two orders different, as to be totally different.
The low output impedance of a voltage driving solid-state NFB amplifier locks energy into the cable/crossover/loudspeaker system, and energy remains *after* the audio has moved on; just as happens with a ported loudspeaker.
This loudspeaker system trapped energy distorts reproduction. Period.
It is not the amplifier that generates the different sound, but the way it is being used.
This would happen with current drive if there were any parallel tuned loudspeaker system circuits, but it would not happen with a non-global feedback designs as long as it did not have too low an output impedance.
The difference in sound with a SS NFB amp driving some loudspeakers when compared to a tube amp is so great as to be much worse than the 1% (or more) distortion possible with a SE tube amplifier, but it is not the SS amplifier that is distorting.
Hi Andy,
Incoherent posts ?
Its the easiest thing in the world for a reader to blame a writer if they do not comprehend, but the writer can do nothing about it if the reader merely comments about the writer, and not the text itself !!!
Please understand one thing though, in the absence of other similar contributions at least I am genuinely trying.
The endless confusion is actually everybody's problem, though some don't seem to realise it.
Maybe I should just give up........
Cheers .......... Graham.
Andy_c
"This is too bad, because his class A/AB/B amplifier is very cool."
what i liked most about his amplifier was the one output transistor feeding from the vas emitter resistor, carlos
mentioned it got hot as hell and was there to change mode
to class-a
"This is too bad, because his class A/AB/B amplifier is very cool."
what i liked most about his amplifier was the one output transistor feeding from the vas emitter resistor, carlos
mentioned it got hot as hell and was there to change mode
to class-a
Hi Mauro and Andy,
Something I learned long ago, you cannot 'know' anything until you can explain it to yourself on paper. Symbols are not the same as publishable words.
It has never been the case that I don't share, sure my circuits are here, but then they are self explanatory.
It was just that I was never able to get the 'words' sufficient for me to share knowledge, and complaints from others who don't have the 'words' either, is grossly unfair.
Yes the class-A/AB/B appears to have potential, but it did not sound right by the time I got to 42V rails, I think due to the fact that each 'B' transistor was changing gain and stability on a per half conduction basis. I might well have been able to overcome it, but by then my circuit had become more complex than intuition suggested was necessary.
My class-A//AB sounds fine, with nicely-low low-level distortion plus mostly resistive damping to minimise internally induced error voltage generation from leading back-EMFs when playing loudly.
Hi Jorge,
My A//AB amplifier has a low output impedance.
One of my three-ways is bothered by the low source resistance and its treble is dynamically rendered 'too energetic'.
A series 0.22 ohm output resistor literally 'switches' that loudspeaker trapped energy 'off', whilst the other loudspeaker shows me that the amplifier itself is fine and thus not the problem.
As I said at the time, Post#145 was my first attempt at explanation.
It contained the essence of my Post#153, as well as this below.
Maybe too much in one go, but as far as I am aware, all based upon 'basic principles'
Another pointer,
Parallel tuned circuits give a high impedance resonance peak;- voltage.
Series tuned (loudspeaker) circuits give a low impedance resonance peak;- current.
With crossovers, those series tuned resonances produce greatest current flow when the amplifier has a low output impedance. They are sufficient to induce fractional voltage drop across loudspeaker cable impedance wrt the amplifer end; different cable - different fractional voltage.
Loudspeaker system resonant energy dissipates after music has moved on, but it still induces within the cable a voltage greater than amplifier thd, and thus there is a change in the naturality of reproduction. With cable inductance, that voltage increases with energising frequency, and with so many 'hi-fi' speaker crossovers being at cable end, it is possible for the cable induced voltage caused by the back-EMFs of one crossover section (mid), to modify amplifier drive to another (treble).
Not only that, but there is the initial waveform current induced cable voltage error as the amplifier energises the composite loudspeaker system in the first place, and it is always greatest at hf.
For best reproduction bi or tri wiring should be used with a single solid-state NFB amplifier to overcome this very common 'common-cable' induced voltage error.
Cheers .......... Graham.
Something I learned long ago, you cannot 'know' anything until you can explain it to yourself on paper. Symbols are not the same as publishable words.
It has never been the case that I don't share, sure my circuits are here, but then they are self explanatory.
It was just that I was never able to get the 'words' sufficient for me to share knowledge, and complaints from others who don't have the 'words' either, is grossly unfair.
Yes the class-A/AB/B appears to have potential, but it did not sound right by the time I got to 42V rails, I think due to the fact that each 'B' transistor was changing gain and stability on a per half conduction basis. I might well have been able to overcome it, but by then my circuit had become more complex than intuition suggested was necessary.
My class-A//AB sounds fine, with nicely-low low-level distortion plus mostly resistive damping to minimise internally induced error voltage generation from leading back-EMFs when playing loudly.
Hi Jorge,
My A//AB amplifier has a low output impedance.
One of my three-ways is bothered by the low source resistance and its treble is dynamically rendered 'too energetic'.
A series 0.22 ohm output resistor literally 'switches' that loudspeaker trapped energy 'off', whilst the other loudspeaker shows me that the amplifier itself is fine and thus not the problem.
As I said at the time, Post#145 was my first attempt at explanation.
It contained the essence of my Post#153, as well as this below.
Maybe too much in one go, but as far as I am aware, all based upon 'basic principles'
Another pointer,
Parallel tuned circuits give a high impedance resonance peak;- voltage.
Series tuned (loudspeaker) circuits give a low impedance resonance peak;- current.
With crossovers, those series tuned resonances produce greatest current flow when the amplifier has a low output impedance. They are sufficient to induce fractional voltage drop across loudspeaker cable impedance wrt the amplifer end; different cable - different fractional voltage.
Loudspeaker system resonant energy dissipates after music has moved on, but it still induces within the cable a voltage greater than amplifier thd, and thus there is a change in the naturality of reproduction. With cable inductance, that voltage increases with energising frequency, and with so many 'hi-fi' speaker crossovers being at cable end, it is possible for the cable induced voltage caused by the back-EMFs of one crossover section (mid), to modify amplifier drive to another (treble).
Not only that, but there is the initial waveform current induced cable voltage error as the amplifier energises the composite loudspeaker system in the first place, and it is always greatest at hf.
For best reproduction bi or tri wiring should be used with a single solid-state NFB amplifier to overcome this very common 'common-cable' induced voltage error.
Cheers .......... Graham.
Hi All,
Anyone who wants to check out my above text could simulate it.
Ideal amplifier;
driving two cables having R and L;
one driving 12dB/oct tweeter crossover section plus tweeter;
other a bass/mid 12dB/oct section plus same tweeter circuit.
Place voltmeter between tweeter (+)s and feed amplifier with changing signal - not steady sine !
This same arrangement will show how a series output choke distorts changing waveforms.
I did this back in 2003, but cannot find it now.
Cheers .......... Graham.
Anyone who wants to check out my above text could simulate it.
Ideal amplifier;
driving two cables having R and L;
one driving 12dB/oct tweeter crossover section plus tweeter;
other a bass/mid 12dB/oct section plus same tweeter circuit.
Place voltmeter between tweeter (+)s and feed amplifier with changing signal - not steady sine !
This same arrangement will show how a series output choke distorts changing waveforms.
I did this back in 2003, but cannot find it now.
Cheers .......... Graham.
Graham, doesn't worry you.
For me it is normal doesn't understand or doesn't share any hypothesis, this does part of all the cycles of learning and analysis.
For what concerns me, am using some of your test on my circuits.
Your " reverse driven " and the consequent damping linearizations, are the good results even on my multi NFB circuits...
Ciao
Mauro
For me it is normal doesn't understand or doesn't share any hypothesis, this does part of all the cycles of learning and analysis.
For what concerns me, am using some of your test on my circuits.
Your " reverse driven " and the consequent damping linearizations, are the good results even on my multi NFB circuits...
Ciao
Mauro
If you want see the ameliorations that have introduced creditable to your test...
http://www.diyaudio.com/forums/showthread.php?postid=683428#post683428
Ciao
Mauro
http://www.diyaudio.com/forums/showthread.php?postid=683428#post683428
Ciao
Mauro
Graham Maynard said:
What about choke distortion? Why is it bad? Many would consider it not a problem as it is linear distortion and mainly outside audio band.
regards
- Status
- Not open for further replies.