Glen and Bonsai, you are not adding anything substantial to the discussion. Either prove us wrong or give up the criticism.
john curl said:
Well, I'd claim otherwise. I think I've largely dispensed with your 'audible ringing' claim.
A resonate circuit rings at its resonate frequency. This is elementary, dear Watson.
Suppose we have a 2uH inductor and a 1uF load.
2uH and 1uF have a series resonate frequency of 113kHz.
I can't hear ringing at 113kHz.
Under what conditions can you hear it?
Cheers,
Glen
Folks, it may be that this thread is coming to an end.
I would like to give a summary of what we know about output coils and their audibility in power amps.
First of all, output coils have been used for at least 40 years, maybe longer. It is well documented that the Crown DC-300, a popular power amp back in 1970 used a 2uH output coil that was paralleled by some sort of power resistor. This resistor initially used to be about 10 ohms.
Given the output devices available, and the need for high feedback to control crossover distortion to some reasonable value, this was not much of a tradeoff.
As TIM distortion became recognized to be somewhat important, the slew rates of power amplifiers tended to increase. Even improved amp design techniques and somewhat faster power transistors (2-4MHz), the 2uH output coil was considered useful and necessary, although some experimented with the parallel resistor and reduced it in value to perhaps 1 ohm. This better damps the coil and keeps it from ringing.
As time passed and output devices became even faster, people started to reduce the output coil to a smaller value, such as .5-1uH. I know that Dr. Hawksford was doing this many years ago.
Unfortunately, the coil does tend to make the amplifier more stable under difficult loads, and it takes a bit of effort to keep the amp stable, but it can be done.
Is a 0.5uH coil really audible? Probably it would be hard to prove, but IF you can remove the coil, then why not? Can people hear a 2uH or larger coil. I think so, and probably as much because of what Graham is talking about on this website, as much as any ringing or any other interaction with the loudspeaker load.
Our advice is to get rid of the coil, if you have the design sophistication to do so. However, if you need a coil, then maybe 0.5uH with a 1 ohm or smaller resistor is probably OK. It is up to each designer.
I have not used an output coil in my designs for the last 15 years, but your results may vary.
I would like to give a summary of what we know about output coils and their audibility in power amps.
First of all, output coils have been used for at least 40 years, maybe longer. It is well documented that the Crown DC-300, a popular power amp back in 1970 used a 2uH output coil that was paralleled by some sort of power resistor. This resistor initially used to be about 10 ohms.
Given the output devices available, and the need for high feedback to control crossover distortion to some reasonable value, this was not much of a tradeoff.
As TIM distortion became recognized to be somewhat important, the slew rates of power amplifiers tended to increase. Even improved amp design techniques and somewhat faster power transistors (2-4MHz), the 2uH output coil was considered useful and necessary, although some experimented with the parallel resistor and reduced it in value to perhaps 1 ohm. This better damps the coil and keeps it from ringing.
As time passed and output devices became even faster, people started to reduce the output coil to a smaller value, such as .5-1uH. I know that Dr. Hawksford was doing this many years ago.
Unfortunately, the coil does tend to make the amplifier more stable under difficult loads, and it takes a bit of effort to keep the amp stable, but it can be done.
Is a 0.5uH coil really audible? Probably it would be hard to prove, but IF you can remove the coil, then why not? Can people hear a 2uH or larger coil. I think so, and probably as much because of what Graham is talking about on this website, as much as any ringing or any other interaction with the loudspeaker load.
Our advice is to get rid of the coil, if you have the design sophistication to do so. However, if you need a coil, then maybe 0.5uH with a 1 ohm or smaller resistor is probably OK. It is up to each designer.
I have not used an output coil in my designs for the last 15 years, but your results may vary.
Graham,
This back emf thing could be the answer to (some of) my prayers. I haven't had time to work this through yet but I do have a problem - this may not be a problem by the time I'm done but just to check I'm not missing something fundamental:
The speaker force, acceleration, sound pressure should all be in phase with drive current. Speaker velocity and back emf then lag drive current by 90 degrees. The current being fed back to the amplifier input is then this back emf modified by the system impedance = speaker impedance + output impedance. So the output coil is effectively adding to the inductance of the cables and speaker with no differentiation as to the source of the inductance. This assumes no crossover. This total inductance then adds additional phase lag to the back driven current. I can see how this could play havoc with discontinuous signals but the magnitiude of the contribution of the output coil is then only its fraction of the total inductance.
Is this correct or am I missing something?
I do have a separate question which I'll cover in a separate reply.
Tony
This back emf thing could be the answer to (some of) my prayers. I haven't had time to work this through yet but I do have a problem - this may not be a problem by the time I'm done but just to check I'm not missing something fundamental:
The speaker force, acceleration, sound pressure should all be in phase with drive current. Speaker velocity and back emf then lag drive current by 90 degrees. The current being fed back to the amplifier input is then this back emf modified by the system impedance = speaker impedance + output impedance. So the output coil is effectively adding to the inductance of the cables and speaker with no differentiation as to the source of the inductance. This assumes no crossover. This total inductance then adds additional phase lag to the back driven current. I can see how this could play havoc with discontinuous signals but the magnitiude of the contribution of the output coil is then only its fraction of the total inductance.
Is this correct or am I missing something?
I do have a separate question which I'll cover in a separate reply.
Tony
Graham,
Second question: I've noticed a couple of things about the way certain speakers sound. On various units I've measured and listened to there appears to be a strong correlation between measured IMD and the unit's sound clarity. I've also noticed that the IMD performance and audible clarity of some speakers can be improved by driving the units from a current source.
On other units there was no real improvement in either. The units that showed no real improvement tended to be the more expensive examples from the likes of Scan Speak. I had summized that some of the changes made to the design of these were to mitigate the fact that the speakers were being voltage driven when their basic operation was as a current controlled device. I did notice, however, that these units also had a field shunting feature that greatly reduced the effective inductance.
I'm now unsure of my conclusions and am wondering what part the reduced inductance might be playing in the performance of these units. When I finally get my measurement system back together I'll take some measurements. Meanwhile, would any of your experiences shed light on this?
Tony
Second question: I've noticed a couple of things about the way certain speakers sound. On various units I've measured and listened to there appears to be a strong correlation between measured IMD and the unit's sound clarity. I've also noticed that the IMD performance and audible clarity of some speakers can be improved by driving the units from a current source.
On other units there was no real improvement in either. The units that showed no real improvement tended to be the more expensive examples from the likes of Scan Speak. I had summized that some of the changes made to the design of these were to mitigate the fact that the speakers were being voltage driven when their basic operation was as a current controlled device. I did notice, however, that these units also had a field shunting feature that greatly reduced the effective inductance.
I'm now unsure of my conclusions and am wondering what part the reduced inductance might be playing in the performance of these units. When I finally get my measurement system back together I'll take some measurements. Meanwhile, would any of your experiences shed light on this?
Tony
john curl said:
Hi John.
So Halcro lacks design sophistication? Also, are you even going to try to rebut my contention to your "audible ringing" claim, or is this discussion now going to come to a close on "probable" speculation?
Cheers,
Glen
I share Glen's exasperation, though some of it actually generated by Glen's (non technical) argument.
I did the auditions first and then sought to find out WHY a choke causes fractional degradation.
Everyone here is starting with sine theory/maths, then stating theory says the problem cannot exist, and now there is talk about ringing at RF frequencies ???
Will those who have not done so please go and audition for switched in/out choke degradation FIRST.
You need a top flight source, transient/metallic/percussion/sibilant source material and a chokeless loudspeaker sited mono-block to hear the difference with a dynamic loudspeaker system having good mid and tweeter.
As regards filters for digital audio/DACs has it not taken 20+ years to make the 'perfect system' more acceptable ???
Hi Ingrast,
>> Your post here <http://www.diyaudio.com/forums/showthread.php?postid=1206545> shows Delay ch-R, Delay ch-LS and BACK EMF voltage probes, the plot has no label whatsoever, thereby my insistence on having the point clarified. <<
This is clear to me, so I fail to understand your enquiry. Only two group delay traces are shown.
Below I add the group delay trace for the same choke//resistor with your equivalent LS.
Hi Workhorse.
Indeed what you write is so correct. Though the choke effect is directly related.
I am sure many of us have heard reproduction changes between hope amps/spkrs sound with short cables and a domestic arrangement, then the same gear used on a stage with long cables extending to either side of it.
The change does not relate to HF only but it modifies tonalities at frequencies directly related to loudspeaker system current phase change.
Hi Tony,
Yes.
By using a transducing voltage-current amplifier, the amplifier does not generate or fail to correct accurately for voltage errors that are not related to drive current (signal input).
The fractional errors do not become added/subracted, but the reproduction difference can be significant.
However speakers designed for voltage drive do not sound right with current drive, and vice-versa. In full-rangers the effect of impedance rises with inductance and resonance become meaningless. That is the drive at resonant frequency is the same for a first cycle and all subsequent.
The electromotive force at bass resonance with voltage drive starts off normal but becomes reduced after the first half cycle, and can be reduced even more during the next cycle due to a tuned loudspeaker enclosure. With current drive the bass will become louder (boomy) at resonance as more undiminished current drive energy becomes loudspeaker system stored in the resonance.
Cheers ........ Graham.
I did the auditions first and then sought to find out WHY a choke causes fractional degradation.
Everyone here is starting with sine theory/maths, then stating theory says the problem cannot exist, and now there is talk about ringing at RF frequencies ???
Will those who have not done so please go and audition for switched in/out choke degradation FIRST.
You need a top flight source, transient/metallic/percussion/sibilant source material and a chokeless loudspeaker sited mono-block to hear the difference with a dynamic loudspeaker system having good mid and tweeter.
As regards filters for digital audio/DACs has it not taken 20+ years to make the 'perfect system' more acceptable ???
Hi Ingrast,
>> Your post here <http://www.diyaudio.com/forums/showthread.php?postid=1206545> shows Delay ch-R, Delay ch-LS and BACK EMF voltage probes, the plot has no label whatsoever, thereby my insistence on having the point clarified. <<
This is clear to me, so I fail to understand your enquiry. Only two group delay traces are shown.
Below I add the group delay trace for the same choke//resistor with your equivalent LS.
Hi Workhorse.
Indeed what you write is so correct. Though the choke effect is directly related.
I am sure many of us have heard reproduction changes between hope amps/spkrs sound with short cables and a domestic arrangement, then the same gear used on a stage with long cables extending to either side of it.
The change does not relate to HF only but it modifies tonalities at frequencies directly related to loudspeaker system current phase change.
Hi Tony,
Yes.
By using a transducing voltage-current amplifier, the amplifier does not generate or fail to correct accurately for voltage errors that are not related to drive current (signal input).
The fractional errors do not become added/subracted, but the reproduction difference can be significant.
However speakers designed for voltage drive do not sound right with current drive, and vice-versa. In full-rangers the effect of impedance rises with inductance and resonance become meaningless. That is the drive at resonant frequency is the same for a first cycle and all subsequent.
The electromotive force at bass resonance with voltage drive starts off normal but becomes reduced after the first half cycle, and can be reduced even more during the next cycle due to a tuned loudspeaker enclosure. With current drive the bass will become louder (boomy) at resonance as more undiminished current drive energy becomes loudspeaker system stored in the resonance.
Cheers ........ Graham.
My last response to Tony reveals why we ought not rely upon steady sine based measurements and time isolated graphical representations, because whether with current or voltage drive, the curves reveal nothing about the sonic variations we hear.
Hi Rodolfo.
I forgot the illustration, and add this from the post you reference which explains my failure to understand your request;-
>> The straight line response below is the voltage group delay at R due to a series output choke and steady sine excitation, which is of course what everyone measures and bases their understanding upon !!!
The widely variable voltage group delay is that at the (typical)loudspeaker terminal due to another series output choke (both driven by a perfect source), yet this includes a negative delay component caused by leading back-EMF generated by the loudspeaker. <<
Cheers ........ Graham.
Hi Rodolfo.
I forgot the illustration, and add this from the post you reference which explains my failure to understand your request;-
>> The straight line response below is the voltage group delay at R due to a series output choke and steady sine excitation, which is of course what everyone measures and bases their understanding upon !!!
The widely variable voltage group delay is that at the (typical)loudspeaker terminal due to another series output choke (both driven by a perfect source), yet this includes a negative delay component caused by leading back-EMF generated by the loudspeaker. <<
Cheers ........ Graham.
Attachments
What I still do not understand is how a coil (2uH or thereabouts) can have any effect on the sound when the lumped inductance of the entire speaker system including cabling is much greater than this. Someone already showed that the cross over inductors on the tweeter side might be as high as 180uH, to say nothing of the speaker inductance, cabling inductance etc.
Graham, you talk about a 'chokeless' speaker driven 'monobloc'' style. This is not the real world. Any amplifier needs to deal with a whole range of output loads without hiccupping - I doubt an amplifer would be a commerciall success if we told the purchaser to make sure he used his amp with a 'chokeless' speaker and drove it 'monobloc' style.
I suspect that its not output inductors per se that is causing the difference in sound people claim to hear, but probably the amp misbehaving. And, maybe it sounds better to the listeners ear - but that's a very personal thing because what you deem to sound good might be terrible to my ears
One other point. Who's to say that if you put another (different brand or model) speaker on the output of the amp that the difference in sound would not be greater than with/without output coil, or, that it would sound better or worse than the original speaker, again, with or without the coil. most people agree, speakers have the biggest impact on the overall sound of the system - not the amp, and unlikely a 2uH coil.
I think there are far too many unsubstantiated claims about the sound of this or that component, when in most cases there individual components actually play a very small part in affecting the sound of an amplifier. The 2uH coil is one such component in my view, and you only have to look at the Halcro (damn fine amp, most will concede) and one of JC's designs (bipolar output, no coil and also claimed by some 3rd party's to be 'damn fine') so why all the dogma about the output coil?
Let's approach this thing like engineers - by logically peeling the onion starting from a systems poinjt of view.
I've just put by
on, so fire away!
Graham, you talk about a 'chokeless' speaker driven 'monobloc'' style. This is not the real world. Any amplifier needs to deal with a whole range of output loads without hiccupping - I doubt an amplifer would be a commerciall success if we told the purchaser to make sure he used his amp with a 'chokeless' speaker and drove it 'monobloc' style.
I suspect that its not output inductors per se that is causing the difference in sound people claim to hear, but probably the amp misbehaving. And, maybe it sounds better to the listeners ear - but that's a very personal thing because what you deem to sound good might be terrible to my ears
One other point. Who's to say that if you put another (different brand or model) speaker on the output of the amp that the difference in sound would not be greater than with/without output coil, or, that it would sound better or worse than the original speaker, again, with or without the coil. most people agree, speakers have the biggest impact on the overall sound of the system - not the amp, and unlikely a 2uH coil.
I think there are far too many unsubstantiated claims about the sound of this or that component, when in most cases there individual components actually play a very small part in affecting the sound of an amplifier. The 2uH coil is one such component in my view, and you only have to look at the Halcro (damn fine amp, most will concede) and one of JC's designs (bipolar output, no coil and also claimed by some 3rd party's to be 'damn fine') so why all the dogma about the output coil?
Let's approach this thing like engineers - by logically peeling the onion starting from a systems poinjt of view.
I've just put by
on, so fire away!
Hi Bonsia,
You wrote >> What I still do not understand is how a coil (2uH or thereabouts) can have any effect on the sound when the lumped inductance of the entire speaker system including cabling is much greater than this. <<
Because the choke / LS cables are in series with both parallel and series tuned circuits. Series tuned circuits can develop very high momentary currents when crossover capacitors are charged.
From my experience I would judge 2uH as becoming borderline for audiblity though.
I am not worried about commerciality because there will always be those who want the best they can get.
NFB amplifier 'behaviour' is improved by a series choke and this is a major reason for them being included, though 'behaviour' at the output terminal becomes electrically degraded.
Changing the LS with a series output choke coupled amplier DOES change the sound. Hi-Fi reviewers have to be very careful with their reviews in order to make sure they do not overly disadvantage any loudspeaker or amplifier.
I cannot comment on the amplifiers you mention, but any amp using say a 2uH output choke is very likely to sound equally good as another ?reference? amplifier using a 2uH output choke in A/B testing !!??!!
You wrote >> What I still do not understand is how a coil (2uH or thereabouts) can have any effect on the sound when the lumped inductance of the entire speaker system including cabling is much greater than this. <<
Because the choke / LS cables are in series with both parallel and series tuned circuits. Series tuned circuits can develop very high momentary currents when crossover capacitors are charged.
From my experience I would judge 2uH as becoming borderline for audiblity though.
I am not worried about commerciality because there will always be those who want the best they can get.
NFB amplifier 'behaviour' is improved by a series choke and this is a major reason for them being included, though 'behaviour' at the output terminal becomes electrically degraded.
Changing the LS with a series output choke coupled amplier DOES change the sound. Hi-Fi reviewers have to be very careful with their reviews in order to make sure they do not overly disadvantage any loudspeaker or amplifier.
I cannot comment on the amplifiers you mention, but any amp using say a 2uH output choke is very likely to sound equally good as another ?reference? amplifier using a 2uH output choke in A/B testing !!??!!
G.Kleinschmidt said:
Now thats a strike...
So what are you doing in this thread??
If you have built several (different, I assume) amps and there is no audible difference, then it's either
1. You have built the ultimate amps, absolutely perfect.
2. Your listening abilities are not adequate
3. You are experiencing what many are accused of here: You hear what you want to hear, and as you are someone who seems to believe only in numbers, you'll not hear any difference, because all your designs are at 0.000....% distortion. Even if there is a difference.
In case 1, you should immediately go and sell these amps, they will sell like....
In case 2, what are you discussing here, with your listening abilities? You're not up to the task.
In case 3, you should show a little bit more respect towards the experienced designers here and stop winding them up.
In any case, I and many others are diyers here and would like to be able to continue to listen to JC, GM and all the others. Guys like you have already managed to drive away enough experienced people...
So, give it a rest.
Tino
Hi Tino,
OR;-
(4) Maybe Glen's amplifiers all have the same value of series output choke. I don't know.
Cheers ........... Graham.
OR;-
(4) Maybe Glen's amplifiers all have the same value of series output choke. I don't know.
Cheers ........... Graham.
zinsula said:
Now thats a strike...
So what are you doing in this thread??
If you have built several (different, I assume) amps and there is no audible difference, then it's either
1. You have built the ultimate amps, absolutely perfect.
2. Your listening abilities are not adequate
3. You are experiencing what many are accused of here: You hear what you want to hear, and as you are someone who seems to believe only in numbers, you'll not hear any difference, because all your designs are at 0.000....% distortion. Even if there is a difference.
In case 1, you should immediately go and sell these amps, they will sell like....
In case 2, what are you discussing here, with your listening abilities? You're not up to the task.
In case 3, you should show a little bit more respect towards the experienced designers here and stop winding them up.
In any case, I and many others are diyers here and would like to be able to continue to listen to JC, GM and all the others. Guys like you have already managed to drive away enough experienced people...
So, give it a rest.
Tino
Maybe we should put this in perspective. So far we have:
- JC who gives a lot of unsubstantiated claims, who changes his position everytime he is challenged, and in general refuses to put his money where his mouth is;
- GM who insists to report things as a result of a test signal that has no relevance to audio, his infinitely-fast first cycle impuls, and who refuses to repeat his simulations/measurements with an audio-band signal;
- GK who gives his views, backs them up with arguments and measurements, and is the only one reasoning consistently here.
Who did you say was chasing away whom?
Jan Didden
Jan, I design amps for a living. That is MY statement! I don't need output coils, so most who do, ESPECIALLY with values over 0.5uH used by Halcro and Bob Cordell, show flawed design skills compared to the contemporary designs of today.
If you have a problem with me, I hope to hear about it, but don't expect me to be polite to you in my response.
If you have a problem with me, I hope to hear about it, but don't expect me to be polite to you in my response.
Graham,
whether the output inductance is in the form of a 2uH coil on the output of the amp, or whether its as a result of the speaker cable, surely the behaviour of the amp output stage is the same? An inductance is an inductance, right?
My point is
1. the output inductance is small compared to the overall inductances connected to the amp output
2. It is likely that other quantifiable variables have a much greater effect on the 'sound' - e.g. speakers - change them and viola - totally different sound.
John,
we all respect you, but please, because we use output coils does not mean we design flawed amplifiers. On the contrary, we seek to design robust, well engineered products. That said, I hold nothing against a designer who does not use one. Prove all the objectivists wrong by showing us of your inductorless designs driving a 2uF load//8 Ohms 10KHz square wave (2-3V pk pk would be ok) -then we can really learn something.
😉
whether the output inductance is in the form of a 2uH coil on the output of the amp, or whether its as a result of the speaker cable, surely the behaviour of the amp output stage is the same? An inductance is an inductance, right?
My point is
1. the output inductance is small compared to the overall inductances connected to the amp output
2. It is likely that other quantifiable variables have a much greater effect on the 'sound' - e.g. speakers - change them and viola - totally different sound.
John,
we all respect you, but please, because we use output coils does not mean we design flawed amplifiers. On the contrary, we seek to design robust, well engineered products. That said, I hold nothing against a designer who does not use one. Prove all the objectivists wrong by showing us of your inductorless designs driving a 2uF load//8 Ohms 10KHz square wave (2-3V pk pk would be ok) -then we can really learn something.
😉
Jan,
Your armchair post is so unproductive. Why not do something pro-active, and stop making conceptual assumptions.
I have yet to see anyone other than Andy-C try any properly suitable technically based examination.
Do you not see Jan, that your own thoughts about 'infinitely fast first-cycle impulses' are creating a 'steady-state' internal barrier related to concept ?
How can you begin to consider the dynamically energised group delay disturbances caused by loudspeaker current flow through a series output choke, and how do inner loops within the closed global loop of a NFB amplifier affect both output impedance and damping, and thus reproduction ?
Cheers ......... Graham.
Your armchair post is so unproductive. Why not do something pro-active, and stop making conceptual assumptions.
I have yet to see anyone other than Andy-C try any properly suitable technically based examination.
Do you not see Jan, that your own thoughts about 'infinitely fast first-cycle impulses' are creating a 'steady-state' internal barrier related to concept ?
How can you begin to consider the dynamically energised group delay disturbances caused by loudspeaker current flow through a series output choke, and how do inner loops within the closed global loop of a NFB amplifier affect both output impedance and damping, and thus reproduction ?
Cheers ......... Graham.
Hi Bonsai,
Please tell me. What is the reproduction relevance of a 2uF//8R test load.
Name any loudspeaker with anywhere near those characteristics.
By wanting a choked amp that will drive 2uF all day long you are shooting your system before you listen to it !!!
We all know (I hope you don't infer otherwise) about the horrendous loudspeaker problems there are in relation to those of amplifiers, but you can still be distracted by the tiny amplifier/electrical disturbances even when listening via compromised loudspeaker designs.
Cheers ......... Graham.
Please tell me. What is the reproduction relevance of a 2uF//8R test load.
Name any loudspeaker with anywhere near those characteristics.
By wanting a choked amp that will drive 2uF all day long you are shooting your system before you listen to it !!!
We all know (I hope you don't infer otherwise) about the horrendous loudspeaker problems there are in relation to those of amplifiers, but you can still be distracted by the tiny amplifier/electrical disturbances even when listening via compromised loudspeaker designs.
Cheers ......... Graham.
Maybe I could add an armchair comment of my own here.
This is like an egotistical - - mine's bigger than your's thing(knowledge bank of course) - - and everyone is so satisfied with their own designs they don't feel any need to try other's.
That does not mean to say that other designs do not drive loudspeakers more accurately, and maybe even other designs which have not been considered here !!!
Cheers ......... Graham.
This is like an egotistical - - mine's bigger than your's thing(knowledge bank of course) - - and everyone is so satisfied with their own designs they don't feel any need to try other's.
That does not mean to say that other designs do not drive loudspeakers more accurately, and maybe even other designs which have not been considered here !!!
Cheers ......... Graham.
- Status
- Not open for further replies.