"femto next gen." Gasp, all that speed to send videos of stupid cats, useless tweets, and porn.
I would wonder about the power dissipation of these devices if used in a linear circuit. I am guessing they are used in switching, off or saturated. We all know how unreliable GBIC's are in their normal use.
I would wonder about the power dissipation of these devices if used in a linear circuit. I am guessing they are used in switching, off or saturated. We all know how unreliable GBIC's are in their normal use.
I prefer to bear in mind that English is not the first language of several participants in this discussion
Now, what is the correct name for the kind of delay Thorsten was talking about? Is that "group delay", or is it something else again?
No offense was intended, just clarity.
John, once you invoke non-linearity the linear systems theory goes out the window.
We've been through this before. Degeneration is feedback, therefore the maths is the same. If you get a different outcome, then you started from a different place e.g. a forward open-loop gain characteristic which happens to give a good outcome.ThorstenL said:
Only a problem if the designer is not aware of how a feedback amp works. Input stage degeneration is the usual solution, by making the part before the integrator more linear it can handle the apparently massive overshoot. I say 'apparently massive', because if there is sufficient OL gain then it is still small but not as tiny as it would otherwise be.john curl said:
Hi,
Based on material cited previously degeneration and looped feedback behave differently, even if only applied to a single stage.
However, no such extra citation is needed...
IF degeneration and looped feedback where identical they would for example have the same effect on a given stage's output impedance. But they patently have not.
However, as you persist on making the, repeatedly debunked claim again and again I believe the burden of proof is upon you. So please provide such proof, or with respect, save me the time of having to debunk your claim again.
Ciao T
Based on material cited previously degeneration and looped feedback behave differently, even if only applied to a single stage.
However, no such extra citation is needed...
IF degeneration and looped feedback where identical they would for example have the same effect on a given stage's output impedance. But they patently have not.
However, as you persist on making the, repeatedly debunked claim again and again I believe the burden of proof is upon you. So please provide such proof, or with respect, save me the time of having to debunk your claim again.
Ciao T
If loop over several stages versus local degeneration in a single stage are derived and applied the same way, why wouldn't they have the same output impedance?
Sy,
Lets walk before we run and analyse a SINGLE stage.
Do you really suggest the degenerated device (BJT, FET, Pentode, Triode et al) will have the same output impedance from the resulting circuit for the application of degeneration and the application of a feedback loop from output to input?
Ciao T
Lets walk before we run and analyse a SINGLE stage.
Do you really suggest the degenerated device (BJT, FET, Pentode, Triode et al) will have the same output impedance from the resulting circuit for the application of degeneration and the application of a feedback loop from output to input?
Ciao T
Time for a new 'straw man'?
@qusp made an interesting statement in another thread.
That got me thinking because bridged is, IMO, superior to SE, while balanced is superior to bridged. This is a true with cans, at least. But why?
Let's try this. Balanced IS bridged but with common noise cancellation in the circuitry as well as at the can. Bridged provides common noise cancellation at the can. SE provides neither. Electrical measurements can demonstrate that balanced is better than se but, it takes the human ear to detect the improvement that differential signals make on the transducer.
The persistence of the topic of objective v. subjective is so old as to have become the holy war of audio. Lets all agree that leaving out conversion of electrical signals to compression waves limits applicability of the measurements of the former as a predictor of the subjective preferences of the receptors of the latter.
@qusp made an interesting statement in another thread.
That got me thinking because bridged is, IMO, superior to SE, while balanced is superior to bridged. This is a true with cans, at least. But why?
Let's try this. Balanced IS bridged but with common noise cancellation in the circuitry as well as at the can. Bridged provides common noise cancellation at the can. SE provides neither. Electrical measurements can demonstrate that balanced is better than se but, it takes the human ear to detect the improvement that differential signals make on the transducer.
The persistence of the topic of objective v. subjective is so old as to have become the holy war of audio. Lets all agree that leaving out conversion of electrical signals to compression waves limits applicability of the measurements of the former as a predictor of the subjective preferences of the receptors of the latter.
No, I didn't. I asked, "If loop over several stages versus local degeneration in a single stage are derived and applied the same way, why wouldn't they have the same output impedance?"
Just to make sure there's no language barrier, do you understand the meaning of feedback derivation and application? For example, what one would mean by "series derived, parallel applied"?
Of course, there is a typo under drawing 3. It should be read, "Parallel feedback", instead of "parallel input..."
As you may see, all feedbacks are around the single stage except the last one where feedback is both a degradation and a loop. Also, the same resistors represent load for the 2'nd transistor.
As you may see, all feedbacks are around the single stage except the last one where feedback is both a degradation and a loop. Also, the same resistors represent load for the 2'nd transistor.
Envision a battery powered scope or DVM, there is no way to differentiate between the two, a mc transducer can't know the difference. I would think the amplifiers grounding, supply wiring, etc. would cause more of a difference between se/bal.
Last edited:
Everyone, if you really want to get down to the fundamental equations, then look at this IEEE article from 1966: Daugherty, D. Greiner,R.
'Some Design Objectives for Audio Power Amplifiers. pp. 43-48
10.1109/TAU.1966.1161838
IEEE TRANSACTIONS ON AUDIO AND ELECTROACOUSTICS VOL 14 issue 1
It is the LAST page that is most important and the ONLY one that I am missing. You might look at THEIR conclusions, before making any of your own.
If anyone has this article and could send me the last page, I would be grateful.
'Some Design Objectives for Audio Power Amplifiers. pp. 43-48
10.1109/TAU.1966.1161838
IEEE TRANSACTIONS ON AUDIO AND ELECTROACOUSTICS VOL 14 issue 1
It is the LAST page that is most important and the ONLY one that I am missing. You might look at THEIR conclusions, before making any of your own.
If anyone has this article and could send me the last page, I would be grateful.
Sy,
You constructed a case that has nothing to do with the topic. Hence my suggestion to stick to simple circuits so we are clear what is being debated.
Yes, I understand, however, such feedback would not be degeneration, but looped feedback, if it is derived like looped feedback.
So, let's go back to basic electronics, given is one stage, with one active device, inverting.
We can apply degeneration to the device or looped feedback.
If the looped feedback (miller or full range) if used to produce the same gain reduction as degeneration will produce different results regarding harmonic spectrum, output impedance, input impedance and bandwidth, can we conclude safely that applying degeneration to a device is not the same as looped feedback?
After that we can consider if the same still holds if we use multiple stages with degeneration applied or with global looped feedback applied...
Ciao T
You constructed a case that has nothing to do with the topic. Hence my suggestion to stick to simple circuits so we are clear what is being debated.
Yes, I understand, however, such feedback would not be degeneration, but looped feedback, if it is derived like looped feedback.
So, let's go back to basic electronics, given is one stage, with one active device, inverting.
We can apply degeneration to the device or looped feedback.
If the looped feedback (miller or full range) if used to produce the same gain reduction as degeneration will produce different results regarding harmonic spectrum, output impedance, input impedance and bandwidth, can we conclude safely that applying degeneration to a device is not the same as looped feedback?
After that we can consider if the same still holds if we use multiple stages with degeneration applied or with global looped feedback applied...
Ciao T
Gladly when I am back at work.
See snip from Otala, why must we remain with our heads in the past? 50KHz follower bandwidths are long gone.
Attachments
Hi,
Very nice and colourful. I am quite familiar with the issues, but it seems that there are some here in dire need of such a diagram.
As things are so nicely colour coded, Yellow is degeneration, orange is (incomplete) looped feedback and purple is a circuit that not only uses more than one stage, but also combines degeneration and looped feedback, but is so far not yet on the agenda for debate...
I note that you have commented on the relative impacts on the impedances, but alas, not on harmonic distortion...
Ciao T
Thorsten;
let's start from basics:
here is the drawing I did for kids long time ago, when my English was even worse than now. I hope you are agree with this basics.
Very nice and colourful. I am quite familiar with the issues, but it seems that there are some here in dire need of such a diagram.
As things are so nicely colour coded, Yellow is degeneration, orange is (incomplete) looped feedback and purple is a circuit that not only uses more than one stage, but also combines degeneration and looped feedback, but is so far not yet on the agenda for debate...
I note that you have commented on the relative impacts on the impedances, but alas, not on harmonic distortion...
Ciao T
Well I must say that I was surprised when I won 'phonostage of the year' using 2 inexpensive IC's, (neither were ADI) to make it. However, I have also listened to my JC-3 phonostage, and compared it to my best Vendetta Research phono stage. I am still using the Vendetta research. AND I hope to make even better products in the future with both discrete and advanced IC designs (they are already in process).
Still, as there is no substitute for cubic inches in displacement in auto engines, there is no substitute for standing current, low noise jfets, and fully complementary circuits, for example. It is just that you can, most likely, keep most people happy without using every possible design technique, and that is what we do at Parasound, BUT we do not try to just 'cheapen' the product by using IC's when we can still use discretes.
We satisfy our customer base with pretty good products, much as Honda might.
However, some out there need, and are willing to pay for 'more' and they will get it through other sources, like Constellation Audio. You know, the folks who drive Porsches and Bentleys, etc. Trust me, a Bentley is more 'car' than a Honda. (at 10 times the price).
Still, as there is no substitute for cubic inches in displacement in auto engines, there is no substitute for standing current, low noise jfets, and fully complementary circuits, for example. It is just that you can, most likely, keep most people happy without using every possible design technique, and that is what we do at Parasound, BUT we do not try to just 'cheapen' the product by using IC's when we can still use discretes.
We satisfy our customer base with pretty good products, much as Honda might.
However, some out there need, and are willing to pay for 'more' and they will get it through other sources, like Constellation Audio. You know, the folks who drive Porsches and Bentleys, etc. Trust me, a Bentley is more 'car' than a Honda. (at 10 times the price).
??
funny thing that all those major electronics manufacturers have made polystyrene extended foil caps, resistors to "audiophile flavors", carbon vs metal film, wire wound, TaN, bulk metal foil...
I never knew the high end audiophile market drove so much of the tech I use in precision industrial/scientific instrument circuits - clearly no engineer ever asked for them ever for any other reason than their sound - according to your mythology
funny thing that all those major electronics manufacturers have made polystyrene extended foil caps, resistors to "audiophile flavors", carbon vs metal film, wire wound, TaN, bulk metal foil...
I never knew the high end audiophile market drove so much of the tech I use in precision industrial/scientific instrument circuits - clearly no engineer ever asked for them ever for any other reason than their sound - according to your mythology
Last edited:
- Status
- Not open for further replies.