Compensation testing method for amplifiers

[PAA]

JJ, what can you say about Cyrill Hammer's statement regarding the operation of amplifiers with negative feedback?

 

[SIY]

JJ, what can you say about Cyrill Hammer's statement regarding the operation of amplifiers with negative feedback?

Worthy of The Absolute Sound.

Anyone with a basic grasp of control theory knows why this is marketing blather.

 

[PAA]

Anyone with a basic grasp of control theory knows why this is marketing blather.

David Hafler, using the XL-280 amplifier as an example, showed that this is not marketing talk, that it is possible to actually make an amplifier with zero signal delay time in almost the entire audio range, and even with negative group delay without losing stable operation.

 

[Cbdb2]

David Hafler, using the XL-280 amplifier as an example, showed that this is not marketing talk, that it is possible to actually make an amplifier with zero signal delay time in almost the entire audio range, and even with negative group delay without losing stable operation.

So his Amp is faster than light? How can you have ZERO delay time?
And that Amp has global feedback which I believe is what Sys comment was partly about.
That amp also has almost no phase margin and would probably ring if it dosnt go completely unstable. Especially after you make it real, not a Sim.

 

[PAA]

So his Amp is faster than light? How can you have ZERO delay time?
And that Amp has global feedback which I believe is what Sys comment was partly about.
That amp also has almost no phase margin and would probably ring if it dosnt go completely unstable. Especially after you make it real, not a Sim.

Your speculations are not supported by anything. This amplifier was mass-produced. Here is its loop gain with the same tuning modes of the trimmer capacitor.
As you can see, its phase margin is stable regardless of the Gdelay setting.

 

[PAA]

JJ, are you familiar with the book "Operational Amplifiers" by Jiri Dostal_1993
If yes, please express the speed distortion of the linearly increasing voltage using FFT.

 

[solderdude]

That requires infinite bandwidth (the start of the rising input signal).
Something real world amps do not have so all amps will do this in some severity.
In an FFT (and FR) you will thus see the higher harmonics slope downwards and square-waves being rounded off.

And again.... this thread is about things that cannot be measured. Your example is not something that cannot be measured and a mystery.

 

[SIY]

If yes, please express the speed distortion of the linearly increasing voltage using FFT.

First, define "speed distortion" precisely. If you mean that as "delay," it absolutely shows in a Fourier transform.

For about the 700th time, you seem to not understand Fourier analysis, and you'll spend at least another 20 years going in circles and annoying people until you get that very basic concept.

 

[Jimbob54]

JJ, are you familiar with the book "Operational Amplifiers" by Jiri Dostal_1993
If yes, please express the speed distortion of the linearly increasing voltage using FFT.

This is beginning to feel like you just want to test the understanding and try to get one up on some of our senior members. I doubt that will happen but I'm not sure behaviour like this is beneficial to readers of this thread or the membership generally. Politely asking you to change tack here.

Also. Please. @ the person you are responding to and ideally use the reply function to the specific post you are responding to.

That second point goes to all members. We have at least 2 separate discussions going on here and it's obviously causing some confusion in what is already a dumpster fire of a thread.

Thanks all.

 

[PAA]

First, define "speed distortion" precisely. If you mean that as "delay," it absolutely shows in a Fourier transform.

For about the 700th time, you seem to not understand Fourier analysis, and you'll spend at least another 20 years going in circles and annoying people until you get that very basic concept.

SIY, open the book and figure it out. Rate error or Speed distortion and delay are different concepts.

The book is more than 30 years old, but during this time you still haven’t understood the difference.
Once you figure it out, then show a specific example of how Rate error is reflected in the Fourier transform.

 

[PAA]

This is beginning to feel like you just want to test the understanding and try to get one up on some of our senior members. I doubt that will happen but I'm not sure behaviour like this is beneficial to readers of this thread or the membership generally. Politely asking you to change tack here.

I just suggested getting acquainted with the material on the compensation testing method, which allows measuring those distortions ("Things that cannot be measured") that are not measured by standard tests. Instead of getting acquainted with the material, I am told again and again something like this: "this cannot be because it cannot be" - and this is once again the main argument.
The compensation method was studied by Otala, Baxandall and many other authoritative developers who were disappointed in the standard methods of measuring distortions.

 

[Jimbob54]

I just suggested getting acquainted with the material on the compensation testing method, which allows measuring those distortions ("Things that cannot be measured") that are not measured by standard tests. Instead of getting acquainted with the material, I am told again and again something like this: "this cannot be because it cannot be" - and this is once again the main argument.
The compensation method was studied by Otala, Baxandall and many other authoritative developers who were disappointed in the standard methods of measuring distortions.

Im going to suggest you need your own thread for that topic. Want me to move your posts and direct replies to a new thread titled "Compensation testing method for amplifiers" or similar? You are in a catch all thread here and the discussion is going to get disjointed as you're discussing something very specific. I may yet do that anyway but would prefer your consent.

 

[SIY]

SIY, open the book and figure it out.

It's your term, define it. Then I can try to teach you how to use a Fourier transform to determine it.

For others, always remember that frequency domain and time domain are completely equivalent and can be freely interconverted. Each contains exactly the same information.

 

[PAA]

It's your term, define it. Then I can try to teach you how to use a Fourier transform to determine it.

The definition of velocity error is given in the book, and Figure7-16 shows the physical meaning.

7.4 Rate Error

The rate error is familiar to those involved with servomechanisms: Asa result of dynamic limitations, the output shaft follows the control voltage with some delay, with an error proportional to the rate.

 

[SIY]

Asa result of dynamic limitations, the output shaft follows the control voltage with some delay, with an error proportional to the rate.

OK, then this will show up in the Fourier transform as a varying offset to the phase.

 

[NTK]

The definition of velocity error is given in the book, and Figure7-16 shows the physical meaning.

7.4 Rate Error

The rate error is familiar to those involved with servomechanisms: Asa result of dynamic limitations, the output shaft follows the control voltage with some delay, with an error proportional to the rate.

What the figure in the book shows is just the response of a bandwidth limited system to an infinite bandwidth signal. There is no real life infinite bandwidth system nor signal. When the bandwidth is enough for the application, it is enough.

Otherwise, can you explain how anti-missile interceptors (such as Iron Dome or Arrow) can possibly work?

 

[RayDunzl]

Nope, probably no infinite frequency:

Looks pretty high, though.

Wouldn't need a very big antenna to catch it.

 

[Cbdb2]

Your speculations are not supported by anything. This amplifier was mass-produced. Here is its loop gain with the same tuning modes of the trimmer capacitor.
As you can see, its phase margin is stable regardless of the Gdelay setting.

As I can see you changed the plots. WTF? The first plot you showed totally support what I said, almost no phase margin. This new plot is better but nit great. If you want people to believe you stop lying so obviously.

 

[NTK]

Your speculations are not supported by anything. This amplifier was mass-produced. Here is its loop gain with the same tuning modes of the trimmer capacitor.
As you can see, its phase margin is stable regardless of the Gdelay setting.

What you showed is just the response of a typical lead-compensated system.

 

[Holmz]

….
For others, always remember that frequency domain and time domain are completely equivalent and can be freely interconverted. Each contains exactly the same information.

While ^this^ is true, sometimes analysis is better done in the time domain.