Dollars poorer, but I have my bookzine!
Thankfully, I have a brother-in-law who works at the USPS. Many secret payoffs later, and after crossing the treacherous and mighty Brushy Creek from Austin, I have my bookzine! It is exactly as described and worth the effort. My wife asked me last night if I was paying attention to her while she was talking. No! hehehe
Although I did find a red herring in the envelope. jacco, was that you??
Thankfully, I have a brother-in-law who works at the USPS. Many secret payoffs later, and after crossing the treacherous and mighty Brushy Creek from Austin, I have my bookzine! It is exactly as described and worth the effort. My wife asked me last night if I was paying attention to her while she was talking. No! hehehe
Although I did find a red herring in the envelope. jacco, was that you??
jacco, I'm glad you have a sense of humor.
I feel like the small boy given his first chance to walk from home. I look back and still see my home, but looking forward, I've yet to see my destination. It's a long road, I know. But I give credit to three people on this site who have answered my questions with great patience and have kept me on the right path.
Could you ask your brother-in-law who I have to blow here in Sacramento to get mine?
se
Could you ask your brother-in-law who I have to blow here in Sacramento to get mine?
se
My guess is that the USPS will agree to those terms and have yours delivered immediately!
Seriously, my bro-in-law said that they see several interesting magazines that they didn't know existed. He said a few employees have been known to pull them out of the bundle and read them for a few days on their breaks, then deliver them once done.
Self's compensation article
In Self's article "Inclusive compensation..." he references Marshall Leach's article found here:
http://users.ece.gatech.edu/~mleach/papers/Feedforward.pdf
Quoting Self's article, he says "A similar input-inclusive compensation configuration was put forward by Marshall Leach [in the above article]"
Perhaps it is a matter of semantics, or maybe something I don't quite understand, but I never interpreted Leach's paper as input-inclusive compensation. Also, it does not seem to meet Self's definition of compensation.
Self shows compensation as an ADDITIONAL local loop to go along with global feedback. In Leach's case, the internal loop is an ALTERNATIVE global loop that is taken before the output stage. It is either/or on the global path, not compensating for the global path.
Any clarification on this?
In Self's article "Inclusive compensation..." he references Marshall Leach's article found here:
http://users.ece.gatech.edu/~mleach/papers/Feedforward.pdf
Quoting Self's article, he says "A similar input-inclusive compensation configuration was put forward by Marshall Leach [in the above article]"
Perhaps it is a matter of semantics, or maybe something I don't quite understand, but I never interpreted Leach's paper as input-inclusive compensation. Also, it does not seem to meet Self's definition of compensation.
Self shows compensation as an ADDITIONAL local loop to go along with global feedback. In Leach's case, the internal loop is an ALTERNATIVE global loop that is taken before the output stage. It is either/or on the global path, not compensating for the global path.
Any clarification on this?
I believe that Doug calls both cases 'input-inclusive compensation' because the compensation is not returned to the Vas input but to the input stage.
I see this also in Leach's paper (figs 1 & 2).
It also seems that what Leach calls feedforward is 'taking the (high freq) feedback signal from a point "forward" (before) the output stage'. I wouldn't call that feedforward.
Where's Edmond when you need him ?
jan didden
I see this also in Leach's paper (figs 1 & 2).
It also seems that what Leach calls feedforward is 'taking the (high freq) feedback signal from a point "forward" (before) the output stage'. I wouldn't call that feedforward.
Where's Edmond when you need him ?
jan didden
Last edited:
But I don't see any "compensation" as I understand it in Leach's approach.
The Leach Amp - Feedback Network
It does not "compensate" the overall global feedback path when the global feedback path is in operation. It replaces the "global" feedback path starting at the driver stage at high frequencies.
Again, semantics may play a part, here. I was never comfortable with Leach's term "feedforward compensation" for this, either. This reminds me of the transconductance doubling debate they had between them. When it came down to it, they both had a different definition of what class AB was.
The Leach Amp - Feedback Network
It does not "compensate" the overall global feedback path when the global feedback path is in operation. It replaces the "global" feedback path starting at the driver stage at high frequencies.
Again, semantics may play a part, here. I was never comfortable with Leach's term "feedforward compensation" for this, either. This reminds me of the transconductance doubling debate they had between them. When it came down to it, they both had a different definition of what class AB was.
Well, it compensates the global amp without the output stage. As the output stage is nominally unity gain, you could argue that it compensates the amp gain transfer function (because that's what it is about; it's not about compensating the feedback loop as such).
But I agree that his use of 'feedforward' is confusing if not plain wrong.
jan didden
The compensation in the Leach design is the two Miller dominant-pole capacitors C3 and C4, and this appears to be conventional. As Pooge rightly said, the Leach scheme replaces the outermost or global feedback path with another path fed from the VAS stage at high frequencies. The first diagram caption refers to the "feedforward feedback network" which is a touch opaque. I think it's a misuse of the word "feedforward" and what was meant was that feedback is taken from somewhere forward of the output.
It would appear the design was simulated but not measured. Making an amplifier is not just about Nyquist stability around the global feedback loop- that's relatively easy. What's harder is making sure that reactive loads don't trigger local parasitic oscillations in the output stage. SPICE does not, in my experience, offer enough detail in its transistor models to simulate this, and I can't help wondering how well the design would work in reality.
Output inductors are not that expensive in production. I always use them.
It would appear the design was simulated but not measured. Making an amplifier is not just about Nyquist stability around the global feedback loop- that's relatively easy. What's harder is making sure that reactive loads don't trigger local parasitic oscillations in the output stage. SPICE does not, in my experience, offer enough detail in its transistor models to simulate this, and I can't help wondering how well the design would work in reality.
Output inductors are not that expensive in production. I always use them.
Last edited:
The Leach amp incorporating this scheme has been around for over a dozen years. So yes, it has obviously been measured. There is also a SPICE for his circuit on this page:
Leach Amp Plans - Part 1
However, the alternate global was moved from the VAS to the predriver, for reasons mentioned in his article.
Leach Amp Plans - Part 1
However, the alternate global was moved from the VAS to the predriver, for reasons mentioned in his article.
Then it is not a matter of semantics. You just haven't read it.
And to think, you guys invented English.
I already posted this link.
The Leach Amp - Feedback Network
My complaint is that you steered the discussion off topic by dissing me in post #116 by stating my reference examples were somehow faulty because they were not the same amplifier (because the the feedback wasn't from the VAS and it had an output inductor). The original article, in the first paragraph of page 338, states:
"This positive feedback could be minimized by choosing the v, node to be a lower impedance point in the circuit, e.g. C2 could be connected to the emitter of Q7 (node 17) rather then to its base."
Therefore, Leach's amp clearly follows this tweek in the article, and therefore CLEARLY follows the article's teaching. I'm pretty sure MOST people would regard it as the same amp, at least for the benefit of the present discussion. The fact that he added an output inductor as an extra measure is totally irrelevant!
You have dissed the output triple for years as being characteristically unstable, yet the Leach amp is a very well known, rock stable design, even into electrostatic loads. Curiously, in your writings, you promote the advantages of a buffered VAS, which resembles the predriver of a triple.
As Leach's writings have been publically available for at least a decade on-line, and I know you are fully capable of understanding them, and many many amps based upon this design have been successly built, it makes me suspect that you have not read his available publications, which would surprise me, or that you have some other motive for minimizing the technology he has published.
"This positive feedback could be minimized by choosing the v, node to be a lower impedance point in the circuit, e.g. C2 could be connected to the emitter of Q7 (node 17) rather then to its base."
Therefore, Leach's amp clearly follows this tweek in the article, and therefore CLEARLY follows the article's teaching. I'm pretty sure MOST people would regard it as the same amp, at least for the benefit of the present discussion. The fact that he added an output inductor as an extra measure is totally irrelevant!
You have dissed the output triple for years as being characteristically unstable, yet the Leach amp is a very well known, rock stable design, even into electrostatic loads. Curiously, in your writings, you promote the advantages of a buffered VAS, which resembles the predriver of a triple.
As Leach's writings have been publically available for at least a decade on-line, and I know you are fully capable of understanding them, and many many amps based upon this design have been successly built, it makes me suspect that you have not read his available publications, which would surprise me, or that you have some other motive for minimizing the technology he has published.
Pooge,
nobody is "dissing" anybody... I looked at your links, and to me they don't appear to show what you are trying to discuss. The feedback comments relate to the circuit you posted the link too, and that doesn't have the feedback network you are talking about.
It's not up to the reader to go hunting through pages of stuff (where ever they may be)... and I haven't a clue what and where "page 338" is.
Don't assume knowledge or that everyone knows what you are talking about
nobody is "dissing" anybody... I looked at your links, and to me they don't appear to show what you are trying to discuss. The feedback comments relate to the circuit you posted the link too, and that doesn't have the feedback network you are talking about.
It's not up to the reader to go hunting through pages of stuff (where ever they may be)... and I haven't a clue what and where "page 338" is.
Don't assume knowledge or that everyone knows what you are talking about