The Beauty of Long-Tailed Pairs
Yes, it is truly beautiful. The simplicity of long-tailed pairs (and simple variations thereof) is, indeed, something that should be appreciated by every electronic engineer. Long-tailed pairs are one of those, 'A-ha!' things. When you first come to understand them, and the myriad possibilities they offer (op-amp applications and more), you can't help but marvel at the power of their elegant simplicity and symmetry.
Indeed, the ('cross') coupling of the two cathodes/emitters/sources/whatever is fundamental to it. If it doesn't have that coupling, it's not a long-tailed pair. I think this is what some here are missing. A long-tailed pair isn't just a pair of amps with long 'tails'. A long-tailed pair is a pair of amps coupled together, with a single tail. That's absolutely fundamental to them. It's what they are. And it's quite normal - indeed, very common! - for them to be used with feedback.
Christer said:
Yes, it is truly beautiful. The simplicity of long-tailed pairs (and simple variations thereof) is, indeed, something that should be appreciated by every electronic engineer. Long-tailed pairs are one of those, 'A-ha!' things. When you first come to understand them, and the myriad possibilities they offer (op-amp applications and more), you can't help but marvel at the power of their elegant simplicity and symmetry.
Indeed, the ('cross') coupling of the two cathodes/emitters/sources/whatever is fundamental to it. If it doesn't have that coupling, it's not a long-tailed pair. I think this is what some here are missing. A long-tailed pair isn't just a pair of amps with long 'tails'. A long-tailed pair is a pair of amps coupled together, with a single tail. That's absolutely fundamental to them. It's what they are. And it's quite normal - indeed, very common! - for them to be used with feedback.
Simon,
Interesting though your slow disection of this design is, there seems to be a non technical angle on this; that you cant tolerate patents. Patent law is much misused, and it is understandable that someone with stong feelings about this can become hot under the collar. However, you are going after the wrong man here. Without Nelson Pass, there might not have been this forum for you to argue your case. He is a major benefactor of diyAudio, and has supported us in many ways over the years. In no way can he be described as a ”greedy patent” grabber.
Let’s make an analogy:
If Nelson ran Microsoft, he would (for personal use):
Give the software away – free..
Include the source so you could adapt it – free.
Patiently give you one to one advice on using and modifying it – free.
Showcase your adaptations of his ideas – free.
Give you badges to make it look authentic – free.
Almost GPL
Oh, just one more thing:
Having spent time and money developing his designs, he wants to keep hold of the right of ownership, so you don’t go into commercial competition with him.
Unreasonable?
Interesting though your slow disection of this design is, there seems to be a non technical angle on this; that you cant tolerate patents. Patent law is much misused, and it is understandable that someone with stong feelings about this can become hot under the collar. However, you are going after the wrong man here. Without Nelson Pass, there might not have been this forum for you to argue your case. He is a major benefactor of diyAudio, and has supported us in many ways over the years. In no way can he be described as a ”greedy patent” grabber.
Let’s make an analogy:
If Nelson ran Microsoft, he would (for personal use):
Give the software away – free..
Include the source so you could adapt it – free.
Patiently give you one to one advice on using and modifying it – free.
Showcase your adaptations of his ideas – free.
Give you badges to make it look authentic – free.
Almost GPL
Oh, just one more thing:
Having spent time and money developing his designs, he wants to keep hold of the right of ownership, so you don’t go into commercial competition with him.
Unreasonable?
Nelson Pass said:
Okay, I think I had it backwards before and was confusing this circuit with its predecessors.
So it seems there are two ways to make error equal and cancel across the two sides of a differential shunt-feedback amp. One is to amplify common-mode error, which in turn requires common-mode gain. This is how the problem was approached historically, and there are several variations of this in the literature.
Another way is to selectively amplify the error so that the errors of one side are magnified *only* by the other and vice versa. As it turns out, this happens inherently as a result of the failure of each side to "see" and amplify its own error while passing it to the other side, where it is "seen" and amplified.
This failure of each side to amplify its own error is not due to a feature of the circuit architecture, but to the less-than-perfect nature of the input devices themselves. Is that closer to it?
Re: Help
My simulation would be a simulation of X. At least, that was what I intended. Perhaps I guessed wrong about the input impedance of the second stage? Perhaps it's a much lower impedance than I thought?
Nelson mentioned the impressive performance when used in a two-stage power amplifier. I understood the first stage to be essentially the same kind of thing as on the patent cover page, but with feedback taken from the outputs of the second stage (not included on that cover page schematic) instead of the first. My bet was that that second stage has quite high input impedance.
Nelson suggested I actually build it and try it out. Presumably, this is so that I'll be enlightened as a result (or at least puzzled when it doesn't work the way I thought, leading me one step closer to enlightenment).
If Nelson provides me with, or points me to, the relevant schematics, I'll happily run the simulations to put my predictions to the test, and report back here (Or I could produce my own schematic, of course, and provide that to check whether or not I've implemented it correctly.)
MikeW said:If your simulationis better than the X, send the schematic. I will get a layout and the get board made.
My simulation would be a simulation of X.
At least, that was what I intended. Perhaps I guessed wrong about the input impedance of the second stage? Perhaps it's a much lower impedance than I thought?Nelson mentioned the impressive performance when used in a two-stage power amplifier. I understood the first stage to be essentially the same kind of thing as on the patent cover page, but with feedback taken from the outputs of the second stage (not included on that cover page schematic) instead of the first. My bet was that that second stage has quite high input impedance.
Nelson suggested I actually build it and try it out. Presumably, this is so that I'll be enlightened as a result (or at least puzzled when it doesn't work the way I thought, leading me one step closer to enlightenment).
If Nelson provides me with, or points me to, the relevant schematics, I'll happily run the simulations to put my predictions to the test, and report back here
(Or I could produce my own schematic, of course, and provide that to check whether or not I've implemented it correctly.)
Re: Re: Help
After hurting the man's feelings, after questioning the man's reputation...you ask his help to prove your point? Nelson didn't bash whoever is Mr. LTP to come up with the SuSy.
I think it's best for your own interest if you come up with your own circuit and let Nelson dissect it. You're on your own man.
Simon G Best said:
If Nelson provides me with, or points me to, the relevant schematics, I'll happily run the simulations to put my predictions to the test, and report back here
After hurting the man's feelings, after questioning the man's reputation...you ask his help to prove your point? Nelson didn't bash whoever is Mr. LTP to come up with the SuSy.
I think it's best for your own interest if you come up with your own circuit and let Nelson dissect it. You're on your own man.
dhaen said:
Straw man. I have never described Nelson Pass "as a 'greedy patent' grabber".
Our various political positions on patents and patentability, and Pass' generosity in openly sharing his designs, etc, are irrelevant to the question of what it was that was new and non-obvious about Pass' claimed invention(s) at the time of the patent application. Such political issues, and the like, are red-herrings. That seems to be very much your point! So why raise a straw man? Why introduce such irrelevancies?
Pass' generosity does not mean the patent is valid. Neither do my political beliefs render the patent invalid. These are irrelevant issues (irrelevant to the question of what the patent's actually for), introduced here by others, but not by me.
Simon G Best said:
Well, I couldn't possibly know exactly how you have misunderstod me, could I?
As I said before, in my understanding, an LTP (or similar) is not necessary for implementing the SYSY concept (whether the patent requires that or not, is another issue). We have previously seen an implementation using two CFB op amps, as somebody has already mentioned. In my, possibly wrong, understanding, the major idea is that the amplifier amplifies a balanced signal and the feedback signals are fed to the other half of the amplifier instead of being fed back to the input of the same half, as is usually done. This can be described on the level of block schematics where no such details as LTPs or even transistors exist. Elegant? Yes, I think so. Was it a novel idea when Nelson patented it? Don't ask me.
Yes, I quite agree with you that LTPs are usually used with feedback, but not in this way. Usually we are amplifying an unbalanced signal, so one base is the signal input and the other is the feedback input, and the LTP works as a summing node. In Nelsons case, both bases (or gates) are signal inputs, and the feedback goes to the emitters (or sources), which is what some refer to as current feedback. I have never seent LTPs combined with this feedback technique to amplify balanced signals, but my experience is limited, so maybe there are plenty such cases that I am not aware of.
Now, if we cut off those feedback signals to the emitters and look at the input stage in isolation, then we have two transistors, two separate current sources and a bridge resistor between the emitters. My understanding is that we both agree that this circuit is functionally equivalent to an LTP (which is easy to prove). However, I get a feeling that som people here do not agree with this, but the discussion is to confused to quite understand if that is the case. Trying to clear this out was the main reason for my previous post, but in my understanding, that is not a central issue for SUSY.
Simon,
I think we are many who feel that the patent system doesn't work the way it should work, and there seem to be plenty of examples of granted patents around the world that shouldn't have been approved. However, the system is there, whether we like it or not. You seem to assume that Nelson thinks the system is flawless. Maybe he does, maybe he doesn't. He is a commercial player so it fully understandable that he uses the system whether he likes it or not. A company cannot afford to boycott the patent system just because it doesn't work as well as it was intended to. A company may even abuse the system and exploit it shortcomings. That is bad ethics, but such things usually don't bother many companys (I am not assuming Nelson is doing such things). Does anybody know what happened with that russian patent a few years ago that slipped through the system without the patent office realizing they had granted a patent on glass bottles?
I think we are many who feel that the patent system doesn't work the way it should work, and there seem to be plenty of examples of granted patents around the world that shouldn't have been approved. However, the system is there, whether we like it or not. You seem to assume that Nelson thinks the system is flawless. Maybe he does, maybe he doesn't. He is a commercial player so it fully understandable that he uses the system whether he likes it or not. A company cannot afford to boycott the patent system just because it doesn't work as well as it was intended to. A company may even abuse the system and exploit it shortcomings. That is bad ethics, but such things usually don't bother many companys (I am not assuming Nelson is doing such things). Does anybody know what happened with that russian patent a few years ago that slipped through the system without the patent office realizing they had granted a patent on glass bottles?
The mentioning of your anti patent stance is not a straw man.
You continue to insist that you sincerely are interested in this question that you continue to pose and that you are sincerely trying to understand. For this reason people are giving you the benefit of the doubt and sincerely spending time and effort discussing it with you.
If in fact you have a bias that keeps you from even considering that you might be wrong, then you are wasting people's time trying to make a small point, and people will know that your intent is to attack Nelson, not to understand anything.
Sincerely,
Variac
You continue to insist that you sincerely are interested in this question that you continue to pose and that you are sincerely trying to understand. For this reason people are giving you the benefit of the doubt and sincerely spending time and effort discussing it with you.
If in fact you have a bias that keeps you from even considering that you might be wrong, then you are wasting people's time trying to make a small point, and people will know that your intent is to attack Nelson, not to understand anything.
Sincerely,
Variac
I'm Giving Up With This
I'm giving up with this.
Those of you who will not open your minds to the possibility that Pass' 'invention' was nothing new can remain in blissful ignorance.
Those of you who want to learn about long-tailed pairs, feedback, and basic circuit theory, don't need my help. Google is our friend
And those of you who already understand such things? We already know the sound of one hand clapping
Cheery-Bye!
I'm giving up with this.
Those of you who will not open your minds to the possibility that Pass' 'invention' was nothing new can remain in blissful ignorance.
Those of you who want to learn about long-tailed pairs, feedback, and basic circuit theory, don't need my help. Google is our friend
And those of you who already understand such things? We already know the sound of one hand clapping
Cheery-Bye!
Re: I'm Giving Up With This
Simon,
sorry to see you leave after your 6 days on DIYaudio, the Pass forum and one visit to PassDIY.
Please post your future audio related patents here and on your DIY homepage, we'll be very gratefull if you share it.
For the moment i do not need Google that much, with Mr Nelson Pass around willing to answer everyone's questions.
Even without questions raised he offers homework every 3 to 6 months.
Simon G Best said:
Simon,
sorry to see you leave after your 6 days on DIYaudio, the Pass forum and one visit to PassDIY.
Please post your future audio related patents here and on your DIY homepage, we'll be very gratefull if you share it.
For the moment i do not need Google that much, with Mr Nelson Pass around willing to answer everyone's questions.
Even without questions raised he offers homework every 3 to 6 months.
Well Simon, I think you mostly have yourself to blame since you have constantly been very unclear about what you actually are discussing. You seem to be discussing several different issues at once, mixed up into an pretty incomprehensible soup. That's why I tried to separate some of the issues, hoping that the discussion could be somewhat more structured, but you seem not to want that. I am afraid the only thing you have achieved is to leave most of us in a state of grand confusion about what your point actually was.
It seems to me that his point was that the "X" patent was
either prior art or obvious, and therefore invalid.
As far as I know, the obvious argument is pretty much at the
discretion of the patent examiner. If after 70 odd years of
feedback applications, there is no prior art, then there's not
a lot of argument there.
It's important to keep in mind that the real patent is a series
of formal legal claims at the end of the patent - all the rest of
it is preamble for the non-lawyers in the house. The claims
are not usually the part that any of us actually parse.
Years after receiving the patent, I found out that Hadley created
an amplifier in the 60's which had symmetric feedback, but was
sufficiently different that it does not invalidate the claims of the
patent.
Almost 30 years ago I received a phone call from a wickedly
happy guy who found a typo in my 1976 patent and informed
me that the patent was invalid (and by the way do I call that
a chassis). I advised him to call the Patent Office up and clear it
up for them, and yes I do call that a chassis.
either prior art or obvious, and therefore invalid.
As far as I know, the obvious argument is pretty much at the
discretion of the patent examiner. If after 70 odd years of
feedback applications, there is no prior art, then there's not
a lot of argument there.
It's important to keep in mind that the real patent is a series
of formal legal claims at the end of the patent - all the rest of
it is preamble for the non-lawyers in the house. The claims
are not usually the part that any of us actually parse.
Years after receiving the patent, I found out that Hadley created
an amplifier in the 60's which had symmetric feedback, but was
sufficiently different that it does not invalidate the claims of the
patent.
Almost 30 years ago I received a phone call from a wickedly
happy guy who found a typo in my 1976 patent and informed
me that the patent was invalid (and by the way do I call that
a chassis). I advised him to call the Patent Office up and clear it
up for them, and yes I do call that a chassis.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.