lumanauw said:
Then the patent is easy to work around. Very easy.
If you look at the schematic on the cover page of the patent, you'll see two current sources, 26 and 27, and the connecting resistor, 40. Current source 26 is connected to node 38, as is one end of resistor 40. Current source 27 is connected to node 39, as is the other end of resistor 40.
Now, split resistor 40 (I'll call it 'R_40') into two resistors in series, R_40a, and R_40b, such that R_40a = R_40b, and R_40a + R_40b = R_40. Obviously, that's made no real change to the circuit. There is a new node, though, where R_40a is connected to R_40b (the middle of R_40, in other words). I'll call that new node, 'N_40' (as it's in the middle of R_40).
Next, disconnect the two current sources (26 and 27) from nodes 38 and 39, and connect them to the new node, N_40 (the middle of R_40), instead. They're now two current sources connected together in parallel, equivalent to one current source which gives (or takes) the sum of the currents given (or taken) by the original current sources. So, lets actually replace those two current sources with the single, equivalent current source. Now we've got a single tail!
What difference has this made? None, really. The two circuits are equivalent. (It's just a simple matter of basic circuit theory, really nothing more than Kerchoff's laws and Ohm's law.) But it should now be clear that the overall circuit is (a trivial equivalent of) a simple variation of a long-tailed pair. (The tail was previously 'in disguise', in the form of two current sources, 26 and 27, instead of the single, equivalent current source forming the tail. Resistors 40, 42 and 43 just change the differential and common-mode open-loop gains.)
As for the folded cascodes - ? It's clear from the schematic that they're not really significant when it comes to how the circuit uses negative feedback to reduce distortion. The folded cascodes are not a significant difference.
So, either the patent is easy to circumvent (work around) (by combining the two current sources into a single tail), or it's invalid (because, due to the simple equivalence described above, it's not sufficiently different to what came before, and lacks non-obviousness).
Actually, I think I'm now pretty much convinced that I wasn't missing anything, after all.
As for the rest of your upper post, I really (no joke) cannot capture your point. I don't use English in my daily life here, so pardon me
No worries
The point's not important, now, anyway. Sorry I didn't express it more readably, though.But thank you for your contributions
I think your posts have been some of the most helpful to me in helping me to focus in on the issues in my own mind (if that makes sense).azira said:well... I'm sure this'll be far from exact but let me take a stab at this.
First, IMO a patent serves two purposes.
1) allows the owner exclusive rights to license or produce the particular technology for 10 years.
2) claims the owner as the inventor of the technology.
Now, obviously #2 is important to the petitioner but #1 is the important one.
The truth is, you're right, and there are plenty of patents out there that seem obvious or "duh". But lots of times it's only obvious after the fact that the obviousness comes out and the point really is: the patent was specific enough to get approved.
So if you look at the patent for X/SuSy, yeah it's negative feedback, but it only works for perfectly matched pairs, and the patent is specifically for cascoded LTPs with the feedback taken from the LTP, not after a VAS. And that's a pretty specific criteria. That's why it's patentable. If you were to apply X feedback to a Cascoded matched LTP with a say... a matched Source follower after it, well that'd be a different circuit and you could most likely market it (as long as your lawyer is better than NPs
--
Danny
Minor correction to point one:
Formerly (I can't remember the actual date) but US patents were for 17 years from the date of issue of the patent. Some years ago it was changed to 20 years from (if I remember correctly) the date of filing the patent application. This change aligned US patent policy with the rest of the world.
A comment or two on (non)obviousness (while I'm not an attorney, I have a few patents to my name). As a scientist, I have always thought it rather odd that an invention should be nonobviousness. It seems to fly in the face of why/how one conducts research (you know, scientific method, etc; an invention should be the logical, predicted outcome IMHO).
Sometimes it may seem obvious to try something (particularly after seeing the result) but, prior to doing the experiment or research, the outcome may not be so obvious.
My 2cents on the topic
SteveA
Simon G Best said:
Two points -
1. Pass has the patent. The LEGAL (Legal does not necessarily equal logical) presumption is that it is valid. This places the burden of proof on the challenger - you hop in your wayback machine and prove that at the time the patent was issued that it did not meet the criteria for patentability.
2. There is/was/may be (I have been away from patent issues too long to be certain) a "doctrine of equivalents" that says if the changes you make are immaterial to the function of the invention, you have simply made the same thing, and therefore are violating the patent.
As you point out, arguments can be made on both sides of the issue. But until someone like Harman decides they want to use that topology and battles it out in court we'll never really "know" if the patent is valid.
SteveA - you are correct, the US went to 20 years from date of application to bring us in line with the rest of the world. We don't always ram our way down the world's throat.
You are looking at Obviousness from a "How do I solve this problem" perspective. It may be have been a perfectly logical way to solve your problem, and seemed obvious to you. But, as used by the USPTO, "obvious" is a legal term with a definition different from common usage. It basically means that if something is obvious, virtually everyone will solve the same problem that way.
Say a new material is created that you see could be used to make your product slightly different and perform better. So you apply for a patent on your new product using that material. If the same week you apply the PTO gets 15 other applications that are virtually identical, it will ***probably*** be ruled to be obvious and not patentable. If just your big competitor's application arrives a month or so after yours, Steve is a Hero.
If they got there a week sooner than you, your lawyers will argue against the patent saying that Steve's just a regular guy of slightly below average intelligence - any monkey would have done that, this stuff was SO obvious... (isn't it amazing how fast a lawyer changes direction when it is to his advantage?) 
Simon G
"
Firstly, patents and copyright are different things. You're on about copyright, but I'm on about a patent.
Secondly, from what you've said about "the principle of x [having] been around well before Nelson came up with the idea", it seems the patent is, indeed, invalid. (For the patent to be valid, it's not enough for Pass' amplifiers to be his own work, they've also got to constitute a patentable invention under patent law.)
"
firstly
pattents copyrights same thing i say only this what is their purpose
secondly
Nelson has used them first in the power stage previously i assume these circuits had largely been used in small signal stages
it seems Nelson has pattended/copyrighted every known topology
in the power stage for class-a operation, he was the first to use
them as power application, does that make it his because
he used them as such first, i am little bit confused i am not trying
to disrespect his idea to use them as a power amp circuit like
zen, aleph etc which are all pretty basic with a touch of his own
invention of course
thirdly
Pattents/copyrights are there, it is a reality, the only way to challenge them is going to courts with money and power
fourthly
I think you're being abit of a clown really, bringing such a subject
to this forum and spoiling the spirit of the forum because Nelson
is one of the most generous contributors here, if you wanted to
copy use some of his topologies still under protection you could
contact him and he might grant you some kind of license/loyalties
arrangement
fifthly
you should admire the guy for having the guts to use such things
as zens for power amplification and turning what was previously
an amplifier into a heater
sixthly
i await Nelson to comment on this thread it would be interesting
what he has to say on this subject
cheers
"
Firstly, patents and copyright are different things. You're on about copyright, but I'm on about a patent.
Secondly, from what you've said about "the principle of x [having] been around well before Nelson came up with the idea", it seems the patent is, indeed, invalid. (For the patent to be valid, it's not enough for Pass' amplifiers to be his own work, they've also got to constitute a patentable invention under patent law.)
"
firstly
pattents copyrights same thing i say only this what is their purpose
secondly
Nelson has used them first in the power stage previously i assume these circuits had largely been used in small signal stages
it seems Nelson has pattended/copyrighted every known topology
in the power stage for class-a operation, he was the first to use
them as power application, does that make it his because
he used them as such first, i am little bit confused i am not trying
to disrespect his idea to use them as a power amp circuit like
zen, aleph etc which are all pretty basic with a touch of his own
invention of course
thirdly
Pattents/copyrights are there, it is a reality, the only way to challenge them is going to courts with money and power
fourthly
I think you're being abit of a clown really, bringing such a subject
to this forum and spoiling the spirit of the forum because Nelson
is one of the most generous contributors here, if you wanted to
copy use some of his topologies still under protection you could
contact him and he might grant you some kind of license/loyalties
arrangement
fifthly
you should admire the guy for having the guts to use such things
as zens for power amplification and turning what was previously
an amplifier into a heater
sixthly
i await Nelson to comment on this thread it would be interesting
what he has to say on this subject
cheers
mastertech said:
Not quite sure what you mean, here, but to clarify (somewhat simplified)
Patents protect the function and esence of an invention from copying. You cannot build a copy of a patented circuit without violating the patent. Mr. Pass is gracious enough to allow us copy his designs for non-commercial use. There are others in the industry that are not so willing to have their Patented product built, even for research use. Patents must be applied for and approved by a government agency, and a US patent provides generally no protection against a company in another country making and selling a copy of a US patented item unless there is also a grant of patent rights in that country.
Copyrights protect an EXPRESSION of an idea. In the US you acquire copy rights simply by expressing and publishing an idea. More protection is provided by registering your copyright, but you are not out in the cold, even if you don't put that "copyright me 2005" on the bottom of the page.
Grey was perfectly within his rights to redraw the Pass patented circuit and claim a copyright in HIS drawing, his expression of the idea. Had Mr. Pass published a schematic and Grey's was too similar it might have infringed on Pass' copyright, but not the patent until he built the circuit.
So you cannot make a virtual replica of a Star Wars movie by changing the names, but you can make a futuristic moviewhere a young man joins forces with a woman who turns out to be his sister to fight an evil person who happens to be their father, if the rest of the details don't match too closely. It seems to be a little easier to violate a copyrighted song, though. Again, it boils down to how hard you and the copyright owner are willing to fight and the financial resources you each can bring to bear.
The third form of intellectual property is trademark rights. Trademark rights are acquired by use and by registering with the PTO. The idea is that the "mark" identifies to the consumer the source of the goods. Like copyright, better protection is afforded the holder of a registered trademark than one who simply uses the mark in trade. So if Krell were to come out with a new topology, the called the XA200 (even if it wasn't a patent violation) it would violate the Pass trademark rights in XA as applies to audio amplifiers. I could probably get away with calling my new television projector an XA200, since it is a different market than Pass Audio gear.
Patents, Registered Trademarks and Registered Cpyrights are all voidable, but once issued, they are valid. The burden of proof is on the party claiming that the rights should not have been granted. No matter what you think of the invention described in a patent, the patent is not void until succesfully challenged.
I recall a story claiming that a cable company was issued a patent for a four wire braid, and the claims for the invention were almost identical to those in a patent issued 40 years earlier. Probably the examiner missed that one, but the patent is still valid, and if you want to sell a 4 wire braid for that same purpose, be prepared for an expensive battle.
Hope that this helps
Joe Berry said:
Ah! I see!
Yes, that's what resistors 42 and 43 on the patent cover page are for, yes? To reduce that common-mode distortion that can't be reduced differentially?Is that what the patent's for? I thought it was for the following:-
For a moment, I thought that was very similar to my understanding of it. But re-reading, I think I disagree.
My understanding of it is this:-
If, say, without applying negative feedback, there's a blip of distortion on one of the outputs, that blip has both differential and common-mode components. It's because it contributes both to the difference between the outputs (differential) and the mean of the outputs (common-mode). (A purely differential error would appear on both outputs, with no common-mode component.)
When negative feedback is applied (and ignoring the aforementioned resistors 42 and 43 for the moment), the differential component of the blip is reduced, but the common-mode component is left much as it was. Consequently, the common-mode component appears on both sides of the output (it's common-mode, after all), with the differential component reduced (and no longer hiding the common-mode component on the other side, as it were).
(This isn't the same as converting differential error into common-mode error, just a matter of reducing the differential component more than the common-mode component. (But perhaps I'm being pedantic, and we meant exactly the same thing.))
That, as I understood it, was what resistor 40 (in the schematic on the patent cover page), the application of negative feedback, and the patent itself were for.
(But then there are also the aforementioned resistors 42 and 43. These increase the magnitude of the (negative) common-mode gain, which increases negative feedback's reduction of common-mode distortion, at the expense of reducing the amplifier's common-mode rejection ratio (it's a trade-off). But, as you say, the outputs are used differentially, anyway.)
Reading the patent abstract (again), it seems clear that it's about the differential stuff, not the common-mode stuff. (I'm assuming the abstract is not misleading.) The first paragraph of the "DISCLOSURE OF THE INVENTION" is even more compelling:-
The last sentence of the third paragraph (in that same section), followed by the fourth paragraph, is particularly revealing (emphasis mine):-
That's the aforementioned resistors 42 and 43, then.
Long-tailed pairs, and the like. That's just what they do. Nothing new.
Clearly (answering my question above), the common-mode stuff, including resistors 42 and 43, is not what the patent's for. It's for the differential stuff - resistor 40 and negative feedback. But the addition of the cross-coupling resistor, resistor 40, just turns it into (a simple equivalent of) a simple variation of a long-tailed pair (just connect the current sources 26 and 27 to the middle of resistor 40, instead of to nodes 38 and 39 - it's an equivalent circuit).
So, yeah, I'm quite convinced, now, that I wasn't missing anything, and that the patent's either invalid (and shouldn't have been granted), or easily and defensibly circumventable - just give it a single tail!
Netlist said:
Transform it into what? The long-tailed pair is slap-bang in the middle of the schematic, from top to bottom. It's already clearly a long-tailed pair - no transformation necessary. Nothing between the tail (the top half) and Q5 and Q7, either. I suppose the tail could be split in two, one tail for each half of the pair (each tail giving half the current of the original), with the two tails connected in parallel (so that it's equivalent to the original tail). But it would seem rather pointless, as it's already a single tail as it is.
Perhaps I misunderstand?
Bob, the Supreme Court pretty much defanged the Doctrine of Equivalents a few years back.
If the inventor or examiner cited the earlier patent, then it's going to be expensive, yes. But if they just missed it, a re-examination can be done for a couple thousand bucks.
In theory, yes, but in reality, they'll probably go to 15 different examiners. I can give you an example of two patents that issued (and hence are presumed valid) despite covering the same invention and application dates only one day apart. One of them is mine, the other is my nastiest competitor. So much for the concept of interference...
If the inventor or examiner cited the earlier patent, then it's going to be expensive, yes. But if they just missed it, a re-examination can be done for a couple thousand bucks.
In theory, yes, but in reality, they'll probably go to 15 different examiners. I can give you an example of two patents that issued (and hence are presumed valid) despite covering the same invention and application dates only one day apart. One of them is mine, the other is my nastiest competitor. So much for the concept of interference...
I believe that there is a difference between Symmetry and SuperSymmetry. The difference is focused on the rate of distortion cancellation. If the SuperSymmetry apparently applies a new thing on the top of Symmetry, better cancels the distortion, and is well documented, why no patent?
This thread sounds valueless . . .
Regards
jH
This thread sounds valueless . . .
Regards
jH
Simon G Best said:
It's true that differential error will be reduced up to a point. What remains beyond that, as I understand it, is reflected in an error signal at the junction of the sources of the input differential pair. This residual error signal reflects not only common-mode distortions, but also differential distortion left uncorrected due to system asymmetry or finite loop gain.
Working now as common-gate amplifiers, both sides of the input diff pair amplify this error signal so that it appears in phase at both output terminals. The end result is a net reduction in distortion reduction, measured across the outputs, roughly equal to the square of the loop gain. For example, if the loop gain on either side is 20dB, you get roughly 40dB of distortion reduction - about 20dB due to shunt NFB, and another 20dB due to cancellation.
This behavior is inherent in differential shunt feedback amplifiers, but as far as I can tell, Pass was the first to recognize its utility and codify it as an operating principle for such amplifiers. Though perhaps not perfect, the approach has clear advantages and benefits over the "prior art," notably simplicity.
Joe Berry said:
No. It works the same way as long-tailed pairs - prior art.
In a long-tailed pair (with MOSFETS), with negative feedback applied on both sides, you get the same error signal (current) carried from one MOSFET source to the other as you do with Pass' topology. They both work the same way.
Replace the MOSFETs with BJTs? No relevant change. You've still got that error signal current stuff going on. Triodes? Again, no relevant change. Stick resistors between the MOSFET sources (or BJT emitters, or triode cathodes, or whatever,) and the tail? Still no relevant change. Yes, you change the open-loop, differential gain that way, but you still have the same error current stuff happening. Stick resistors in parallel with the tail, to change the common-mode gain (as well as changing the differential gain)? Again, no relevant change, because that error signal stuff still happens, still working the same way as before.
This stuff about error signals between the MOSFET sources really isn't at all surprising. As far as the long-tailed pair itself (not including externally-applied feedback) is concerned, the negatively-fed-back signal is just another part of the input signal. It gets processed in exactly the same way as the rest of the input. That involves those signal currents - including errors negatively fed back - between the two sides. This is also fundamentally relevant to how long-tailed pairs have different differential gains (large) to common-mode gains (small) (and hence high common-mode rejection ratios). (If you perturb one input, both outputs change.)
So, when negative feedback is applied, it, too, sends stuff across the coupling of the two halves at the top of the tail. (So, when the perturbation of one of the inputs is an error fed back from one of the outputs, it, too, goes across the top of the tail, appearing on the other output - exactly what long-tailed pairs do with their inputs anyway.)
There is no revelation, nothing new to appreciate, because that's what long-tailed pairs do anyway. That they still do that even when negative feedback is applied is quite unremarkable. It's exactly what would be expected anyway.
I think the problem in seeing this with Pass' patent is that he came at it from a different direction. He started off with two, uncoupled amplifiers, used together for the amplification of differential signals. Then, he joined them together at the waist (as it were), adding a resistor between the MOSFET sources. Superficially, it doesn't look like a long-tailed pair, because it doesn't have a single tail. Instead, the tail's split in two, in parallel, with the connecting resistor in between. But that's equivalent to a single tail, connected to the middle of the connecting resistor - a long-tailed pair!
As soon as you see it as the long-tailed pair (or simple variation thereof) that it equivalently is, and knowing and understanding how long-tailed pairs work, Pass' 'invention' turns out to be nothing more than a simple, insignificant change to what must undoubtedly be very well-known prior art.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.