I also did another 'proof of concept' circuit topology as a MC pre-pre application. Also, raw and basic. Because I was working at LLNL, I tried to get it patented thru LLNL lawyers. They said it was patentable However, if it didnt pertain to something I was working on, they could not spend the time and government money to get patent. I was released and free to patent it myself. I decided to publish it instead.
Another such idea/creation was the dc servo which i never applied for a patent. And, back a few pages here we were talking about how to reduce distortion of dynamic drivers and several new ideas.. patentable were produced --- Dude - Not every creative thought gets patented.
Several other patents of mine were assigned to various companies. some are for capacitor design (MultiCap), some are filters used on ac power line. Some are for ac power line protection schemes. Etc etc. I myself never paid for getting any of the patents. I dont care about patents -- companies did though. I do not know how many patents I produced in life time. I had forgotten about many of them. I just came across the patent doc for ac power line protection (tri-mode) a couple days ago. I had completely forgotten about it. I think it has a few more years left on it.
If you can find the exact same topology and current-mode operation (not VFA) described prior to myself or to Comlinear's we would all like to know. Ditto the exact same DC servo, prior to WJung's implementation of it in TAA preamp art. You will also find no foot notes as to its origin by D.Self or R.Cordell. Whether is a contribution to wavelet theory or a unique circuit by an IBM researcher... not all is written down as to who the originator is.
There is an African saying. 'When a man dies, a library burns down'.
THx-RNMarsh
Another such idea/creation was the dc servo which i never applied for a patent. And, back a few pages here we were talking about how to reduce distortion of dynamic drivers and several new ideas.. patentable were produced --- Dude - Not every creative thought gets patented.
Several other patents of mine were assigned to various companies. some are for capacitor design (MultiCap), some are filters used on ac power line. Some are for ac power line protection schemes. Etc etc. I myself never paid for getting any of the patents. I dont care about patents -- companies did though. I do not know how many patents I produced in life time. I had forgotten about many of them. I just came across the patent doc for ac power line protection (tri-mode) a couple days ago. I had completely forgotten about it. I think it has a few more years left on it.
If you can find the exact same topology and current-mode operation (not VFA) described prior to myself or to Comlinear's we would all like to know. Ditto the exact same DC servo, prior to WJung's implementation of it in TAA preamp art. You will also find no foot notes as to its origin by D.Self or R.Cordell. Whether is a contribution to wavelet theory or a unique circuit by an IBM researcher... not all is written down as to who the originator is.
There is an African saying. 'When a man dies, a library burns down'.
THx-RNMarsh
Last edited:
I don't know how is in the USA but It's God damn expensive to patent something today in the EU !There's a whole topic about it.I found out that you can patent anything for free in India and it is recognized world wide, you probably need to go there in person, i think...
I'm sorry to say that, but here we have just another case of religious electrons.
Well...I spent a few minutes reading back about three pages of comments and what was interesting was that in fact this is just another case of faithful disciple that is revering its rightful master up to a point where he becomes irrational attacking all his opponents as the most dangerous people in the universe.
I'm sorry to say that, but here we have just another case of religious electrons.
Same here... which is why i assign it to a company to pay for it... or publish. Anyway...... this is supposed to be about JC's original topologies. I only bring up my topology because JC stated simpler but not too simple was best.
JC... was there ever a patent on any or all of your original designs or topologies? How about the one you last put up to discuss?
-RNM
Last edited:
I don't have the patience for that.I didn't put a blame on anybody .I did the same with many people over the time...We're human nature...
By the way, i saw the comments on headphones chip amps ... testing lots of headphones amplifier topologies , for my 250 ohms DT880 , akg (600 ohms professional) tpa6120(ths6012) is perfect...I built the o2 and other chip based versions(class a biased bd139/bd140 driven by drv134 , studer a827 , paralleled njm4556 inverted and non inverted, transformerless and otp valve amps, mosfet and bipolar , etc...)but nothing sound better than tpa6120(ths6012).I only have a Kenwood amplifier(A63) , 2x30w which sounds as good as tpa6120, maybe a little bit closer to the truth than tpa6120...I just saw some comments about chips and headphones.I tried everything about it and i kinda know this territory a little bit better than others.
By the way, i saw the comments on headphones chip amps ... testing lots of headphones amplifier topologies , for my 250 ohms DT880 , akg (600 ohms professional) tpa6120(ths6012) is perfect...I built the o2 and other chip based versions(class a biased bd139/bd140 driven by drv134 , studer a827 , paralleled njm4556 inverted and non inverted, transformerless and otp valve amps, mosfet and bipolar , etc...)but nothing sound better than tpa6120(ths6012).I only have a Kenwood amplifier(A63) , 2x30w which sounds as good as tpa6120, maybe a little bit closer to the truth than tpa6120...I just saw some comments about chips and headphones.I tried everything about it and i kinda know this territory a little bit better than others.
Last edited:
Absolutely. I will never understand this VFA/CFA war. It is surrealistic.
Why not a war against cascodes, diamonds, I don't know what.
As it turned up in another thread these attribution discussions are not productive. The audio community has a long history of not paying attention to what is going on in the rest of the engineering world. Certainly the so called "current mode" operation of amplifiers to enhance BW and distortion performance has been around since tubes and WW2.
Another complicating factor is the patent office's doctrine of obviousness which is very hard to apply. Tubes certainly had no complements and early semiconductor processes had poor ones. So is taking an existing circuit and simply making a complementary version not obvious, who knows.
Another complicating factor is the patent office's doctrine of obviousness which is very hard to apply. Tubes certainly had no complements and early semiconductor processes had poor ones. So is taking an existing circuit and simply making a complementary version not obvious, who knows.
Like a lot of things... it can go undone for years or decades. It becomes obvious once it is done, though. Then people slap their head and say Why didnt I think of that! Its so obvious.
-Richard
Current feedback has been with us since Black at Bell Telephone. It was the most natural way to add loop feedback to a tube amplifier. In fact, it was taken over in the audio solid state amplifiers and preamplifiers until about 1970 by most people. This was because differential input was more expensive (added active part) and everything was cap coupled anyway. I worked on both current feedback bipolar (1967) then voltage feedback, (complementary differential) from 1968-1973. The biggest advantage of comp. differential was DC stability. The complementary jfet was tried in 1972-73 for power amps, and in theory, was one of the best ways to make a power amp, primarily because of 'stage count'. Still, the problem of DC stability. That is why, I switched to complementary differential jfets in 1973, for the GD modules, because they were more DC stable and they were slightly more versatile. I stick with it, because it is very linear, and very DC stable, BUT the single sided complementary input is potentially better, just because it uses less active parts in its audio path.
Patents are a real problem. While they give a claim to originality, they are usually owned by the big company one is usually working for, and in my case, that was Ampex.
I offered the circuit to them on two occasions and they thought it was too complex in both cases, so I took it with me when I left in 1969.
The Levinson JC-1 patent was another of my ideas that I developed before I worked with Mark Levinson, but he wanted to patent it, and of course he wanted to own it, but I would not assign it to him, just license it to him, so he paid for the patent, but never gave me the paperwork. It wasn't really worthwhile to patent a circuit that you don't have the money to defend. Ortofon just stole the circuit, and claimed it five years later.
Patents are a real problem. While they give a claim to originality, they are usually owned by the big company one is usually working for, and in my case, that was Ampex.
I offered the circuit to them on two occasions and they thought it was too complex in both cases, so I took it with me when I left in 1969.
The Levinson JC-1 patent was another of my ideas that I developed before I worked with Mark Levinson, but he wanted to patent it, and of course he wanted to own it, but I would not assign it to him, just license it to him, so he paid for the patent, but never gave me the paperwork. It wasn't really worthwhile to patent a circuit that you don't have the money to defend. Ortofon just stole the circuit, and claimed it five years later.
Patents mostly benefit lawyers, patent trolls, and corporations that wield them as weapons. You can get just about any garbage patented, especially regarding software. The system is broken. See Eastern district of Texas and Delaware.
I wonder why Samsung built an outdoor ice rink, for free, in Marshall, TX?
I wonder why Samsung built an outdoor ice rink, for free, in Marshall, TX?
Last edited:
Here lays a semantics problem and why IEEE does not call that kind of circuit Current-Mode amplifier. But call it CFB.... it is not the Current-Mode of operation being used today for high speed, high BW designs. Reread Bonsai for the distinctive characteristics. I know Scott gets it too and many others. It can be your term for CFB (feedback to a relatively low Z port, only) and not be operating as a CMA.
Never-the-less, as there are many VFB circuits and topologies..... so there are with CMA. Mine is one and Comlinear another etal. Most of the new designs are all MOSFET circuits which are optimised for RF/uwave apps.
I did not know my topology would have the characteristics of what is called a CMA until I measured it. My requirments for design just coincided with CMode operation. I wanted modest gnfb, fast, wide BW circuit which was dc coupled. The main reason no one did this before IMO is because the popular 2 transistor circuit (also fb to low Z port) with cap coupled in and out was considered good enough for audio.
I originally wanted better caps. And worked to figure them out and WJung and I wrote an article about caps together. But I decided after eliminating them in preamps using IC opamps and thought the sound was more accurate without any coupling caps. I figured how to make this direct-coupled as I showed. That lead also to dc servo for high gain (higher dc offset drift) circuits like RIAA.
BTW -- IIRC, after my 1980 preamp article came out, I also got a call from JC. My first of many. he too wanted to know why I chaged topology for the line stage.
THx-RNMarsh
Last edited:
Although Tubecad claims that everything published on his site is his property, i found some circuits of it published by Tektronix and others in the 50's, not to mention RDH4...
And indeed , the same valve based topologies were re-patented when silicon entered the market, just because of the principle of claimed component operation which was slightly different between valves and bipolar transistors(exponential vs square law) but similar in nature, while j-fet transistors simply copied all the tube circuits developed for 50 years before they appeared on the market.
Because the bipolar transistor was clearly defined by current gain while valves by transconductance and they also had different equations , they couldn't assign them the same operation modes , but in some instances they perform the same.
It's like with the sub-treshold region of j-fets where they perform the same as bipolar transistors govern by exponential law which made the manufacturers to make use of them in differential input or cascodes mixed with bipolar amplifiers giving a more predictable behavior of the whole circuit when feedback would enter the scene .The digital world evolved separately as they needed very different valves initially and then later different mos-fets than the others so they developed their circuits in a completely different world , but actually the most interesting patents of cascode and current operation were developed in the digital world.
And indeed, the radio world aren't using the exact same terms as the audio world does.
There's one more interesting aspect:
The military can't go public about their own patents and research telling that We did it first!
Here you have a simple example of a circuit that was patented for both , valves and mos-fet transistors, for both input and output stages:
US4647872A - Cascode amplifier
- Google Patents
And indeed , the same valve based topologies were re-patented when silicon entered the market, just because of the principle of claimed component operation which was slightly different between valves and bipolar transistors(exponential vs square law) but similar in nature, while j-fet transistors simply copied all the tube circuits developed for 50 years before they appeared on the market.
Because the bipolar transistor was clearly defined by current gain while valves by transconductance and they also had different equations , they couldn't assign them the same operation modes , but in some instances they perform the same.
It's like with the sub-treshold region of j-fets where they perform the same as bipolar transistors govern by exponential law which made the manufacturers to make use of them in differential input or cascodes mixed with bipolar amplifiers giving a more predictable behavior of the whole circuit when feedback would enter the scene .The digital world evolved separately as they needed very different valves initially and then later different mos-fets than the others so they developed their circuits in a completely different world , but actually the most interesting patents of cascode and current operation were developed in the digital world.
And indeed, the radio world aren't using the exact same terms as the audio world does.
There's one more interesting aspect:
The military can't go public about their own patents and research telling that We did it first!
Here you have a simple example of a circuit that was patented for both , valves and mos-fet transistors, for both input and output stages:
US4647872A - Cascode amplifier
- Google Patents
Can somebody translate in plain English what is the audio community understanding of a “current mode amplifier”?
The only definition that makes sense was provided by Gilbert:
http://cas.ee.ic.ac.uk/people/dario/files/E416/gilbert-voltagemode-currentmode.pdf
with the caveat of “pure voltage mode or current mode do not exist”.
It appears to me that “current mode” is used by some high end audio priests as yet another buzzword, in fact blurring Gilbert’s definition (in the context of translinear principle) to the point of completely obfuscating any engineering meaning.
The only definition that makes sense was provided by Gilbert:
http://cas.ee.ic.ac.uk/people/dario/files/E416/gilbert-voltagemode-currentmode.pdf
with the caveat of “pure voltage mode or current mode do not exist”.
It appears to me that “current mode” is used by some high end audio priests as yet another buzzword, in fact blurring Gilbert’s definition (in the context of translinear principle) to the point of completely obfuscating any engineering meaning.
Last edited:
@Syn08 no chance. Even the dedicated thread on CM vs VM just descended into circular arguments with a few sin bins without the two sides reaching a common understanding
Wasn’t that thread about CFA vs. VFA? These two concepts are rather well defined by the circuit properties. However, Marsh and other self appointed high end audio gurus are insisting in using a “current mode amplifier” concept, well outside Gilbert’s definition, and without any clear alternative definition. Hence my question.
Or am I jet lagged after crossing the ocean?
Last edited:
I call it a 'current mode amplifier' on my website because . . . . if I say 'current feedback amplifier ' I get emails telling me 'feedback is bad', 'feedback collapses the sound stage' etc etc. For me they are the same thing, the IEEE definition notwithstanding but IIRC that's more to do with high speed IC design concepts and nothing to do with audio.
Even Dennis Feucht (highly regarded analog designer) and Prof. Sergio Franco were not aligned on the EDN discussion on this subject prompted by one of the articles
As to whether its a CMA, CFA or VFA who cares and why dwell on it? As long as you get it and can use the topology effectively, that's all that matters.
(I also got an email from ex-member mikeks berating me for using CFA/CMA topology since it was 'entirely inadequate for audio' and there was 'no such thing as CFA'
Oh dear. Anyway, I am OUT of this discussion.
Even Dennis Feucht (highly regarded analog designer) and Prof. Sergio Franco were not aligned on the EDN discussion on this subject prompted by one of the articles
As to whether its a CMA, CFA or VFA who cares and why dwell on it? As long as you get it and can use the topology effectively, that's all that matters.
(I also got an email from ex-member mikeks berating me for using CFA/CMA topology since it was 'entirely inadequate for audio' and there was 'no such thing as CFA'
Oh dear. Anyway, I am OUT of this discussion.
Last edited:
- Status
- Not open for further replies.