Figure 2 is only to show the wire breaking, place where T2 will be inserted. Cartoon style. You really think that I thought fig.2 is job done. Presentation 1, 2, 3.
Lazy Cat,
As usual picture tells more than words. T2 is the one to blame for the loss of organic quality of musical information in spite of claims that it is linearizing something and in spite of the fact that it cancels even order harmonics. It may do so but at what cost? Solution is worse than not solving the original problem.
This is all connected with civilization's Myth of progress. Technical developments are in numerous cases steps backward. People that listen Horowitz LPs and CDs are usually dissatisfied with quality of recordings done with "superb" SS equipment, and more than pleased with recordings made with "retro" valve technology. Specs do not tell the whole story.
Circuit designers are hard pressed to follow certain routine just because they are afraid that their product will be marked as having bad specs. The only way to prove that they are competent and that their product is "better" is to present some measurements And that is how we came to the point that most amps are VF, filters for musical emotion. Nothing comes through.
As usual picture tells more than words. T2 is the one to blame for the loss of organic quality of musical information in spite of claims that it is linearizing something and in spite of the fact that it cancels even order harmonics. It may do so but at what cost? Solution is worse than not solving the original problem.
This is all connected with civilization's Myth of progress. Technical developments are in numerous cases steps backward. People that listen Horowitz LPs and CDs are usually dissatisfied with quality of recordings done with "superb" SS equipment, and more than pleased with recordings made with "retro" valve technology. Specs do not tell the whole story.
Circuit designers are hard pressed to follow certain routine just because they are afraid that their product will be marked as having bad specs. The only way to prove that they are competent and that their product is "better" is to present some measurements And that is how we came to the point that most amps are VF, filters for musical emotion. Nothing comes through.
JL Hood never called his design a "current feedback" amplifier; that's because it is nothing of the sort.
Quite correct, IMHO.
It is also in my view not a good amplifier, but let's not stray too far from the topic.
If anyone wants to discuss it please let us have a new thread.
Last edited:
I don't really have a dog in this fight. I see the point Michael is making, but I am quite happy to treat "what is normally described as a CFA" as a CFA.
I haven't heard anything yet to alter my view that CFA is just a more difficult way to make an amplifier.
Yes, that is rather the elephant in the drawing room. Let me lead it forward by the trunk....
Seki, K “Multistage Amplifier Circuit” US Patent 4,145,666, 20th Mar 1979
Assigned to Hitachi.
While it has a deliberately obfuscatory title if I ever saw one, that patent is actually wholly devoted to TMC/Output-Inclusive compensation. The granting of the patent seems to show it was novel at the time, but as others have pointed out, it is a very simple circuit (once you know it) that I would have thought opamp designers would have discovered in the 1960's. A search through opamp stuff (and there is a huge amount of it) might trace its history back further. There is probably no point in going as far back as valve circuitry, because unity-gain output stages were not normally used.
Patent examiners are not infallible, and the granting of a patent does not guarantee that said patent is not invalid because of prior usage. On the other hand, I have found them to be pretty thorough in Britain. I had to tweak one of my patents to avoid a bit of prior art in 1909- by some fellow called Marconi.
Not so. There will be in the new book extensive coverage of several kinds of push-pull VAS, double input stages, Borbely-Lender, and so on. Which is best? The answer might surprise you...
You know I don't do proof by assertion.
Right. Let us have a list of every CFA you know of, so we can compare the numbers with how many VFAs exist.
With buffers power amps exhibits better specs , particularly
the ones that matters for sound quality.
Quite so.
I am not sure that a car radio chip would be my first choice in aiming at high quality audio.
I am however sure, having spent more time than I care to contemplate in double-blind listening tests, that none of these qualities you list have any existence. Think what you are saying: an amplifier can alter the rhythm of music? Bossa nova in, Tango out?
Hello Bob
Examples are in Fig 6.31, Fig 7.18, and Fig 8.9. (5th edition) These are all pretty much the same, apart from differences in the output stage, and whether or not protection circuitry is shown.
Fig 10.19 is also Blameless, but the circuitry is complicated by the precision Class-A biasing system.
My favorite test for audio amps is to play orchestral recording with two instruments playing in unison. For instance, Freddie Hubbard and Dexter Gordon intro theme in Hancock's "Watermelon Man" or two violin intro of 4th Movement of Handel's Concerti Grossi. VF amps never differentiate two instrument colors in proper way, one violin is only vague echo of the other. CF amps shine with such difficult tasks, plus pace is much better. You feel like dancing while listening to music.
Rather than clutter up this thread, perhaps you could level match them, record the signals out of the amps, and post the original and Diffmaker files in a new thread?
Not in this position, it just worse the input stage, T2 (fig.3, post #160) in VFA makes the amp of no use. Do we make amps to amplify trivial signals from the test equipment or to amplify audio signals to be listenable?
I agree it gets fairly close to Blameless performance, but it's a good deal more complicated. I think it's an awkward way to make an amplifier, but I certainly do not assert you cannot make a good amplifier that way.
I hope they weren't the sort of IGBT's that used to latch hard on above a certain current.
Seems no one has read your 5th edition, from the bottom of page 132 chapter 5 till page 134 a CFB topology is shown although you refer to it as " Series differential input configuration. Some pitfalls are discussed but not linearity.
Just as youve tweaked the old lin topology into the blameless having very low THD performance so this CFB topology can be tweaked to perform equal and in some aspects better than your standard blameless circuit using miller compensation, although use of TMC swings it back in favour of blameless.
Most commercial offerings using CFBs are not aiming at lowest THD so you dont see optimised circuits using the topology. As an example, introducing a beta enhanced vas as used with the circuit shown on page 134 (figure 15.4) markedly improves its performance and at frequencies higher than 5 KHZ no standard blameless can match its linearity. For ultimate low THD I find the Vfb topology best but cfb´s have constant low thd accross all audio frequencies and at higher frequencies will better standard vfb types such as the blameless.
Hi Doug,
Not sure what the complication is that you refer too? At first sight, the MA CFA looks to be a rather strange, unusual arrangement.
1) Comparing PCB area,# of components wise?
MA/CFA implementation has less brd area, than the ckt I see referenced in fig 5.24, p134,ed 3. of APAD Handbook (Load Invariant) and what is posted at Signal Transfer Co., (http://www.signaltransfer.freeuk.com/invarint.htm)comparing a TH part implementation.
2) Circuit complexity? as in pcb nets (not including what is inside the SSM2131) , MA CFA, is also less for a MA/CFA than Blameless/Load Invariant.
3) Complexity in analysis? I agree the MA CFA is more complicated to analyze, since I could not Pspice it and could for the Blameless/LI!! But that is only my limited abilities.
What´s the difference between them???Do we make amps to amplify trivial signals from the test equipment or to amplify audio signals to be listenable?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.