In 1976 I was in my 4th and 5th semester of the Faculty of Electrical Engineering
. I was building amplifiers, was aware of TIM (slew rate I would say) problems and had my own experience that uA741 was completely unusable in power amplifiers in a signal path 
I know I can add the buffer. I am not sure if the decent buffers were available in 1976
Regarding noise, try it with very sensitive tweeter, add 10k to your low noise input.
Hi,
Why not?
Funerals are the perfect place.
You know, memento mori and all. Great opportunity for hitting on the YL using the carpe diem line... You can always recite "To his Coy Mistress" by Andrew Marvell outside the church...
"Had we but world enough and time..."
Ciao T
Why not?
Funerals are the perfect place.
You know, memento mori and all. Great opportunity for hitting on the YL using the carpe diem line... You can always recite "To his Coy Mistress" by Andrew Marvell outside the church...
"Had we but world enough and time..."
Ciao T
Hi,
Yet it happens frequently. More recently I remember the "improvements" to the Goldmund clone where exactly that happened (and had enthusiastic support from some self proclaimed audio designers extraordinaire)... So, it seems the strawmen are alive, well and walking.
Ciao T
Yet it happens frequently. More recently I remember the "improvements" to the Goldmund clone where exactly that happened (and had enthusiastic support from some self proclaimed audio designers extraordinaire)... So, it seems the strawmen are alive, well and walking.
Ciao T
I know I can add the buffer. I am not sure if the decent buffers were available in 1976
Regarding noise, try it with very sensitive tweeter, add 10k to your low noise input.
The final commercial product, the ML-2 (yes I know it
was modified in several ways compared to the JC-3, but
anyway) a buffer was used. It´s production started 1977.
Attachments
Hi,
Ideal does not exist, so it is untrue in that sense. Mathematically it is only true if use extreme abstractions that make the math useless for real world situation.
Once we add all the terms needed for a full analysis of the problem (which quickly becomes very complex) even mathematically: No.
Maybe not a "scientific" one, but from empirical one. I am very aware that empiricism is mainly frowned upon by the current scientific establishment, but that does not make an observation derived from empirical methods invalid.
Well, this is like telling Washing Powder manufacturers they should tell their customers just the cold hard facts, modern, capitalist and competitive markets do not work like that.
Ciao T
Ideal does not exist, so it is untrue in that sense. Mathematically it is only true if use extreme abstractions that make the math useless for real world situation.
Once we add all the terms needed for a full analysis of the problem (which quickly becomes very complex) even mathematically: No.
Maybe not a "scientific" one, but from empirical one. I am very aware that empiricism is mainly frowned upon by the current scientific establishment, but that does not make an observation derived from empirical methods invalid.
Well, this is like telling Washing Powder manufacturers they should tell their customers just the cold hard facts, modern, capitalist and competitive markets do not work like that.
Ciao T
I introduced that topology, too. (the comp-symmetry follower) back in 1970. We used it for making filters and buffers back in 1972, when I worked with the Grateful Dead, Wall of Sound. I also used it with IHEM in Switzerland, for crossover filters, and Mark Levinison used them in the LNC-2 electronic crossover, as well as the JC-2 phono preamp output follower.
Tom C. added it to the design of the JC-3 as part of the 'makeover' to increase the input Z.
Tom C. added it to the design of the JC-3 as part of the 'makeover' to increase the input Z.
Many of us use as little global negative feedback as possible, because we have found, that all else being equal, they sound better with less. We are fully aware of how feedback works and we normally use it. Personally, I prefer the gain structure of video IC's better, over audio IC's, because it has higher open loop bandwidth, to reduce, guess what? Differential phase distortion or PIM! You know that 'product of my imagination' '-)
PMA, I used the UA741 by the dozen for servo amps to maintain tape tension in a high speed video recorder in 1969. They worked great for that job, and they were the most popular analog IC, just about ever made.
However, it was difficult to PROVE that the UA741 was a lousy audio gain stage, because SMPTE IM looked reasonably OK with it. That is one of the reasons we had to develop the Sine-Square TIM test in 1976. Several audio manufacturers including 3M, used slow IC's like the UA741 for analog circuits. You won't find me making that mistake, but then I have always striven for higher quality to the point of being 'impractical'. What we did use at the time, was the HA911, that was about 5 times faster than the 741, but these failed a listening test with the Grateful Dead in 1971, and I had to go back to the 'drawing board' so to speak. They should have been fast enough, but they had their own problems as well, such as crossover distortion in many units. Mark Levinson, at the suggestion of Dick Burwen, used selected HA911 IC's in the LMP-2 preamp that preceded the Levinson JC-2, by about 1 year. This preamp was more fancy, had meters, tone controls, and YET it never got as popular as the JC-2. Ultimately I developed replacement modules for the LMP-2 as well, based on a discrete op amp design, with less operating current than the JC-2 modules. Not as good as the JC-2, but better than the Harris IC's. That is the fact of the matter.
However, it was difficult to PROVE that the UA741 was a lousy audio gain stage, because SMPTE IM looked reasonably OK with it. That is one of the reasons we had to develop the Sine-Square TIM test in 1976. Several audio manufacturers including 3M, used slow IC's like the UA741 for analog circuits. You won't find me making that mistake, but then I have always striven for higher quality to the point of being 'impractical'. What we did use at the time, was the HA911, that was about 5 times faster than the 741, but these failed a listening test with the Grateful Dead in 1971, and I had to go back to the 'drawing board' so to speak. They should have been fast enough, but they had their own problems as well, such as crossover distortion in many units. Mark Levinson, at the suggestion of Dick Burwen, used selected HA911 IC's in the LMP-2 preamp that preceded the Levinson JC-2, by about 1 year. This preamp was more fancy, had meters, tone controls, and YET it never got as popular as the JC-2. Ultimately I developed replacement modules for the LMP-2 as well, based on a discrete op amp design, with less operating current than the JC-2 modules. Not as good as the JC-2, but better than the Harris IC's. That is the fact of the matter.
You said it not me. Let me check around I might be able to propose a simple experiment with the AD844. Charles Hansen uses it and he's on your side. I can only provide facts, if you folks are going put up unblind listening as the final judgement then I'm wasting my time.
Statements of measurable fact have been made concerning these circuit techniques that can be proven as measurable phenomena to be wrong.
I think it's funny people quote Cherry's work when much of it concerns making amplifiers with extra gain stages and GOBS of low frequency gain stable in a closed loop application. This Maxim app note shows the two opamps in series are better than one basic idea. I think LT used this idea in their "distortionless" oscillator, it is an old idea used in our modules for making cheap 10^7 AOL op-amps out of junk parts. It's the sledge hammer approach of ultra high feedback.
Build High-GBW Op-Amp From A Dual Video Amplifier - Maxim
BTW John that JC-3 is a very nice circuit, parts efficient, symmetrical, and probably sounds fine. If you want to criticize some of GlenK and sny08's designs as being over engineered that's fine. If you say that they are no different than the Goldmund cloners (I don't read those threads) I disagree.
Last edited:
That's fine, but I mentioned a signal path, not the DC servo. They were unusable for the signal path with gain, for the reason that they started to make a triangle from sine at 10Vp somewhere near 8kHz.
Scott, I am not interested in blind testing or anything else that takes me off the path. It is pointless, and I would have given up further design improvements, if I were forced to, years ago.
When I have to, I use modern IC op amps, perhaps not the most expensive, in some of my latest designs. This month, at least one formal review of the Parasound JC-3 phono stage should be out. It uses 'pretty good' IC's and the review was VERY successful, because it is great 'bang for the buck'. In my evaluation of the prototype, I could hear differences in different IC's used in the input section of the design. I am not supposed to hear any differences, but I did. By the way, your 797 sounded OK in the unit, but I decided not to use it, due to the extra high cost of the device, and I found something that sounded as good, at 1/2 the price. But they did NOT sound exactly the same (for some reason) and a third device had to be rejected, because it did not sound right at all.
When I have to, I use modern IC op amps, perhaps not the most expensive, in some of my latest designs. This month, at least one formal review of the Parasound JC-3 phono stage should be out. It uses 'pretty good' IC's and the review was VERY successful, because it is great 'bang for the buck'. In my evaluation of the prototype, I could hear differences in different IC's used in the input section of the design. I am not supposed to hear any differences, but I did. By the way, your 797 sounded OK in the unit, but I decided not to use it, due to the extra high cost of the device, and I found something that sounded as good, at 1/2 the price. But they did NOT sound exactly the same (for some reason) and a third device had to be rejected, because it did not sound right at all.
I guess I can always expect a comment on the most irrelevent point. I asked if you would want to examine some of the claims made in a purely technical way. That is when someone says, "This linearizes X and lowers distortion" and it doesn't I don't want to get back, "But it made Jerry Garcia happy".
- Status
- Not open for further replies.