Once again, slew rate higher than requested by input signal is a necessary condition. No sane designer sticks with only one parameter, but must fulfill the necessary condition. Everyone knows that only a complex set of parameters makes a good amplifier.
IMO you are playing game with words only and you are trying to amaze crowds totally unacquainted with analog design, like amateurs and audio magazine readers
Regarding 'limitations' - probably a learning process? I believe that in a month or two you will be more familiar with your new test equipment. It is more and more tough as we are getting older
Ah, that's interesting and just how I remember it when I tested it with regard to slew rate.
"IMO you are playing game with words only and you are trying to amaze crowds totally unacquainted with analog design, like amateurs and audio magazine readers "
That's exactly the issue here. This thread is frequented by serious, very well educated STEM professionals along with amateurs (mostly highly competent ones at that). So, trying the 'back in '74 ' routine, or proposing things that do not stand up to scientific or engineering scrutiny is going to lead to one outcome only: gurus will not go unstoned
That's exactly the issue here. This thread is frequented by serious, very well educated STEM professionals along with amateurs (mostly highly competent ones at that). So, trying the 'back in '74 ' routine, or proposing things that do not stand up to scientific or engineering scrutiny is going to lead to one outcome only: gurus will not go unstoned
Last edited:
I would be interested in whether DA is always linear in behaviour, or whether it can manifest non-linear characteristics ...
I wrote a short little thinker piece many years ago titled: "How Linear distortion can sound non-linear". Anyone remember seeing that piece?
THx-RNMarsh
Amplifying the input error signal on an op-amp to measure distortion (a la Bob Pease's LME data sheets) was the first thing I was shown at work. The test box was dated 1969. Slew enhancement at the input did not improve distortion because the input voltage to current transfer had a kink in it as it transitioned from the underlying slow slew to the boosted slew. In addition the GBW was still that of the slow amp. Degeneration OTOH was a far smoother function (as were JFET's for that matter) which became our design standards of choice.
If you read the real 1741S data sheet you will notice, GBW is still only 1MHz and distortion for power BW is 5% (no not .5%), a part almost designed to be obsolete from the get go. No TIM test necessary all the info is right there.
So again I ask can anyone show a commercial audio power amp where someone wanted just a big slew number for the spec page and used a switching slew boost rather than degeneration or JFET's? There might be one, I have not seen it but I don't research the matter.
Last edited:
J. C.
Sorry I have been off doing real work. Sorry to miss the dump on you spurt. So here is my contribution:
All this thirty year old stuff doesn't matter anymore! (P.S. thanks for the direct box circuit you used for the Grateful Dead works brilliantly even though it is made with obsolete FETs. Testing it here against a modern one wasn't even a close contest.)
What nonsense that a larger power transformer in a preamp would make a difference. The reports that it improved resolution would not stand up to a double blind test. (It is just coincidence that measurements show the larger transformers have much less saturation distortion, and some tests show that this propagates also as an excess magnetic field.)
Your use of capacitors as shunts in phono equalization stages instead of the normal series use can't have any difference as they are still in the circuit. (Of course a careful analysis showing that shunt versions of the circuit have lower distortion is at a level that obviously you can't hear in a DB test.)
Now I should also poke some fun at Dick. It is complete nonsense that changing a .22 ohm emitter resistor to .1 ohms would have any difference. I don't care that J. C. has mentioned that .1 ohm is the optimum value for a number of reasons. And that bit about distortion levels of -100 db thd can't be heard. Everyone knows that 2nd harmonic pretty much isn't a problem below -20 db. (Of course if we look at 16th harmonic, the musical energy at 150 Hz is about 15 - 20 db greater than at 2400 hz and Fletcher Munson did a bit of research that says the 150 hz notes would appear to be of equal loudness to a 2400 hz tone 20 db down. So we are perceiving a signal to noise ration of 60 db or so. Surely enough for state of the art fidelity!)
ES
Sorry I have been off doing real work. Sorry to miss the dump on you spurt. So here is my contribution:
All this thirty year old stuff doesn't matter anymore! (P.S. thanks for the direct box circuit you used for the Grateful Dead works brilliantly even though it is made with obsolete FETs. Testing it here against a modern one wasn't even a close contest.)
What nonsense that a larger power transformer in a preamp would make a difference. The reports that it improved resolution would not stand up to a double blind test. (It is just coincidence that measurements show the larger transformers have much less saturation distortion, and some tests show that this propagates also as an excess magnetic field.)
Your use of capacitors as shunts in phono equalization stages instead of the normal series use can't have any difference as they are still in the circuit. (Of course a careful analysis showing that shunt versions of the circuit have lower distortion is at a level that obviously you can't hear in a DB test.)
Now I should also poke some fun at Dick. It is complete nonsense that changing a .22 ohm emitter resistor to .1 ohms would have any difference. I don't care that J. C. has mentioned that .1 ohm is the optimum value for a number of reasons. And that bit about distortion levels of -100 db thd can't be heard. Everyone knows that 2nd harmonic pretty much isn't a problem below -20 db. (Of course if we look at 16th harmonic, the musical energy at 150 Hz is about 15 - 20 db greater than at 2400 hz and Fletcher Munson did a bit of research that says the 150 hz notes would appear to be of equal loudness to a 2400 hz tone 20 db down. So we are perceiving a signal to noise ration of 60 db or so. Surely enough for state of the art fidelity!)
ES
I haves seen it spelled audiophliac .Thus defined as an unnatural love of audio. as opposed to audiophilephiliac an un natural love of audiophiles.
Scott, I might agree with you that the 1741S (now that it is recognized as actually existing) was a short term solution to a real problem, i.e. how do you get a good slew-rate out of a process that can only make the equivalent of 741's normally? I think it is a brilliant, but flawed solution to the problem, 40 or more years ago.
In principle, why can't modern op amps utilize a similar technique? Then perhaps 20V/us might become 200V/us or even more, and with the now significant change in the rest of the topology or processing? This (breakthrough) would create a new improved IC and complete the process of design to phase 3. You know: you invented it. '-)
In principle, why can't modern op amps utilize a similar technique? Then perhaps 20V/us might become 200V/us or even more, and with the now significant change in the rest of the topology or processing? This (breakthrough) would create a new improved IC and complete the process of design to phase 3. You know: you invented it. '-)
- Status
- Not open for further replies.