Hi to everybody!
I just finished to read the douglas Self' book -Audio Power Amplifier Design Handbook.
It was my first "Audio Book".
By reding it, i found out that the ideas of Daglas about the circuity topology are are very different from what i have been set about audio equipments.
He considers the feedback as a weapon to use.
He suggests to use as much feedback as you can!
I have been used from the magazines to consider high feedback as a bad thing.
I 've been always thought that class a amp with low feedback is the best choice.
I would like to know your view point about this guys.
Just a thought, i would have rather preferred to read togheter with the technical datas, also sounds impression about the different configurations and not just tdh measuements.
I'm very curious to read your posts.
Best Regards,
Stefano.
I just finished to read the douglas Self' book -Audio Power Amplifier Design Handbook.
It was my first "Audio Book".
By reding it, i found out that the ideas of Daglas about the circuity topology are are very different from what i have been set about audio equipments.
He considers the feedback as a weapon to use.
He suggests to use as much feedback as you can!
I have been used from the magazines to consider high feedback as a bad thing.
I 've been always thought that class a amp with low feedback is the best choice.
I would like to know your view point about this guys.
Just a thought, i would have rather preferred to read togheter with the technical datas, also sounds impression about the different configurations and not just tdh measuements.
I'm very curious to read your posts.
Best Regards,
Stefano.
of course, there are a lot of good points to learn and take into account!
What i would like to know is: what the people here on the forum think about his way of design amp.
..about the feedback.....i would expecially like to know the effects on the sound.
how would an amp with wide feedback would sound compare with one at low feedback.
Furthermore i would like to know what do you guys think about class b and class a: the difference on the sound.
I would also like to know .... what's the best audio authour for you guys!
Regards,
Stefano.
What i would like to know is: what the people here on the forum think about his way of design amp.
..about the feedback.....i would expecially like to know the effects on the sound.
how would an amp with wide feedback would sound compare with one at low feedback.
Furthermore i would like to know what do you guys think about class b and class a: the difference on the sound.
I would also like to know .... what's the best audio authour for you guys!
Regards,
Stefano.
Doug Self writes a lot of good stuff but doesn't seem to believe in listening to things other than maybe the final developed design. I think you should always listen as well as measure all along the way. I also wonder if he even knows what to listen for.
I have to say most magazines I have seen have been written by clueless/blinded by paradigm people really, perpetuating myth and trying to look good by using scientific concepts, although wrongly most of the time.
I think that ways of doing things is just as important if not more than simply taking feedback as a metric to judge sound quality. Same goes for class-a vs class-b.
I rate John Linsley Hood as an author and an audio engineer, but there are many good and bad people about.
I have to say most magazines I have seen have been written by clueless/blinded by paradigm people really, perpetuating myth and trying to look good by using scientific concepts, although wrongly most of the time.
I think that ways of doing things is just as important if not more than simply taking feedback as a metric to judge sound quality. Same goes for class-a vs class-b.
I rate John Linsley Hood as an author and an audio engineer, but there are many good and bad people about.
hi
good topic = negative feedback level
my own amateur DIY approach:
Total open loop gain = 100 x closed loop gain
( this is a feedback of 40dB )
I have figured this will give good quality
while avoiding some of the bad drawbacks of very high feedback
such as extensive compensation filtering
why overdo things?
too much cocaine will kill you
too much of anything is bad
Try the sensible way of:
"Just use whatever much you need, and no more!"
as is said by
the old familiar lineup proverb, expression
good topic = negative feedback level
my own amateur DIY approach:
Total open loop gain = 100 x closed loop gain
( this is a feedback of 40dB )
I have figured this will give good quality
while avoiding some of the bad drawbacks of very high feedback
such as extensive compensation filtering
why overdo things?
too much cocaine will kill you
too much of anything is bad
Try the sensible way of:
"Just use whatever much you need, and no more!"
as is said by
the old familiar lineup proverb, expression
Hi,
In the standard amplifier topology considered by D. Self in his book
and used in the vast majority of commercial amplifiers in existence,
the use of feedback is implicit, it allows the topology.
D. Self gives lots of insights into this topology and how it can be
optimised in objective performance terms, and avoids getting
embroiled in subjective and meaningless debate.
Nevertheless there is plenty of information is his book regarding
optimising subjective sound quality (depending on the principles
you choose) if you can read between the lines of the information
he gives.
However the coverage of amplifier types is deliberately limited.
Subjective performance is not directly considered either, the
point of the book is how to avoid / minimise THD, as its the
most quoted parameter, and the mechanisms to do this.
It is not a good book as an introduction to "the sort of amplifiers
manafacturers foist upon us with dubious claims of superior
performance for reasons that cannot be proved and are usually
simply wrong" or understanding valve amplifier topologies where
low feedback is a given, as high feedback cannot be used, and
the common topologies used do allow no feedback to be used.
The fact is nearly all music you listen to has passed through
high feedback amplifiers many times before it gets to your
system, so high feedback in itself is not implicitly wrong.
/sreten.
In the standard amplifier topology considered by D. Self in his book
and used in the vast majority of commercial amplifiers in existence,
the use of feedback is implicit, it allows the topology.
D. Self gives lots of insights into this topology and how it can be
optimised in objective performance terms, and avoids getting
embroiled in subjective and meaningless debate.
Nevertheless there is plenty of information is his book regarding
optimising subjective sound quality (depending on the principles
you choose) if you can read between the lines of the information
he gives.
However the coverage of amplifier types is deliberately limited.
Subjective performance is not directly considered either, the
point of the book is how to avoid / minimise THD, as its the
most quoted parameter, and the mechanisms to do this.
It is not a good book as an introduction to "the sort of amplifiers
manafacturers foist upon us with dubious claims of superior
performance for reasons that cannot be proved and are usually
simply wrong" or understanding valve amplifier topologies where
low feedback is a given, as high feedback cannot be used, and
the common topologies used do allow no feedback to be used.
The fact is nearly all music you listen to has passed through
high feedback amplifiers many times before it gets to your
system, so high feedback in itself is not implicitly wrong.
/sreten.Usually feedback amps have open loop gain which is very variable with frequency (to achieve stability), and since feedback is the more effective the more gain you have, this means that in certain regions of the spectrum the residual error and all the effects of feedback will be higher than in other regions. Feedback also needs very good circuit layout to avoid feedthrough, parasitic oscillations or (even worse) behaviors at the edge of instability, all frequency-dependent so that sound is affected in unpredictable ways. Moreover, you should dimension things so that in all circumstances you will never go anywhere near to clipping, which is much nastier in fed-back circuits (is this term OK? I have invented it right now).
I am not an expert, but I suspect that if people knew how to use feedback it would not have got its bad fame. As a matter of fact, who knows how many audiophile records have been recorded through tens (100's) of operational amplifier circuits, each with its own feedback, but seriously designed so that professionals could use them reliably.
--st.r.
I am not an expert, but I suspect that if people knew how to use feedback it would not have got its bad fame. As a matter of fact, who knows how many audiophile records have been recorded through tens (100's) of operational amplifier circuits, each with its own feedback, but seriously designed so that professionals could use them reliably.
--st.r.
Stefanoo said:
You can learn a lot from that book. However, do not make the mistake of taking everything he says as "gospel". Remember, Self earns a living as a design consultant to the "Big Box" manufacturers, and he has to play by their rules if he expects future employment. His "Blameless" has been built by many a DiYer, and the sound is frequently described as "lifeless". IOWs, it sounds like a Big Box system.
Self's other big error is in being way too doctrinaire. His mantra: If you can't measure it then it doesn't count, is just plain wrong. This has led him to make an unwarranted denunciation of the quasi-complementary final. The "numbers" say that this topology is inferior. However, the ears say otherwise. I've tried it both ways -- full complementary and quasi-complementary. I prefer Q-C by a wide margin, and would never consider anything else.
He's partially right about this. Negative feedback is highly useful, but like anything else, can easily be misused. The most common abuse of NFB is using it to cover for a fundamentally bad open loop design. This is what gives NFB a bad reputation. This need not be the case.
Again, he's partially right about this. In my experience, a BJT implementation can stand a very large NFB factor. Indeed, a BJT amp seems to demand this if it's to sound good. I've got a set of 3.0W BJT monoblocks here that are serving as computer speaker boxes. The main limitation to the sonics is the speakers (car radio replacements) not the amp itself. Self discourages the use of MOSFET finals. Given the most likely implementations, he's right to do so.
A MOSFET based design can't tolerate quite so much NFB, and the big mistake with these devices is the adaptation of designs originally intended for Class AB BJTs. These designs sound just plain Gawd-awful. MOSFETs aren't like BJTs, and so require different topologies to work right.
Vacuum tube designs don't tolerate very much NFB at all. Throw in excessive NFB, and an otherwise good sounding VT amp will sound as bad as any solid state amp, if not worse.
Anyone can publish a magazine. That doesn't mean that they know what they're talking about, and frequently they propagate lots of unsubstantiated audiophool folk "wisdom". Believe maybe 50% of what you see, and 25% of what you read and you can't go too wrong.
Maybe yes; maybe no. For a MOSFET implementation, that's probably the best way to go, given the nasty crossover behaviour of this device. For VTs, it's usually correct, but not always, in that there are lots of Class A(2), AB(1), and AB(2) implementations that do produce excellent sonics. For BJTs, it's not so important.
In the end, that's what truly counts when it's time to put away the test equipment, connect the speeks, and see if all that design work paid off in the form of something you like to listen to.
I would like to support everything Miles just wrote in his excellent summary.
BTW I'm not sure what Self is measuring when he decries quasi complementary but I'm able to measure the superior performance of this topology, all other things being equal.
The key to success lies in determining what "bad" means to feedback. The reason for different tolerance, if you will, in BJT, FET and tube systems is entirely deterministic and measurable.
BTW I'm not sure what Self is measuring when he decries quasi complementary but I'm able to measure the superior performance of this topology, all other things being equal.
janneman said:
I almost never use very large bandwidth.
Because I cant hear too high frequencies
and if i could, most 95% of music and sounds are within the audible spectra.
hehe - obvious, isnt it ....
But lets say my target values for amplifiers are:
10 - 100/200 kHz.
The actual value is of course dependent on
what kind of amplifier I will build.
A power amplifier seldome needs the bandwidth of a preamp, for example.
But I am more into the policy of making very good open loop performance ( a lot of emitter resistors )
within a more narrow bandwidth,
than being hunting for enormous figures in terms of high freq.
Very high freq may look good at your oscilliscope screen
with a pure resistive load,
but in real working environment such circuit can run into serious troubles.
One simple issue is radio waves transmitters causing unwanted effects!
Now the mentioned 100x, 40dB figure is not an absolute number for me.
It can be different depending what type of amp.
But what I try to avoid is MAXIMAL open loop gain.
---------------------
By the way, a question:
What is the lowest negative feedback factor possible?
If you want to use any negative feedback for correcting the signal.
In my mind it is 6dB.
Open loop gain is 2 x closed gain.
Is this correct?
At least I have drawn several such pre-amplifiers, on my papers.
Miles,
Top stuff. Agree emphatically about Doug. His dogmatic approach is a weak point; blinds him to the bleeding obvious.....
# that two amps can have the same fb factor yet sound different.
# that lower distortion measurements fail to correlate with listener preferences, taken over large samples.
# that the engineering elegance of the CFP is not matched by its sonic performance, particularly when strapped capacitively for good stability.
# that lag compensation, and it's first cousin, phase lead, are probably more useful weapons to 'voice' the sonics than any other single parameters.
A particularly damning omission in Self's work is his glossing over the input stage. He does not draw attention to either memory effects (owing to power variations in the tiny dies in this and the VAS stages) or to the variance of the Vbe across the diff pair with current variation through these devices.
My comments, as always, are qualitative; I've not done the math but there is no doubt that a diff pair set up for constant current operation has sonic benefits.
Cheers,
Hugh
Top stuff. Agree emphatically about Doug. His dogmatic approach is a weak point; blinds him to the bleeding obvious.....
# that two amps can have the same fb factor yet sound different.
# that lower distortion measurements fail to correlate with listener preferences, taken over large samples.
# that the engineering elegance of the CFP is not matched by its sonic performance, particularly when strapped capacitively for good stability.
# that lag compensation, and it's first cousin, phase lead, are probably more useful weapons to 'voice' the sonics than any other single parameters.
A particularly damning omission in Self's work is his glossing over the input stage. He does not draw attention to either memory effects (owing to power variations in the tiny dies in this and the VAS stages) or to the variance of the Vbe across the diff pair with current variation through these devices.
My comments, as always, are qualitative; I've not done the math but there is no doubt that a diff pair set up for constant current operation has sonic benefits.
Cheers,
Hugh
my view on this subject is that Self is a Legend my only disappointment is he is not perfect,who is?
i think Self is up there with pass
my view on feedback is that it does effect greatly amplifier
performance so does no nfb, an amplifier in oder to sound
good must be designed good and this is a very complicated
matter but on a schematic might appear as a simple one
as self has said demystifying amplifier design has been his
goal with blameless and nfb, linearity and distortion are his
prime subjects and thats what his little basics book is all
about people say is an advanced book of course it is if
youre a complete beginner self will appear as giant in his
book
measurements are important and thats where self has based
the facts in his book
john
That's a little hard on Nelson
You want to design and build your own high performance sports car. Where do you go? Do you visit the guy down the road in the small shop which builds go-carts and bicycles and sells a book entitled: "high performance road vehicle design guide". No. Instead, you phone up the design department of Ferarri (or the design house you most respect) and ask them to share their expertise with you. They refuse but at least they do tell you exactly where to go!
Such is the nature of performance audio.
You want to design and build your own high performance sports car. Where do you go? Do you visit the guy down the road in the small shop which builds go-carts and bicycles and sells a book entitled: "high performance road vehicle design guide". No. Instead, you phone up the design department of Ferarri (or the design house you most respect) and ask them to share their expertise with you. They refuse but at least they do tell you exactly where to go!
Such is the nature of performance audio.
darkfenriz said:
you are probably right
but good thing I cant here much at 'borders of audio sepctrum'
it is in the very word of 'audio'
audio is something else for other animals ... like bats and dolphins
-----------------------------------------------
'nother very important thing, not mentioned, yet:
Where is this gain located?
Say 3 stages, which is not unusal.
If I have 100x feedback, that is for an amplifier with voltage gain=20
an open loop gain of 2.000
----------
Amp A.
input: x2
voltamp: x1.000
endstage: x.95
sum gain: 1.900
Amp B.
input: x20
voltamp: x1.00
endstage: x.95
sum gain: 1.900
----------
These two amplifier, A and B, although both have ~40dB feedback
are in electrical sense different
and surely in performance end results.
I have touched the subject before:
Open loop gain in stages of amps.
in this article and most surely in some other postings of mine.
http://www.diyaudio.com/forums/showthread.php?postid=839294#post839294
lineup - Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.