Good book!
very in depth! slightly over my head in parts,im gona get it out again after my eletrotech exams so i can read properly
topics
the 'fear' of -ve feedback (hehe)
small input stages
output stage 1,2
compensation,slew rate and stability
power supplies and PSRR
class A
FET output stages
Thermal compensation and thermal dynamics
Protection
Grounding and practical matters
i like the 8 distortions part of the book,very good.action packed goodness!
he has distortion down to 0.005%and finds out that a certain technique lowers it by four fold by complicating the circuit by about 4x, but the distortions get very low from what he does.
He goes through each topology ,input /output stage type and its features- its good! i havent seen a book like it here in NZ
Cheers!
have u read? what do u think?!!!
very in depth! slightly over my head in parts,im gona get it out again after my eletrotech exams so i can read properly
topics
the 'fear' of -ve feedback (hehe)
small input stages
output stage 1,2
compensation,slew rate and stability
power supplies and PSRR
class A
FET output stages
Thermal compensation and thermal dynamics
Protection
Grounding and practical matters
i like the 8 distortions part of the book,very good.action packed goodness!
he has distortion down to 0.005%and finds out that a certain technique lowers it by four fold by complicating the circuit by about 4x, but the distortions get very low from what he does.
He goes through each topology ,input /output stage type and its features- its good! i havent seen a book like it here in NZ
Cheers!
have u read? what do u think?!!!
Great book.
I think this is a great book. It does examine the distortions as you said. It's not all maths and lots of it has measured results.
The only catch is that while one can keep trying to get distortions down , will it reallly show up well in listening tests ? The no feedback tube amps sound very good with large doses of distortion. Probably there should be a chapter devoted to the audible implications of distortions with some mention of some great amps and their audible performance and their electrical specifications.
I have this book and it needs to be read many times .
Cheers.
I think this is a great book. It does examine the distortions as you said. It's not all maths and lots of it has measured results.
The only catch is that while one can keep trying to get distortions down , will it reallly show up well in listening tests ? The no feedback tube amps sound very good with large doses of distortion. Probably there should be a chapter devoted to the audible implications of distortions with some mention of some great amps and their audible performance and their electrical specifications.
I have this book and it needs to be read many times .
Cheers.
Self's stuff is very good indeed. It's not comprehensive, but in context, enormously useful.
I'm unconvinced that reducing distortion from, say, 0.1% to 0.001% makes much difference to the ear, but it's an interesting exercise to see how it's done. The biggest missing x-factor in Self's work is the thing that I've found actually makes real audible differences between otherwise-well-designed amps: overload recovery and clipping behavior. The comprehensive book on that for solid state circuits has yet, to my knowledge, been written. (Crowhurst repeatedly pounded that drum for tube circuits, and he was absolutely right)
I'm unconvinced that reducing distortion from, say, 0.1% to 0.001% makes much difference to the ear, but it's an interesting exercise to see how it's done. The biggest missing x-factor in Self's work is the thing that I've found actually makes real audible differences between otherwise-well-designed amps: overload recovery and clipping behavior. The comprehensive book on that for solid state circuits has yet, to my knowledge, been written. (Crowhurst repeatedly pounded that drum for tube circuits, and he was absolutely right)
Yes, I agree that vanishingly low (sinewave) distortion is not mandatory.SY said:
Sonically acceptable overload behaviour however is mission critical to good sonics.
Many amplifiers suffer 'stiction' when recovering from momentary overload, and this is both ear and speaker damaging.
Clean and benign overload behaviour is a lot of what differentiates a good sounding amplifier from a 'bad' one.
Soar protection circuitry can also be a culprit here.
Eric.
Books
I found the book (second edition) very interesting. I like the way it’s based on experiments combined with reason. I’m looking forward to reading the latest third edition.
However, I don’t understand why Douglas Self completely avoids the topic about symmetric amplifiers.
This topic is covered to some detail in Randy Slones book on power amplifiers, but I would like some more theory and detail in this book.
Also, I need a book (maybe just a chapter in one of the mentioned books) on how different sounding amplifiers can measure more or less the same. I’d also like to read more about no feedback amps.
\Jens
I found the book (second edition) very interesting. I like the way it’s based on experiments combined with reason. I’m looking forward to reading the latest third edition.
However, I don’t understand why Douglas Self completely avoids the topic about symmetric amplifiers.
This topic is covered to some detail in Randy Slones book on power amplifiers, but I would like some more theory and detail in this book.
Also, I need a book (maybe just a chapter in one of the mentioned books) on how different sounding amplifiers can measure more or less the same. I’d also like to read more about no feedback amps.
\Jens
The first difference would be that most published distortion figures are into a resistive load, and most amplifiers change when driving a reactive load.
Perhaps a measurement standard should be devised for measuring distortion when driven into a specified amount of overload - this may be revealling of overload behaviour in spec sheets.
Eric.
Perhaps a measurement standard should be devised for measuring distortion when driven into a specified amount of overload - this may be revealling of overload behaviour in spec sheets.
Eric.
I was about to post a new thread on Self's Power Amp book and then found this one.
I started reading the 6th edition online and am enjoying it so much that I'm going to
buy it. I borrowed an earlier edition from the library over 5 years ago and skimmed
it but this version I am really enjoying.
I'm going to have some questions but wanted to finish reading the relevant sections.
Why so little posting in this thread as compared to Bob's?
I started reading the 6th edition online and am enjoying it so much that I'm going to
buy it. I borrowed an earlier edition from the library over 5 years ago and skimmed
it but this version I am really enjoying.
I'm going to have some questions but wanted to finish reading the relevant sections.
Why so little posting in this thread as compared to Bob's?
I have the second edition, and it was good reading at that time. I now mainly consult it for practical building considerations, also useful for heavily changed or completely different approaches ...
Does the last edition contain in-depth analysis of other topologies than the very classical one?
Matthias
Last edited:
So, maybe, it's worth to look at it again.
What I really liked already in the 2nd edition: he doesn't take anything for granted or possibly trivial. Good approach, even if it may become a bit restricting in the still limited space of a single book.
Does 6th edition do more on the Overload & Clipping Behaviour that SY mentions in #3?
I suscribe to his point of view that this is probably the most important aural distinction between amps of about the same power.
If you wanna know why various Golden Pinnae amps sound different, just check how they overload/clip on real speakers.
You don't need Golden Pinnae to do this. Just an oscillator, scope and an old guitar speaker you don't mind blowing up.
I suscribe to his point of view that this is probably the most important aural distinction between amps of about the same power.
If you wanna know why various Golden Pinnae amps sound different, just check how they overload/clip on real speakers.
You don't need Golden Pinnae to do this. Just an oscillator, scope and an old guitar speaker you don't mind blowing up.
I've done a bit on that. In my previous life, I was a DBLT guru and can pontificate at length on AUDIBLE speaker distortions.
For amplifiers, 0.01% is probably well into inaudibility .. even with true golden pinnae and mild xover distortion
At 0.1%, a VERY small number of people can distinguish these from 0.01% amps. Here, the 1969 JLH Class A amp, with its nice 2nd harmonic distortion profile dropping rapidly with level, is often 'liked'.
Of course, these comments are about unclipped amps which are stable into real speakers.
The other audible distinguishing feature of Golden Pinnae amps is that most (all?) show bursts of oscillation on real speakers ... depending on the level, thermal & musical history too.
On the other hand, there is modern 'music' (?) where there is no audible difference if the amp is clipping 50% of the time.
Last edited:
Thirty-odd years ago, the National research Council of Canada (NRC) conducted several tests that revealed overwhelmingly that most listeners preferred the sound of an audio system that had flat frequency response and low-distortion - the flatter the better and the lower-THD the better.
Electronic distortion is inherently alien to our ear since we did not evolve processing such distortions. Instead, we evolved to overcome distortions and masking that interfere with recognition of sounds in our natural environment - predators, food, friends. The kinds of distortion a circuit can make is unlike what happens in nature and depending on how extreme that electronic distortion is, it takes a lot of cerebral energy to make what we are listening to intelligible.
On that basis, I have to agree with others who believe that making any and all distortions in a circuit as small as possible is the right way to go design-wise. Yes, some circuits have "pleasing" distortion, but only with airy music and at low loudness. JLH turns to mud with pop and rock that is heavy in the mids but sounds great with solo voice or chamber music.
Self is a bit imperious in his views and his writing. His design work concentrates on the ubiquitous Lin amp with diff-input - some say this is a Thompson? Anyway, it is very inexpensive to build (which appeals to his English thrift) and to also get much better performance from than what Self shows.
With all the editions of Self's book you get to pay again for the same information that was in the 1995 Wireless World magazines, with no corrections to reflect later changes in his thinking. The 6th has a chapter about symmetric amps where he cripples it by using MPSA42/92 throughout - not how anyone builds them - and another chapter about class-G amps of one format. There are other ways designers use to step the output than what he shows but his design is certainly the least expensive way to do it.
He has added some stuff from other articles he has written over the years, so you can pay for those again, along with material about line-inputs from the Small Signal design Handbook - which itself is a good book.
I learned a lot from these books but they also have huge gaps in circuit explanations that you have to fill in from other sources. I believe his target audience is not hobbyists but novice audio engineers.
Electronic distortion is inherently alien to our ear since we did not evolve processing such distortions. Instead, we evolved to overcome distortions and masking that interfere with recognition of sounds in our natural environment - predators, food, friends. The kinds of distortion a circuit can make is unlike what happens in nature and depending on how extreme that electronic distortion is, it takes a lot of cerebral energy to make what we are listening to intelligible.
On that basis, I have to agree with others who believe that making any and all distortions in a circuit as small as possible is the right way to go design-wise. Yes, some circuits have "pleasing" distortion, but only with airy music and at low loudness. JLH turns to mud with pop and rock that is heavy in the mids but sounds great with solo voice or chamber music.
Self is a bit imperious in his views and his writing. His design work concentrates on the ubiquitous Lin amp with diff-input - some say this is a Thompson? Anyway, it is very inexpensive to build (which appeals to his English thrift) and to also get much better performance from than what Self shows.
With all the editions of Self's book you get to pay again for the same information that was in the 1995 Wireless World magazines, with no corrections to reflect later changes in his thinking. The 6th has a chapter about symmetric amps where he cripples it by using MPSA42/92 throughout - not how anyone builds them - and another chapter about class-G amps of one format. There are other ways designers use to step the output than what he shows but his design is certainly the least expensive way to do it.
He has added some stuff from other articles he has written over the years, so you can pay for those again, along with material about line-inputs from the Small Signal design Handbook - which itself is a good book.
I learned a lot from these books but they also have huge gaps in circuit explanations that you have to fill in from other sources. I believe his target audience is not hobbyists but novice audio engineers.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.