Your opinions are sought on Audio Power Amplifier Design: 6th Edition. Douglas Self

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
The Sixth Edition has been out there for about a year now, and i would be very glad to get some feedback on what people think of it. I imagine that everyone who wants it has bought it by now- this not an advert. This is also NOT a trailer for a 7th edition, which is not even planned.

I would specifically be interested in reactions to:

Chapter 7, the detailed analysis of VAS distortion, carried out in collaboration with Samuel Groner.

Chapter 8, which seems to demonstrate that a push-pull or balanced VAS is always inferior to the single-ended VAS of the Blameless power amplifier.

Chapter 12, which shows how very effective output-inclusive compensation is at reducing distortion in a real amplifier, getting it below 0.0001% up to at least 10 kHz at 50W/8R.

Chapter 16,which shows that the real efficiency of a Class-A amplifier with a music signal is no better than 1%, which brings into question whether they should ever be built at all.


I would of course also like to know what you think is missing, but, as I say, there is no prospect of a new edition in the foreseeable future.
 
Disabled Account
Joined 2012
Hi,

I haven't bought the newest edition yet. But, now I will just because of your questions.

It seems that efficiency is always measured at full load to make comparison. But at less than full load, the efficiency differences are are not so great and are more alike. Pls comment.

Thank you for producing an excellent book. -R.Marsh
 
Hi,

Among a lot of informations which were not in the previous editions, I was happy to see a chapter about the Input Stage Common-mode Distortion (pages 141 and 142, figures 6.16 and 6.17, page 143). I would like to discuss about it a bit.

Common mode distortion at low closed loop gain can be solved by a "cascode follower" (as Linsley-Hood called it once) as seen on the circuit suggested by Stochino at the right of Figure 6.16.

However, looking at the harmonic distortion on figure 6.17, we see that the cascode stage is better than the simple stage only above 5 kHz.
I wonder why.

I suppose the reason may rely in the biasing scheme R5-R6-C3 which may significantly alter the impedance at the output of constant current source I1.

Maybe this is why R6 is sometimes replaced by another constant current source.

Regards.
 
Hi,

It seems that efficiency is always measured at full load to make comparison. But at less than full load, the efficiency differences are are not so great and are more alike. Pls comment.

Thank you for producing an excellent book. -R.Marsh

Actually, in APAD6 efficency is examined at all output levels between zero and maximum, for Class-A, Class-B and other modes like Class-G. Class-A efficiency with music is much worse than with sinewaves (see p418) and gets still worse as the operating level decreases from maximum. It really is shockingly inefficient at normal listenng levels, say 15dB below maximum.
 
Member
Joined 2011
Paid Member
T
...
Chapter 12, which shows how very effective output-inclusive compensation is at reducing distortion in a real amplifier, getting it below 0.0001% up to at least 10 kHz at 50W/8R.
...

Dear Douglas

thank you for this really great book!

Can't find "0.0001%" in chapter 12 - maybe you meant "below 0.001% up to at least 10 kHz at 50W/8R"?

BR, Toni

:)) owner of APAD5, APAD6 and some other books from you!).
 
I can see absolutely no way that this could be true. Would you care to propose a mechanism?

As I understand, the noise from the zeners and LEDs masks the low levels of sound and so cannot be heard. I have assembled a shunt regulator for my low TIM Leach amp, and connected it in place of the 2x20 Volt Zeners. Low level highs has improved a lot.

My opinion is that this cannot be simmed as noise component from the zener or LED is not taken into it.

You have to assemble it and hear for yourself.

This is my common sense and understanding, what must be happening.

Gajanan Phadte
 
As I understand, the noise from the zeners and LEDs masks the low levels of sound and so cannot be heard. I have assembled a shunt regulator for my low TIM Leach amp, and connected it in place of the 2x20 Volt Zeners. Low level highs has improved a lot.

My opinion is that this cannot be simmed as noise component from the zener or LED is not taken into it.

You have to assemble it and hear for yourself.

This is my common sense and understanding, what must be happening.

Gajanan Phadte

Hi Gajanan,

Noise level can be easily measured. Did you try to support your theory with some objective evidence? First of all, evidence that LEDs and zeners lead to noise level increase. Second - what noise levels are we talking about? If, for example, using a zener will result in overall s/n ratio increase from 120db to 110db - you will not notice that in a "blind" listening test. Just no way.

Also, to what extent zeners and LEDs influence the overall noise level, very much depends on where exactly they are used in the circuit. In a well-designed circuit there will be no increase.

I mean, just a statement that LEDs and zeners mask the details is simply incorrect (based on my experience). It may be a problem of particular circuit, but this is a different story.

Cheers,
Valery
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.