Audio Power Amplifier Design book- Douglas Self wants your opinions

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

AKN

Member
Joined 2005
Paid Member
not if you're half competent, know what to measure - even a good US$150 sound card could "see" the various capacity PS differing loaded ripple through a LM3886 ~100 dB -PSRR with fc rignt at mains frequency

and that's just if you restrict "measurement" to audio input test signals, look at audio output under load

Hi,
One can measure PSU output under like I did Here, quite revealing.
 
Give a hint for some values.

Say 60VA & 4700uF+4700uF for a 60W amplifier, then 180VA & 15mF+15mF (for 3times the current for the same 60W amplifier and finally a pair of 180W SMPS, or should that be a pair of 30W smps for the dual polarity supply?

Is that what you mean?

Hi Andrew . I think Dvv and I would deliberately make it so that it measured the same yet sounded different . As others will say if carefully measured the difference will appear . My feeling is the measurements would be beyond what ordinary people can do on the money they can afford to spend .That's assuming they have the knowledge to know what to look for . This was especially so when we had the bland attitude of the past that a few chosen measurements told us all we wanted to know . I remember review after review saying yes it is hi fi . Never did anyone say I think this is rather good . In the case of a few budget amps they had room for comment . They were dreadful and that was it . None of those ever get discussed these days . Amstrad IC2000 was a treasure of badness . Now if we worked on one how good could we make it ? Amstrad the guilty amplifier . If I remember Amstrad reviews were in watts per pound rather than any hi fi virtue . Sorry Mr Sugar , you are a Millionaire and some ( me not ) . So what do I know ?

I will give you my take on this . I suspect that class B amps given the PSU for class A work better . This is because we no longer have to use common mode rejection to clean up the signal which would otherwise be ripple modulated . The big illusion is ripple is almost zero at idle . Thus we do not imagine it to be a problem . Turn the bias up and suddenly it is there at idle . Now if the majority of the amp is regulated with the output dumpers given raw DC that will get very close and can use a small PSU . Small PSU's usually have capacitors better suited to HF . Not always big is best . Lots of transformer always works well . Also well designed transformers .
 
Dvv . A challenge for anyone who has bits in the shed . Find a chip amp with a circuit reasonably like the blameless design in the handbook . Build three amps all identical . Three power supplies . One , the minimum required , Next the best you can with lets say 200% spare current . The third a switch mode . I bet they all measure the same ? And the next question is , do they sound the same ?

Nige, you know how I do it.

Assume an amp rated at 50/100 into 8/4 Ohms. Assume energy requirements, 1-2 Joules/10W, to be 2 Joules/10W of dissipated power, meaning an evil load. That means 20 Joules per channel. Assume my supply voltage is say +/- 37V. This means I would need to use roughly 30,000 uF PER SUPPLY LINE PER CHANNEL.

Where else but with me do you even see such numbers?

But I'll bet it would pump out around 200W into 2 Ohms in impulses.

No wonder. I am not happy with just one pair of power tranmnies at the output, so it would have to be two pairs. And since I practically use only Motorola/ON Semi MJL 3281/1302, rated at 200W per device, you could say I have the balls for the job.

As a matter of fact, that's exactly what I HAVE done, the schematic is with the PCB artwork guy, I expect them around July 10 or so. Which means you and I are going to have a very interesting automn, because there are TWO 100WPC projects also with the same guy. I decided to test two different topologies for the hell of it.

Once finished, I'll post it here as well, just to yank the chain of the local topic radicals and naysayers.
 
What I notice is we now have an agreement on things which we can leave to measurements and things we might have fun in using our ears as substitute test gear ( Remember the English and irony here ) . Sometimes the sound of a great amp was the power supply . Why we got deceived was the designer was an obsessive and we though knowing the circuit would save us the expense of paying the costs of retail distribution . Now with that secret known we can do exactly that . LAB 47 is how to do it . The power supply is from another planet . Not technically which is superb . No , it is a work of art .
 
What I notice is we now have an agreement on things which we can leave to measurements and things we might have fun in using our ears as substitute test gear ( Remember the English and irony here ) . Sometimes the sound of a great amp was the power supply . Why we got deceived was the designer was an obsessive and we though knowing the circuit would save us the expense of paying the costs of retail distribution . Now with that secret known we can do exactly that . LAB 47 is how to do it . The power supply is from another planet . Not technically which is superb . No , it is a work of art .

No idea why is Lab47's power supply art - perhaps a picture, or a link?

You and I disagree on several things, Nige, but the one thing we have never ever disagreed over is the importance of the PSU for audio circuits. The only question is how to do it.

One of the tricks I have found to work each and every time is to add a mid sized capacitor (ma favorite size being 2,200 uF) to each and every output device's collector, as near to the device as possible. This always seems to clear the sound up and subjectively make the amp faster, more nimble.

The first time I saw that done in practice was an Italian made Galactron (a hell of a company!) amp sometime in the 70ies. I tried it and liked it.
 
This doesn't do it justice .
6moons audio reviews: 47 Laboratory Model 4706 GainCard

I found out about them when I repaired a FAL speaker they were using . The magnet had slipped and I put it back on centre .

The guy is a neighbour of yours from Bosnia . I lent my pair of FAL ( loaned to me also by FAL via my freind Martina ) to his partner Ishvan . He died and that was by by FAL's . It was the repaired ones . I didn't loose money . I lost an opportunity . My boss loves LAB 47 , he knows nothing of my history with them .
 
Member
Joined 2010
Paid Member
At long last, mine turned up too, after an initial delivery estimate that was even a month more!
After an initial scan of the new chapters, I realised something had changed (and they really are new chapters in most respects). It shows quite a metamorphosis from neatly clipped prose in the early editions to something we might read here in some of the pro. posts - 'not that I mind either style but the contrast amongst mixed chapters perphaps gives the impression of a different author/editor as much as new ideas and material.

The content of these chapters also seems in direct response to some criticisms often voiced here. 'Flattering to those who made them, I guess, so never let it be said that authors are disinterested in reader feedback. The response in this 6th ed. is good but I'm a little surprised that some material is so devoted to defence of previous data, model choice and opinions. This isn't critcism, since there can be cultural and scientific divisions at all levels of even DIYAudio membership, so you could expect different groups to stack up their own evidence in their own preferred order of significance, as we find on the EEs' threads even, so I'm steering clear.

On a basic level, I have to say it's good to see some real ink on the pages. The QC of the last edition (the first by Focal Press) was b..... awful! At least my copy of this Focal edition, though finer print, is readable....which is what I intend to get on with :cool:
 
Member
Joined 2010
Paid Member
Does he discuss TMC by that abbreviation or did he insist on his own acronym?;)
Does he reference Edmond Stuart's work and/or the prior Hitachi patent?....
Hi David
In a word, no. I haven't checked references to other acronyms used here, but under the subject of compensation, he uses previous terminology from articles in Jan Didden's Bookzine of input inclusive and output inclusive. The Hitachi reference is made to identify that push-pull VAS (the term "balanced VAS" has been dropped)

The treatment of the push-pull VAS in chapter 6 and elsewhere BTW, is really given some broader and more interesting coverage. That Doug Self may beg to differ in his commercial perspectives from some academic discussions here, probably makes it all the more so.
'Hope this is of some help - though perhaps DS should be the one to comment on terminology, because I think his primary readership is now non-academic.
 
Last edited:
Ian .
I am dyslexic and couldn't read and write properly when 12 . I was fascinated to read of Alan Blumlein and his problems . In 2004 I got a job which demanded communication and use of the Internet . In honesty I had not written a thing in my life until then .

Going for a blood test yesterday the nurse seemed ill at ease . After some minutes I said yes Jane I remember you . Problem is she hadn't remembered me although her face said it was bothering her . At school I did everyone's maths homework and they lent me their written things to copy up . Jane was excellent at French and me not . I now speak French well enough to read stuff that interests me . Jane gave up . I feel in her half remembered thoughts her brother saying her preferred she didn't talk to me probably came back from 45 years ago . My final statement was " Jane you were a good student and very correct " . That was in front of her work colleges . She seemed genuinely surprised and a little pleased .

The point is I always thought the previous books to be excellent examples of the very best use of English . My girlfriend who is a clever bunny was taken aback when told of Jane as she knows here . She would have never guessed . Way back in time Jane was gifted and should have developed that . Me who struggled so much can not understand throwing away a gift .
 
Hi David
In a word, no. I haven't checked references to other acronyms used here, but under the subject of compensation, he uses previous terminology from articles in Jan Didden's Bookzine of input inclusive and output inclusive. The Hitachi reference is made to identify that push-pull VAS (the term "balanced VAS" has been dropped)

The treatment of the push-pull VAS in chapter 6 and elsewhere BTW, is really given some broader and more interesting coverage. That Doug Self may beg to differ in his commercial perspectives from some academic discussions here, probably makes it all the more so.
'Hope this is of some help - though perhaps DS should be the one to comment on terminology, because I think his primary readership is now non-academic.


Ian , I seem to have gotten adrift with my reply and other posts .


Double VAS . Anyone want to help me with this ? It is just to stop me buying a nasty kit or using Quad 405 or Crown DC 300 .

http://www.diyaudio.com/forums/lounge/200865-sound-quality-vs-measurements-204.html#post3582224
 
The Hitachi reference is made to identify that push-pull VAS (the term "balanced VAS" has been dropped)

I meant did he refer to the prior Hitachi patent for TMC?
In his Linear Audio article on the subject he does not reference it, or Edmond's work.
Bob Cordell properly attributes credit and it is hard to believe Self has not read that comment.
If the prior art is still omitted in the book then it makes a poor impression and would influence my purchase decision.

Best wishes
David
 
Member
Joined 2010
Paid Member
Chapter 21 - The FET output stage, is among the shortest with the last reference from 1995. Perhaps that answers your question, Edmond.

I think Bonsai's presumption regarding CFAs is correct, Christophe.

To correct my ref. above in #1152, Chapter 7 is the VAS stage and the new and interesting material on the push-pull VAS is chapter 8. Together, these make up quite a slab of the text.

To be fair in comment, it's going to take some time to consider and compare so many aspects of audio design here that I don't have the professional background or necessary understanding to interpret on the evidence anyway. As I said, I'm steering clear.

As a DIY'er though, I know that many of us are off on some tangent of faith in ultimately inferior technologies, solely for their alleged superior subjective qualities. This distinguishes efforts as purely for our personal delectation. Many who have read DS's "slim volumes" over 17 years would have realised this is not his technical interest. FWIW, my view is that texts should indeed be restricted to convention so there is a clear fundamental basis of understanding, perhaps followed up by supplements for variations as the need, fashions and advancing technologies dictate.

It is a second, subjective level of interest that is paramount to many here but much of the pseudo-science and hearsay supporting it is highly opinionated and inconsistent. It seems that controversy and bitterness mars any debate there in that mire of personal experiences anyway.

Actually, I'd be interested to know of any author interested or capable of writing a considered text on subjective audio theory and pratice. Wouldn't you? :joker:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.