Power Amplifier

[horias2000]

I just built a power amplifier module based on the Trimodal Power Amplifier by Douglas Self. This is part of building a complete integrated amplifier. I'm now working on the preamplifier, power supply, PHONO preamp, input switching, speaker protection and maybe remote control unit.
For those not familiar with the trimodal concept by D. Self, this is a power amplifier that can be switched between two modes of operation. Class A operation with a fixed quiescent current (1A in my case) and a class AB/B mode with a much lower quiescent current of 50mA. The amplifier was build after reading his books (highly recommended). I also measured the THD with my QuantAsylum QA401 analyzer and have to say that the class A performance is very good as it can be seen in the images below. In class B I was a bit baffled by the high-ish 2nd order harmonics. It is still good performance by my standards but there might be something with my prototypes. This is still work in progress. The modules I've built have double output devices and the measurements were performed with an 8OHM load at an output of 5W. With 1A quiescent current in class A it can deliver about 12W before going into class AB. In class B it can deliver around 80W of output power with the +/-28Vdc rails. The measurements you see below were performed with a lab power supply and not on the actual power supply I will be using. I will come back with the same measurements performed with the actual unregulated power supply that will be implemented in the final amplifier. I'm expecting some mains frequency components (and their harmonics) but not too much as the design has a very good PSRR.

 

[horias2000]

so -- what's the third mode? Or is the name deliberately ironic?

Well you have Class A mode, Class AB and Class B. So there are three modes but it switches from one to the other on its own.

 

[mhardy6647]

Well you have Class A mode, Class AB and Class B. So there are three modes but it switches from one to the other on its own.

I retracted the question, because it was too snarky. Sorry.
Class B, eh? Seems odd. Pure class B would have pretty substantial crossover distortion (I would think).

 

[horias2000]

I retracted the question, because it was too snarky. Sorry.
Class B, eh? Seems odd. Pure class B would have pretty substantial crossover distortion (I would think).

Well, a well designed class B amplifier is better than an class AB one. The Power Amplifier Design Book by D. Self goes into this at great lenghts (he calls the class B amplifier "Blameless"). Most of the amplifiers you can buy, work in class B. The philosophy of this design is that if you optimize quiescent conditions (and other mechanisms of distortions), you can get very good class B performance (not as good as class A). And as it can be seen from the measurements, it's true. I still have to look into the 2nd order harmonics but even with these, the performance is very good.

 

[mhardy6647]

Count me as skeptical. Best of luck with your build, though, and sorting out your HD problem.

 

[horias2000]

Count me as skeptical. Best of luck with your build, though, and sorting out your HD problem.

It might sound like I'm trying to sell the book but I can't overstate how good it is. I'm an electronics design engineer and I read a few books related to this subject and this one is by far the best (in my opinion). If you have the time, do give it a read. I will come back to this thread once I finish the preamplifier and the PHONO preamp. The preamp itself is a bit different as it also offers the possibility to adjust the cutoff frequency of the LF and HF sections, not only the cut/boot of LF and HF.

 

[gene_stl]

Doug Self had a great preamp project in Wireless World. I think it was in the late eighties or early nineties. I made sure to get all the magazine issues (at a local news stand, I did not subscribe at that time) because it looked like a great project. He had someone making and selling circuit boards for it. I remember they were a little pricey for blank boards, but I still was very interested. I never quite overcame my intertia, on the project though because I already had a very good preamp. But ever since then I have been a Doug Self fan. He appears in a few youhoo tooob
vids.

I have tried with no luck to find the construction articles. I am pretty sure I didn't hallucinate the whole thing.

 

[horias2000]

Doug Self had a great preamp project in Wireless World. I think it was in the late eighties or early nineties. I made sure to get all the magazine issues (at a local news stand, I did not subscribe at that time) because it looked like a great project. He had someone making and selling circuit boards for it. I remember they were a little pricey for blank boards, but I still was very interested. I never quite overcame my intertia, on the project though because I already had a very good preamp. But ever since then I have been a Doug Self fan. He appears in a few youhoo tooob
vids.

I have tried with no luck to find the construction articles. I am pretty sure I didn't hallucinate the whole thing.

Yep, that is the preamp. It's called Precision Preamplifier and it evolved a bit in the meantime. You can still buy pcbs from the signal transfer company and he wrote an article about it in linear audio about an improved version of it. I made pcbs of this preamp and I'm going to test it in the coming days. Mine does not include the phono pream as this will be on a separate board but it is the updated version from the linear audio magazine.

 

[MakeMineVinyl]

Count me as skeptical. Best of luck with your build, though, and sorting out your HD problem.

In an amplifier that wants to avoid crossover distortion, there has to be some overlap at the zero crossing.

 

[mhardy6647]

In an amplifier that wants to avoid crossover distortion, there has to be some overlap at the zero crossing.

That's what I am thinking; certainly the dogma that I've internalized.
I've heard of Douglas Self, but not read his book(s).
I note that class B amplifiers have never taken the audio world by storm* so either there's something else going on here or Self's IP position is tight enough, and his licensing requirements pricey enough, that the industry has (mostly?!) worked around it and stayed (more or less) Class AB -- at least until the fairly recent prolifieration of PWM amplifiers.
Not clear to me (!) whether any of this has found its way into "commercial" hifi.
I did see this thread, but I don't know and I haven't dug too far.

Douglas Self's Blameless amplifier

I'm in the middle of reading Douglas Self's Audio Power Amplifier Design book and he mentions that at the time of writing thousands of amps had already been manufactured using the principles outlined in his book. I'm intrigued. Does anyone know which commercial amps he designed based on the...

audiokarma.org

____________
* Which is not to say that there have been no Class B audio amplifiers commercialized!

 

[MakeMineVinyl]

That's what I am thinking; certainly the dogma that I've internalized.
I've heard of Douglas Self, but not read his book(s).
I note that class B amplifiers have never taken the audio world by storm* so either there's something else going on here or Self's IP position is tight enough, and his licensing requirements pricey enough, that the industry has (mostly?!) worked around it and stayed (more or less) Class AB -- at least until the fairly recent prolifieration of PWM amplifiers.
Not clear to me (!) whether any of this has found its way into "commercial" hifi.
I did see this thread, but I don't know and I haven't dug too far.

Douglas Self's Blameless amplifier

I'm in the middle of reading Douglas Self's Audio Power Amplifier Design book and he mentions that at the time of writing thousands of amps had already been manufactured using the principles outlined in his book. I'm intrigued. Does anyone know which commercial amps he designed based on the...

audiokarma.org

____________
* Which is not to say that there have been no Class B audio amplifiers commercialized!

I might be wrong, but I believe the vintage McIntosh MC275/240 amplifiers ran in class B and they got away with it because of the elaborate output transformer. Anyone know for sure?

I just ran across this on my hard drive - this is the distortion of my MC240 at its rated 40 watts:

 

[mhardy6647]

I might be wrong, but I believe the vintage McIntosh MC275/240 amplifiers ran in class B and they got away with it because of the elaborate output transformer. Anyone know for sure?

I thought so, too -- but a quick googlin' hasn't confirmed that, though. Plus, as you say, they do use a unique OPT design.
I am sure someone here like @SIY (e.g.) knows the actual answer, as opposed to a googlin's-worth!


I did do a... ahem... Self-search (not literally!) here and tuned up a coupla interesting tidbits.

Search results for query: doug self

www.audiosciencereview.com

e.g., this:

Tube Amp with the "Good Kind of Distortion"

I just have Matched Pair (2x) Genalex Gold Lion 12AX7 tube (ECC83) come three hours ago just put them in and listen almost 2 hours, it is okie, but sound not clear like stock tubes, a little bit bright not warm like stock, with my ears these Genalex tube sound not good like stock McIntosh...

www.audiosciencereview.com

 

[audio2design]

That's what I am thinking; certainly the dogma that I've internalized.
I've heard of Douglas Self, but not read his book(s).
I note that class B amplifiers have never taken the audio world by storm* so either there's something else going on here or Self's IP position is tight enough, and his licensing requirements pricey enough, that the industry has (mostly?!) worked around it and stayed (more or less) Class AB -- at least until the fairly recent prolifieration of PWM amplifiers.
Not clear to me (!) whether any of this has found its way into "commercial" hifi.
I did see this thread, but I don't know and I haven't dug too far.

Douglas Self's Blameless amplifier

I'm in the middle of reading Douglas Self's Audio Power Amplifier Design book and he mentions that at the time of writing thousands of amps had already been manufactured using the principles outlined in his book. I'm intrigued. Does anyone know which commercial amps he designed based on the...

audiokarma.org

____________
* Which is not to say that there have been no Class B audio amplifiers commercialized!

I liked the last post in that linked thread, "Not judging a damned thing just saying a hell of a lot talented people out there who didn't bother or want to write a book."

 

[mhardy6647]

I liked the last post in that linked thread, "Not judging a damned thing just saying a hell of a lot talented people out there who didn't bother or want to write a book."

Which could put one in the realm of the 'trade secret'.
The problem with which, of course, is that one's competition might stumble upon one's trade secret.
Pretty uncommon for one individual to have a lock on all of the insight & intelligence in the world vis-a-vis any particular "problem"/concept/issue.

Or, as it was so pithily stated in Shift Happens by Thomas Friedman in The World is Flat:

“The saying in China is "If you're one in a million, there's still 1,300 people just like you.”

 

[SIY]

I thought so, too -- but a quick googlin' hasn't confirmed that, though. Plus, as you say, they do use a unique OPT design.
I am sure someone here like @SIY (e.g.) knows the actual answer, as opposed to a googlin's-worth!


I did do a... ahem... Self-search (not literally!) here and tuned up a coupla interesting tidbits.

Search results for query: doug self

www.audiosciencereview.com

e.g., this:

Tube Amp with the "Good Kind of Distortion"

I just have Matched Pair (2x) Genalex Gold Lion 12AX7 tube (ECC83) come three hours ago just put them in and listen almost 2 hours, it is okie, but sound not clear like stock tubes, a little bit bright not warm like stock, with my ears these Genalex tube sound not good like stock McIntosh...

www.audiosciencereview.com

Not quite Class B but close. Their unity-coupled circuit relied on a fancy multifilar output transformer and was remarkably clever. I’m sure there’s a copy somewhere on the interwebs of Norman Crowhurst’s “Realistic Audio Engineering Philosophy,” which examines it in detail.

 

[horias2000]

Well all I can tell you is that I worked on a few amplifiers so far and most of them if not all were in class B. Almost all the Japanese ones had 5mA or 10mA quiescent current through the output devices. That is as calss B as it can get. Very low quicent current. I did not have a spectrum analyzer back then to look at the thd but I suspect it wasn't great.

 

[bigguyca]

I retracted the question, because it was too snarky. Sorry.
Class B, eh? Seems odd. Pure class B would have pretty substantial crossover distortion (I would think).

Have you read Douglas Self's power amplifier book (currently in the 6th Edition)? From your comment I would guess not. Here is a link from Amazon. It's a great book if you want to learn about power amplifiers.

Audio Power Amplifier Design, Sixth Edition: Self, Douglas + Free Shipping

Audio Power Amplifier Design, Sixth Edition [Self, Douglas] on Amazon.com. *FREE* shipping on qualifying offers. Audio Power Amplifier Design, Sixth Edition

www.amazon.com

You need to understand that what is typically called class A/B on this and most other forums is called class B by Self. Class B in this context means properly biased, or optimum biased. That is, the bias level that minimizes crossover distortion. Self goes into extensive detail on the factors affecting crossover distortion by theory, simulation and measurements.

What is called class A/B by Self would be called something like class A/B biased heavily into class A by most people. As an example, many Parasound amplifiers are biased heavily into class A/B. Self notes that such over biasing adds distortion at the point where the amplifier shifts from class A to class B. Bob Cordell calls this effect transconductance (gm) doubling. Such over-biased amplifiers also typically use a lot of power at idle due to the partial class A biasing that is very energy inefficient.

Bob Cordell's book is also excellent. Reading both books provides really excellent coverage of power amplifier characteristics and design.

Designing Audio Power Amplifiers: Cordell, Bob + Free Shipping

Designing Audio Power Amplifiers [Cordell, Bob] on Amazon.com. *FREE* shipping on qualifying offers. Designing Audio Power Amplifiers

www.amazon.com

 

[audio2design]

You need to understand that what is typically called class A/B on this and most other forums is called class B by Self.

Other forums, every text book ever written, etc.

 

[horias2000]

In theory, class B is when you have no quiescent current going through the output devices. But we have to acknowledge that if you have 5mA or 10mA of quiescent current, that amplifier will run most of the time (99.99%) in class B when you listen to music at a normal level. So yes, in theory this is not textbook class B but during normal operation it works in class B.

 

[gene_stl]

Norman Crowhurst
The Great Man who set me on my multi amping course. Changed my life. In 1969.
See "Radio-Electronics" March and October 1969. Easily available online free download.

https://worldradiohistory.com/Archive-Radio-Electronics/60s/1969/Radio-Electronics-1969-03.pdf

https://worldradiohistory.com/Archive-Radio-Electronics/60s/1969/Radio-Electronics-1969-10.pdf

It was a great magazine with lots of great construction articles.