Class S

[gareth]

Could anyone tell me if the Class S (Sandman) type amps are still in use today ? Are they any good ? And is it worth trying to make one ? If I understand correctly they work in a similar way to the Quad current-dumping technique.

I am looking to build my first dual-mono power amplifier. I have always liked the idea of the current dump and was intrigued by the Class S when I was younger. Truthfully I would like to go down the class A route but they are very inefficient, but good on a winters' evening ! I know that a well designed Class B sounds good, that's what I am using at the moment, but would like to know what the general opinion is out there. Am I going down the wrong path ?

Thanks

Gareth

[ingenieus]

AFAIK the class S was only/mostly used in low powered amps.

Tandberg Audio made this little 15W (25W according to some sources) amp named "Troll" before they folded their tents. I believe it was class S. It is quite rare today.



For building your first amp, it would be best to stick to class B. Less expensive to build and run.

[gareth]

Originally Posted by ingenieus

AFAIK the class S was only used in low powered amps. Tandberg Audio made a 15W (25W according to some sources) one before they folded their tents.



For building your first amp, it would be best to stick to class B. Less expensive to build and run.

Are they that bad ? Didn't Technics have a bash, though not for long ? What about the current dumpers ? Any ideas ? Or stick with Class B ? I have read 'Th eHigh Power Audio Amplifier Construction book' by @forgotten the name' sorry, but he is an American I believe. He mentions all the different output transistor/mosfet configurations to be used and there were three to choose from I think. I fancied a common emitter, again if I remember correctly, and the cascode design because of it's high bandwidth, but after reading about all the pro's and con's he mentioned I was a bit confused as to where to go which is why I wanted to go down the class A route (still would actually) and possibly the current dump.

Yours
Slightly confused

[Nelson Pass]

I think you will find that "Class S" is a trade name for a design, much like
"Class DX", not a category recognized by IEEE.

[gareth]

Originally Posted by Nelson Pass

I think you will find that "Class S" is a trade name for a design, much like
"Class DX", not a category recognized by IEEE.

Thanks Nelson, I thought wrong then. I thought it was actually a type like A, AB, B, D etc. In your opinion do you think I should go down this route ? Initially I wanted to build class A because of it's topology, transistors/MOSFET's being always on and thus no switching distortion as they are biased on and off. Then, having always liked the QUAD current dump I thought of trying one of those.
There is another reason I fancied a Class A and that is because I have no oscilloscope or acccess to one and I thought it may be easier to set up the Class A because of this. Am I thinking along the right road or should I think elsewhere. I would appreciate your honest answer here as you are the well respected Mr. Pass.

[ingenieus]

Class S was was all about efficiency. It is quite similar to Class D. Some say it is exactly the same. Others call it soft switching as opposed to the hard switching of Class D.

[Symon]

Sandman has published a number of designs, the one I believe is usually termed Class S is similar in topology to the QUAD Current Dumper. But operates on different principles.

Audio and Hi-Fi Handbook - Google Books
page 272

this link shows the basic layout, and it's not in any sense class D

[jez]

It's nothing like class d! Closest is the current dumping idea from quad.
Dr Sandman sued Technics over his design in the late 90's... Don't know if he won !?

[martin clark]

Exactly, Dr. Sandman's amp is elegant in a different way - it uses a passive bridge so that a (high power, relatively low-precision) Class B amp drives the majority of the load, and small, ClassA amp provides accuracy via a scaled, complementary output, and the two sum to high accuracy. Another way of looking at it (John Linsley Hood's view I think) is that effectively the big/dumb side drives the output so that the class A side 'see' a very high-impedance load well within its scope. It is not Quad-style 'current-dumping' but a related approach and also elegant.

I believe the principle got appropriated into Technics 'Class AA'

[jez]

Yep.
It basically uses the small accurate amp to make the larger one "see" a high impedance load (so slightly different from what you described) via the bridge network. If you use even a basic push pull emitter follower output stage with enough bias to get round the crossover region then if you are only driving say 1K rather than 8R the distortion is very low.
Yes it was the class "AA" Technics that ripped off his idea. I hope he won...