Class A / AB amps?

Richard Sinkler · From: aka: Rusty Strings -- Missoula, Montana

I see amps referred to as class A and AB. What does this mean and what are the differences?

------------------
Carter D10 9p/9k, NV400, Korg Triton Le88 Synth, Korg CX-3 organ, Yamaha Motif Rack Module, Regal Dobro, Tele, Gretsch Acoustic.

Dave Grafe · From: Hudson River Valley NY

THis is more of a subject for a technical book, Richard, but to I'll try to provide a preamble if not an answer to your question.

There are at this point more classes of amplifer circuits than I can keep up with, but in any case it refers to the electrical method used to amplify the signal.

Class A amps essentially apply the input signal to a single-ended device that increases gain in a linear fashion from zero to a given positive level. Hi-Fi fans are big on this design, it's not always necessarily the flattest but it represents "pure" amplification principles.

Class AB amps are essentially two Class A amplification stages hooked up out of phase in a push-pull configuration and configured to switch off and on alternately, one amplifying the positive half and the other the negative half of the input waveform. Most guitar amps are Class AB, this is why power tubes are most always found in pairs.

There are several other methods of amplification, such as Crown's new class H, which involves digitizing the input and using it to control the output of a high-speed switching power supply, then smothing out all the spikes by running it through a brute-force choke.

Again, these are all crude oversimplifications, but the point is that there are a number of different ways to produce a high voltage or current with reasonable fidelity to a low input signal and each of these basic techiniques is assigned a circuit class.

David Doggett · Posted 25 Jan 2006 8:01 am

Thanks for the technical summary, Dave, but what I think us non-technos need is a summary of the performance characteristics. The current mythology seems to be that class A amps sound better (not sure in what way), but they require a lot of power to run (not sure if that means they use a lot of watts and drive your electric bill up, or they require big transformers and more sound output watts, or both). A class B amp is cheapo to run, but sounds like crap. A class AB amp is a compromise with reasonable efficiency and decent sound. But I have heard technos say this is a sort of myth, like the idea that only medicines that taste bad or sting work, and that class A amps don't necessarily sound much better, and aren't worth the inefficiency. Part of the myth may come from the fact that class A amps are usually lower ouput, and so they get driven harder and produce more tube bloom, etc. So I guess my question is: is the myth true, partly true but exagerated, or mostly bogus?

Dave Mudgett · From: Central Pennsylvania and Gallatin, Tennessee

I think that the idea that class A always sounds "better" depends strictly on what one likes.

In a class A circuit, each tube is working all the time, and there's no switching between tubes. This produces a natural sound, but it pushes everything hard. Since there's no push-pull pair, all distortion harmonics come through, leading to what I call the "chimey" sound. One problem with class A design is that the maximum theoretical efficiency is 50%.

In a "pure" class B circuit, the signal is split using a phase inverter. The positive side of the signal is amplified using one end of a push-pull pair, and the negative side through the other. It can be shown that the maximum theoretical efficiency for sine wave amplification in class B operation is about 78%. In addition, since push-pull pairs work out of phase, certain distortion harmonics are cancelled out. This gives class B amps a different sound.

The difficulty with pure class B is that there is imperfection in the transition region, where the amplification shifts from one side of the push-pull pair to the other. For anyone who has biased an amp, this is referred to as "crossover" distortion. If one was to bias the amp to try to perfectly separate small signals this way, the imperfections would dominate the output. Therefore, in audio amps where this is a problem, amps are biased in a way that small signals are amplified essentially in class A, and as the signal gets larger, the signal is then split and amplified class B. This is somewhat less efficient than pure class B, but still much more efficient than class A, and retains the large-signal sound caused by cancellation of harmonics by the push-pull pair.

As far as which sounds better, that is purely a matter of taste. But they are different sounds. I prefer class AB for lead guitar and traditional pedal steel sounds. I think of it as a harder, more direct sound. But I prefer class A in certain situations to get a more textured sound, as in jangly rhythm guitar for playing folk-rocky tunes like the Byrds. For more distorted slide or steel sounds, I can't decide which I like better - I guess I like both in certain ways. But that's just my take. I guess if I had to just have one, it would be AB. I find I can get the jangle I need by fattening up the midrange and bottom end a bit.

Dave Grafe · From: Hudson River Valley NY

Very nicely laid out, Dave M.

David Doggett · Posted 25 Jan 2006 12:04 pm

Thanks, Dave M and Dave G. That's the best explanation I've ever heard.

Dave D.

------------------
Student of the Steel: Zum uni, Fender tube amps, squareneck and roundneck resos, tenor sax, keyboards

Joseph Meditz · From: Sierra Vista, AZ

Nelson Pass said in his Zen Amp article, "...The characteristic of a single-ended gain stage is the most musically natural. Its asymmetry is similar to the compression/rarefaction characteristic of air, where for a given displacement slightly higher pressure is observed on a positive(compression) than on a negative (rarefaction). Air itself is observed to be a single-ended medium, where the pressure can become very high, but never go below 0. The harmonic distortion of such a medium is second harmonic, the least offensive variety."

J Fletcher · From: London,Ont,Canada

My understanding is that Class A refers to any device which amplifies all 360 degrees of the input signal, so a single ended amplifier will almost always be Class A, and a push pull amp can also be class A. The biasing determines whether or not it is Class A or AB, not the fact that it's single ended, or push pull.
AB is when the device, tube or transistor, amplifies the input signal more than 50% of the time, but less than 100% of the time. So Class AB has quite a range from fairly low idle current, to very high idle current, but not as high as Class A.
I know that doesn't tell anybody how it sounds though.....Jerry

Jim Sliff · From: Lawndale California, USA

Here's a non-technical, incorrect explanation...but it IS what most people mean in the guitar world:

Class A: Vox AC30's, Matchless, Fender Champs. Either EL84-based or single power tube. Run hot, break up with a certain midrange emphasis. Lower wattage, not as efficient, but loud for the rated output.

Class A/B: Blackface and Silverface Fenders, Marshalls. 6L6 and EL34 power tubes. Clean-to-crunchy with more of a scooped-mid sound. More efficient, higher power capabilites.

_____________________________________

That being said, most of the "Class A" amps are NOT Class A. The Champ is.

But you will find amp techs who will swear than ANY EL84-based amp is class A. So the Vox/Matchless sound has kind of become the terminology benchmark for Class A tone, even though it's technically wrong.

Keith Hilton · From: 248 Laurel Road Ozark, Missouri 65721

Do a search on past Forum topics. I wrote some information about amp classes, even explaining about push-pull amps.

Dave Grafe · From: Hudson River Valley NY

Jim S. that was good, too!

Dave Mudgett · From: Central Pennsylvania and Gallatin, Tennessee

Jim, you're right - many guitar players hold this fallacious view. Amplifier class has nothing to do with what nonlinear device is being used. Any tubes, or as Jerry says, transistors, can be set up to run class A, AB, B, or C. In class C, the tube is "on" for only a fraction of each cycle, and is totally irrelevant for high-fidelity amplifiers.

But as long as we're going here, a further technical subtlety is that there are two ways to set up a class A amp. The first and most obvious way is to simply run one or more parallel versions of identical single-ended amps. This leads to the most raw-sounding amp possible, since there is absolutely no cancellation of even distortion harmonics. Each tube must amplify the full signal because there's no other way for it to work. There's no phase inverter, or center-tapped output transformer to put the two inverted signals back together. Think of a bunch of Tweed Champ amps in parallel. This is what I think of when I really think of "Class A".

The second way is to run push-pull pairs, but bias it so that the zero-point between the positive and negative cycles of the signal is set near the middle of the linear range of the tube. In this case, both tubes are on all the time, hence operation is class A. There may be some harmonics cancellation due to push-pull operation, but the effect is quite different than the class B or AB type operation.

To further confuse things, there are different types of class AB operation, which gets into the issues of fixed vs. cathode bias, for example. With cathode bias in what is called class AB1 operation, the amp is run closer to class A operation. I think of the Tweed Deluxe as the classic example of Class AB1. On the other hand, using a fixed bias in what is called class AB2 operation, the amp is run closer to class B, with the resulting improvement in efficiency and clean operation. I think of the blackface and silverface Fenders like the Twin Reverb as the classic examples of AB2 operation. [Correction: BF/SF Fenders run class AB1 - in other words, no appreciable grid current flows. The bias can be set to run very close to class B. There are a few amps that claim to run in class AB2.]

One way to get closer to class A operation in a class AB amp is to set the bias hotter, i.e., a snaller negative bias voltage. The amp runs hotter and gets more harmonic distortion from running more quickly into the upper saturation limit. For country, rockabilly, and jazz guitar or pedal steel, I set the bias pretty cold, as close to class B as possible without getting a lot of crossover distortion, because I want max clean headroom, but still get tube tone.

My first explanation was intended to be a simplified version of the main differences - and not even this addition is the full story. For that, I recommend the "Radiotron Designer's Handbook", edited by F. Langford-Smith. Mine is a 1965 reprint of the 1953 4th edition. I believe it's been reissued recently. Not for the faint of math, but very illuminating if you're not afraid of technical things. Here are some links to fairly inexpensive CD-ROM versions of this book:
http://www.audioxpress.com/bksprods/books/cdrdh.htm
http://www.marshallschematics.com/rdh4.html

It's not hard to find print versions, but they're pretty expensive.

[This message was edited by Dave Mudgett on 27 January 2006 at 05:52 PM.]

John McGann · From: Boston, Massachusetts, USA * R.I.P.

I have a Matchless and several vintage Fenders.

In non-tech terms,the Matchless (class A) in contrast to the Fenders seems to "speak faster"- not as spongy feeling as the Deluxe reverb. The notes articulate faster and sharper, yet the amp has a very warm hi-fi quality that is different from the Fenders, which are generally softer and "spongier". They are all great in different ways. The difference matters, because I feel that with electric instruments, you "play the amp" as well as playing the instrument. YMMV blah blah...

------------------
http://www.johnmcgann.com
Info for musicians, transcribers, technique tips and fun stuff. Joaquin Murphey transcription book, Rhythm Tuneup DVD and more...

[This message was edited by John McGann on 26 January 2006 at 08:34 AM.]

John Daugherty · From: Rolla, Missouri, USA

Class A amplifier circuits are commonly used as preamps. Class AB is commonly used for the power amp output circuit.

------------------
www.phelpscountychoppers.com/steelguitar

Ray Minich · From: Bradford, Pa. Frozen Tundra

For a good reference for the configuration, layout, comparison and contrast of the different tube based species, look at any issue of the RCA Receiving Tube Manual.

[This message was edited by Ray Minich on 26 January 2006 at 09:06 AM.]

Dave Grafe · From: Hudson River Valley NY

Mr. Mudgett you continue to entertain and amaze, my hat is off to you for a concise and insightful explanation of this arcane subject, with a bibliography and everything.

[This message was edited by Dave Grafe on 26 January 2006 at 05:36 PM.]

Richard Sinkler · From: aka: Rusty Strings -- Missoula, Montana

Thanks to all. While some of the tech stuff is a little over my head (I do have some knowledge of electronics), these answers have helped me understand what the difference means to the musician.

------------------
Carter D10 9p/9k, NV400, Korg Triton Le88 Synth, Korg CX-3 organ, Yamaha Motif Rack Module, Regal Dobro, Tele, Gretsch Acoustic.

J Fletcher · From: London,Ont,Canada

If you are interested in understanding tubes and tube amplifiers, as mentioned , the RCA Recieving Tube Manuals are indispensible.
According to my manual, AB1 refers to a class AB amp that draws no grid current, and an AB2 amp draws grid current. Doesn't matter if you bias hot or cold, fixed bias or cathode bias, it can still be AB1. As far as I know, all Fender tube amps that are push pull, run class AB1. The driver tube doesn't supply current, only a signal voltage to the output tube's grid.
I don't know of any amps that run AB2, though I have seen hifi designs that use a driver transformer, and this would be a method of supplying current to the output tubes grid circuit. The AB2 thing is a bit of a mystery to me...Jerry

Dave Mudgett · From: Central Pennsylvania and Gallatin, Tennessee

I stand corrected on the Fender Twin Reverb running class AB2. Jerry is correct - AB2 draws appreciable grid current, and BF and SF Fender amps don't. They get their additional power by goosing up the plate voltage way beyond the rated plate voltage for the 6L6 power tube, or in the case of the very late Twin/Pro Reverbs, using an ultralinear design with special ultralinear output transformer. I have added a correction to my second post.

One point: AB1 may be cathode- or grid-biased, but tend to run closer to class A when cathode-biased. When grid biased, that point can be moved around. On the other hand, AB2 must always be grid-biased.

There are some amps that claim to run in class AB2. Langford-Smith uses the McIntosh amplifier, with bifilar driver/output transformer, as an example of an amp which can be run either AB1 or AB2. This is strictly a hi-fi amp.

For AB2 guitar amps, the Guytron GT-100 guitar amp claims to run in class AB2 - see http://guytron.com/amp_set.shtml - I can't find a circuit diagram on this, so it's not clear how it works.

The mid-90s Fender Prosonic also claims to run class AB2. That claim is in the specs in the Reference and Owner's Manual. A switch on the back of the amp switches between Tube Rectifier - Class A, Tube Rectifier - Class AB2, and Solid-State Rectifier - Class AB2. The phase inverter in the circuit diagram looks like a pretty standard 12AX7 driver circuit, but with three distinct taps in the grid bias circuit. It also looks like the 12AX7 is run hotter than usual, but it's still not clear how this supplies significant grid current. Running the plates at 465 volts, it is rated at 60 watts for 2-6L6 power tubes, which is higher than usual for non-ultralinear operation.

I tried a Prosonic when they first came out. On the clean channel, it definitely had an extreme hi-fi sound, too much so for me.

These amp classes are pretty broad, and I think are intended as a rough guide to how the amp operates. There is especially some ambiguity in the class A designation. But the RCA Receiving Tube Manual and the Radiotron Designer's Manual are the two authoritative guides, IMO. There is no inconsistency in their definitions - the error on the BF/SF Fender designation was strictly mine.

Jon Light (deceased) · From: Saugerties, NY

quote:
Output Stage
• Consists of two seperate 20W power amplifiers running in parallel
• Each 20W power amplifier has its own pair of 6V6 tubes and its own output transformer, and drives its' own 12" speaker.
• Cathode biased operation
• Class-A mode operation up to point of overdrive; Class-AB operation in overdrive.

This is part of the description of my Tone King Comet. Granted, I struggle to understand much if the discussion in this thread but I find the Class A, AB description here baffling. Does this have meaning for anyone here?

Dave Mudgett · From: Central Pennsylvania and Gallatin, Tennessee

I interpret this to say that the amp is cathode-biased class AB1, and the bias circuit is set so that it stays in class A mode until the onset of distortion. In class AB, the small signal amplification is class A - both tubes work on positive and negative signals. As the signal gets larger, there is a transition to class B type amplification, where the signal splits into positive and negative components and each side of the push-pull pair acts separately.

What's different about this amp is that there are two separate parallel push-pull pairs, each with its own center-tapped output transformer which drives a separate speaker. I like that idea a lot - it would be, in principle, possible to set the bias differently, to allow one side to stay cleaner longer than the other. I generally prefer to play with two amps - one which pushes the tubes a bit, and another which stays clean for better definition.

I had a very early 20W Tone King Imperial - a friend and I have traded it back and forth for over a decade now. It has perhaps the finest clean sound I've ever heard in any amp. The gorgeous sea-foam green and white two-tone naugahyde cover was getting beat up when I took it to gigs, so I stopped taking it out and my friend has it back now. Mark Bartel definitely knows what he is doing, IMO.

Forum Index > Electronics		Next oldest topic :: Next newest topic