Class A is biased so that the wave form does not pass through the 0 DC point (thus switching). Pure Class A means that it NEVER passes through, thus the bias is pretty high (particularly for a high voltage amplifier). Running in this mode is very inefficient, but it supposed to yield the best sound quality. Class B is a switching mode which goes through the 0 volt DC point. The switching causes some small level of distortion. The advantage of B is that it runs much cooler and is generally less expensive to build. Many amps use Class A/B which means they run in Class A for the first watt or two and then go into a switching mode (B) operation. This can be very advantageous as the switching is unlikely to degrade the sound quality appreciably once the amplifier is operating at higher voltages (the switching distortion is so small compared to the overall wattage at that point it is almost negligible). As with most things in the Audiophile community the debate as to whether or not A/B can sound as good as pure A is still out. The other class is class D, which is digital amplification. It operates with discrete pulses (or bits), thousands of times a second to generate the power required. Class D has typically been used only for subwoofers, where the pulse is so short compared to the wavelengths being produced the issue of digital vs analog was not a concern (or much of one). However, more recently, companies are producing digital amplification for the full range. Hope that helps.
This is an area that is a little hard to explain if you do not have a little background. It deals with the amount of time that the tube (I tend to talk tubes) conducts when an input signal is applied. That is the only issue involved.
For class A, one could say that anode current is set at a level such that, even with the maximum allowable input signal, anode current never falls to zero. That is, the valve never turns off.
In class B the the anode current only flows during the positive half cycle of the input waveform.
Class AB the tube conducts more that half (180degrees) but less that all (360degrees) of the waveform.
This is less than helpful for some people. So what! you might ask, and rightfully so. If the above sounds like gibberish the best online explanation (for tubes) of it that I know of is here:
You might have to read a dozen of so pages to get enough background to make sense of it.
Above is the technical definition of class operation for an amp. There are endless debates about the advantages of each for producing music.
Here's a "generic" and very simplified way to look at an amplifiers bias. Think of an amplifier as a rotating wheel.
Class A drives the wheel at all points of the rotation. This means that the wheel is always being directly driven and monitored. As such, it should offer the smoothest, most even rotation possible while always being under control. This means less distortion or irregularities in the rotation ( output ) of the wheel.
Class B drives the wheel for only the first half of the rotation and then is left to rotate on its' own. Due to the natural forward momentum of the wheel, the wheel can be left unattended for the second half of the cycle and remain relatively consistent. This is called "the flywheel effect" as it keeps rotating ( producing output ) regardless of whether or not something is energizing it. As the wheel ( output of the amp ) comes around to the beginning of the cycle, drive is re-applied again and then disengaged at a preset point. The benefit to all of this is that the wheel is only driven 50% of the time, so a lot less work and effort is put forth to keep it running on a relatively smooth basis.
The major problem with such a scheme is that there will be a point of "cogging" or "irregular motion" at the time that the wheel is re-engaged to direct drive or when it is released to "free wheel" for the second half of each rotation. Those slight "glitches" would be considered "crossover distortion". This can be somewhat minimized by use of various types of feedback, but that in itself is a controversial subject and a whole 'nuther ball of wax.
Class AB tries to combine the best of both worlds. The amp ( or "wheel" ) is always turned on and fully engaged ( Class A ) for the entire duty cycle of each signal for low level signals. That is, it operates in this fashion up to a given point. Above that point though, the amp switches over to Class B bias. This helps reduce the need for massive heatsinking and increased output devices to sustain such high output levels while increasing efficiency on the amp. From a manufacturer's point of view, it is "better" because it is both cheaper to produce and slightly more reliable due to less thermal problems.
In an AB amp, most of the signal would be reproduced while the amp is in Class A mode. This means that reproduction "should be" as good as possible in terms of linearity and low distortion. The exceptions to this would be if listening with a slightly higher than average spl, using in-efficient speakers or playing music at an average volume that has extremely dynamic peaks involved. All of these conditions require the amp to work harder and put out more power, pushing the amp closer to or well into the Class B range of operation.
Most "hi end" amps are biased to run in Class A for up to several watts, not just the first one or two ( or even just PART of one watt ) that is commonly found on most mass produced gear. Some will go as high as 30 - 50 watts per channel while still operating in Class A mode. My findings are that many of these amps run Class A up to appr 8 - 10 wpc and then switch over to Class B. This is just a generalization though, as there are always exceptions to "the rule". Obviously, the higher the bias, the "purer" the sound for most operating conditions. This also means more heat dissipation to deal with though and more power supply capacity being required. Since metal for larger chassis and heatsinking and higher power transformers and greater quantities of filter caps are not cheap, the manufacturer will typically compromise somewhere along the line. These are products that may work quite well under normal operation, but are built to a price point.
There are quite a few other classes of operation outside of A, B or AB. While Class C is not normally used in audio applications, there are amps that do make use of Class D, Class G, Class H, Class T, etc... Most of these are of high efficiency design, primarily trying to do the most in terms of output with the least amount of lost energy ( heat ) resulting in high efficiency. The different classes are simply variations of the same basic principle ( to a great extent ). All of these use some type of switched design, literally pulsing the "drive" to the "wheel" many times a cycle. Some even switch in auxillary amplification circuits as the earlier stages become overloaded. This can result in a LOT of "glitches" or "crossover distortion" due to the irregular or "twitchy" type of "steering" that is being applied at random intervals to the wheel.
Having said that, much has been learned in the last few years about amplifier technology, hence the arrival of "digital" amps, etc... Some of these are said to rival amplifiers of Class A or very "richly biased" Class AB ( stays in Class A for a longer time than most AB amps ) while offering drastically increased efficiency. While i have heard and actually own a few of these amps, my personal opinion and findings show that well thought out higher bias designs still have an advantage. Most of the advantages are in the areas of bass attack ( slam ), overall dynamics and spacial characteristics ( air, depth, separation of instruments and notes, blackness of background, etc ). Obviously, I can't lay claim to having tested or experienced everything that is currently available though. Others may have different experiences with various ( same or different ) products in different systems. That is why most of us stress the importance of in-home auditions. What you and i find "acceptable" or most / least importance may be quite different overall. Sean
Nice analogy Sean. But is Class B exactly as you describe? I was under the impression that the "drive" is switching over to a second set of transistors or tubes, during the "second" half (or negative swing) of the signal. Or are you not equating the two "halves" with polarity of the signal?
Thanks for the responses. I have a little bit of a back ground in electricity. So, I believe I have a pretty good grasp of what you're saying.
Audiofile9, what you describe is known as a push-pull arrangement in terms of one output device "handing off" to another device as the signal changes polarity. These types of devices, especially SS models, are typically sold or marketed as "complimentary pairs", etc.. This somewhat assures near identical operating conditions for each polarity of the cycle. This is very similar to the way that running matched pairs in parallel assures that they share the load equally. Sean
Sean, your the man. I've read numerous descriptions on the differences of "class", but yours is the most lucid and succint yet. Thanks.
Unsound, right on! Sean, thanks for the clarification. If you would be so generous, perhaps you could explain the difference between "Single Ended" and "Class A" (and how "push pull" relates)? I've always been fuzzy about the differences there. Perhaps like many others, I assumed Class B was synonomous with PushPull, etc. Have seen the many threads that touch upon these subjects, but have not found any bearing your excellent analogy as above.
Audiofile9 makes a very good point (request). I once recall someone telling me that while you may understand something, the true test is being able to explain it in simple terms to others, so that they understand it. That said, I felt I understood class B vs push pull, but clearly Sean's articulate analogy shows he REALLY understands this concept. And while I also feel I understand "single ended", "triode", "pentode", and "OTL" I don't think I could do the explaination justice--but hope that Sean will.
All "single ended" amplifiers are by definition Class A,
if they are high-fidelity amplifiers.
All that single ended means is that the amplifying element
(tube or transistor) is alone, and not in push-pull. So,
this means that it is starting out, when doing *nothing*
at all "halfway" on. In other words, biased Class A.
By being halfway on, it can go all the way ON,and all the
way off, this being the AC signal that it can produce. If you are sitting there thinking, 'hey, if it's on halfway all the time, that's wasting a lot of power' then you've got it right. Class A single ended amps tend to waste about 75% of
their power as heat, not audio output.
If there is an advantage to Class A amps of any type it is that at low levels the inherent linearity of the output stage is better than a class AB or class B which has some non-linearity happening as one set of output devices switches all the way *off* as the signal (AC) swings through "zero" going from + to - and back to + again (through zero twice there).
The linearity issue is small, as is the deviation from linear gain - you can think of it as a dip or peak in the frequency response of a crossover, not summing quite right. But of course, this is not frequency dependent. It is however usually *bias* dependent. Since by definition all distortion is deviation from ideal gain, non-linearity in the "crossover" region is measured as distortion.
Doug Self has an excellent discussion of this on his website and in his book, although I may not agree with all of his conclusions on the subject, his technical work is impeccable. He has a "blameless" Class B design that measures better than most Class AB designs...
But that's not single ended. Single-ended is what is in most
tube preamps, the classic cascade circuit. A series of simple class A stages one run into the next.
Single ended class A tubes require a transformer in place of the resistor used in preamps at the plate, or else a plate resistor or choke and a cap blocking the DC to a transformer (a popular idea these days). In any event the output is created by sinking current through the tube, causing a voltage drop in the resistor, or through the transformer (more or less).
Push-pull is in effect replacing the resistor with an active device that amplifies the signal but *inverted* in polarity - so that when one side is ON, the other is OFF.
Two guys on opposite ends of a big old saw - one pushes, the other pulls (in phase output, out of phase input)! That's
Single ended - one guy on one side of the saw - pushes and
pulls all the time.
Single ended, parallel- two guys on the same side of the saw.
Push-pull class B - One guy pulls, then the other guy pulls - or you can think of it as one guy pushes and pulls *exactly* 1/2 way,
then the other guy completes the cycle, pulling and pushing back 1/2 way (a more accurate description...).
Class AB - as with class B, but in the middle, their efforts
overlap a little bit.
Class H? guys with jack hammers, pound the damn thing from one side to the other with monster pulses... or is that class G... I forget.
Sorry i missed your questions folks. Luckily, Bear stepped in and helped us out. He is far more capable of explaining most technical things than i am. Sean
Bear, great job! Thanks. Sean don't underestimate your self.
Oh, I meant class D, not G or H... I think.
I think Class H is that switched rail thing, like Carver?
And D is switchmode.
That leaves G... hmmmm...
Guess I have to get a newer text book, eh?
Maybe it's time to get those straight in my mind.
Oh yeah - more on "linearity"
If you look at the transfer curve of any device, they tend to be most linear (straight line) in the middle of the curve. That means that for a given input increase, the output changes the same amount for each increase/decrease.
(ie. "linear" - a 1:1 relationship being maintained)
At the ends of the curves (all the way on, and all the way off) the relationship is not so good. In fact the transfer curve looks like an "S."
What this means is that when almost off - at the bottom of the curve the device is not very linear - you need much more drive to get not so much output, UNTIL you get past the curve of the "S" and it gets linear.
So, two things are important here: 1) for class AB and B circuits, you're almost off (near "cut-off") so the curve to begin with is not linear. 2) Class A is in the middle of the most linear portion of the curve.
In class AB, the idea is usually to bias so that you start out just above the non-linear portion, so that all positive going drive sends you only up through the linear portion of the curve. Of course, when any one side of a Push-Pull pair is driven toward "off" it travels back down the non-linear area. Thus the gains do not sum perfectly, and the earlier comment about "crossover distortion" applies.
In class A, you can drive out of the linear region on peaks! This is *part* of the sound you get from pure triode Class A ZERO feedback amps. There is effectively some *compression* when driven hard into the ends of the linear region (more signal does not get you equally more output). Which in part accounts for why small pure triode amps can sound like they play louder than very linear solid state amps.
Feedback changes this, since it forces the input to be whatever is needed to keep things linear (in this example, more drive is required until a 1:1 gain relationship is acheived).
Here when I say "1:1" gain relationship it does not mean that the gain is unity, but if it is 1 input unit then you get out "N" output units, 5 input: 5N output.
Which is best? That's entirely unclear.
To add to the design mix - not all devices, tubes or transistors *ARE LINEAR* by themselves! Most are NOT. Most
are either entirely non linear everywhere or linear over only a small portion of their range. This is a large part of why feedback of all sorts is employed in practice. Keep in mind that the main reason a cathode follower is linear, is that it uses 100% feedback! :- )
Bear, your encore deserves another BRAVO!
This is my understanding of the subject.
Live music are complex vibration waves in air, an event which occur in space and time. Real live music can only happen as "pure class A" events in a mechanical sense. The pioneering audio amplifier engineers got it all right in the beginning. Back then there were mainly SE pure class A tube amps of very low power driving horn load or similar designs high efficiency speakers. (The success/shortcomings of such early speakers is another story for another A'gon thread.)
IMHO the reproduction of this real class A mechanical music event(recorded) can't get closer than through a pure class A electronic amplifier. I will not attempt to enter into the argument of "distortions" harmonic or others, of different types of amp designs, because honestly I cannot hear them. This is due, I can fairly suspect to the real event recording is not distortion free either, nor is my listening room perfect. So the minute amp distortions, added to the overwhelming inherent distortions, has become a non issue. In my own experience, my current SE pure Class A monoblocs provide that uncanny "breath" and "feel" in space and time which blew away all my previous equipment of various classes, AB & B etc. Ironically they all have more impressive distortion measurements in the catalogs than the Class A. Apparently the distortion numbers had not affect the way the "breath" and "feel" being reproduced. Don't take my word fully, go try out some pure Class A amp in your home setup to experience it. Sorry, you may need another set of speakers too.
I earn my bread as an electrical techy, but for my audiophile pursuits I will push aside my scopes and meters and use my ears. I wonder whether the audiophile equipment tradition had it all wrong, paying to much attention to measuring the unimportant parameter. I have no problem with the various class modes of amp operation, but today i take a break cos the posts above done a wonder job. I will attempt it in lay terms.
IMHO classes AB,B,C etc will always be commercial or technical compromises. Some came into scene to drive low efficiency modern speakers. Some such designs rightly belong in professional audio or even public announcing systems, not home audio. Chopping pure signal apart, reflecting mirror images of it, massaging it with some feedback at some point, and then reconstitute it for consumption. This is not real milk.
IMO pure Class A, warts and all is the way to go for the purist audiophile.