If you didn’t know better, you might think the phrase “class A amplifier” was a marketing term to help sell amplifiers. But it is, of course, actually a technical description of an amplifier that doesn’t distort the input waveform because it doesn’t depend on multiple elements to handle different areas of the input waveform. Want to know more? [FesZ] has a new video covering the basics of class A amplifiers including some great simulations. You can see the video below.
A class A amplifier uses a transistor that is always biased on. It never saturates or switches off. This is good for linearity, but not always the best for efficiency so there are other classes of amplifiers, too. However, for many applications, class A is the most common configuration.
There are a number of trade-offs involved with each type of amplifier and [FesZ] covers them in detail. But the real interesting part is the simulations in Spice. Sure, you can build the circuits and look at everything with a meter or scope, but using Spice is much handier.
There is a second video upcoming. We hope he covers other amplifier types too, as you really do want to understand the differences when you need to design something. If you want more Spice stuff, check out some of our previous posts. If for some reason, you don’t like LTSpice, there’s always Micro-Cap 12.
Of course, there is reason for class A. Class C may be more efficient, but you wouldn’t want the distortion for audio or some RF use.
Class C is the canonical form of amplifier for RF power amplifiers in FM service…
Also for AM — but not SSB.
It won’t amplify AM without removing the modulation.
So you use Class C to amplify the RF signal,then p!ate modulate it.
Fesz is far and away my favorite electronics youtuber. He’s helped me, a humble ME, understand multiple AC concepts better in 30-90min than hours and hours of reading did.
I use my 20 wpc First Watt F4 Class A biased amp for my two channel home theater to drive a pair of sealed floor standing speakers with Mitsubishi Diatone drivers and 15″ woofers. Nominal impedance ~ 8 ohms. Sensitivity’s ~ 83db SPL @ 3 ft so zero loading problems in my master bedroom. Presumably, this is partly why I can get by using this emitter follower amp where voltage gain is < 1 https://en.wikipedia.org/wiki/Common_collector#Basic_circuit And it may also be why I can feed the output of my Oppo 95 BD player directly into the F4 amp without need of a line amp, as its input impedance is 47K. I only use this system for Blu Ray movies not for long music listening sessions, so I don't feel too guilty about the heat emissions. And except for times between July & August the amp helps to heat the room for offsetting my partial anemia.
Cool story. Good job.
“actually a technical description of an amplifier that doesn’t distort the input waveform because it doesn’t depend on multiple elements to handle different areas of the input waveform.”
Well… It surely doesn’t have the cross over behavior like class AB amps.
And it doesn’t behave like a class D one.
And isn’t as whacky as the class C that mostly sees use in RF applications.
But saying it doesn’t distort the waveform is a bit brave…
Especially when some class A amps vary their Vcc to save in on power. Becoming some weird class of amplifier of their own… I would be surprised if such schemes didn’t cause a bit of distortion.Though, transistors have non linear behavior to start with, as well as the other parasitics of the circuit. Distortion is hard to just remove.
To be fair, I wouldn’t be surprised if a good quality class D amp can beat a low range class A one when it comes to distortion.
>To be fair, I wouldn’t be surprised if a good quality class D amp can beat a low range class A one when it comes to distortion
I’ve seen something like that documented some time ago, but can’t find it due to search engines getting worse and worse.
@Alexander Wikström said: “To be fair, I wouldn’t be surprised if a good quality class D amp can beat a low range class A one when it comes to distortion.”
Today properly designed and constructed Class-D audio power amps can hands-down beat even highly integrated Class-A and AB amps in all specification categories. Class-D audio power amplifiers are the modern version of yesterday’s analog “Chip-Amps”. [Warning: Tread lightly, we’re in “Audiophile Territory” now where Science and Engineering facts may be easily dismissed.]
The key words here are “properly designed and constructed”. Due to the switching nature of Class-D amps, power supplies, board layout, and component selection are difficult subjects compared with analog Class-A and AB amp design.
Fortunately, the chip manufacturers provide detailed documentation and usually-good reference designs to follow. Do as you are told and you should be fine with Class-D.
See references [1] & [2] below from Texas Instruments for Class-D audio power amp parts introductions and parts selection.
* References:
1. High power class D amplifiers TI slyw048.pdf
https://www.ti.com/lit/pdf/slyw048
2. Class-D Audio Speaker Amplifier Product Selector
Analog Input:
https://www.ti.com/audio-ic/amplifiers/speaker-amplifiers/products.html?keyMatch=CLASS-D%20AMPLIFIER#p1055max=Stereo&p2982=Analog%20Input&p89=Class-D
Digital Input:
https://www.ti.com/audio-ic/amplifiers/speaker-amplifiers/products.html?keyMatch=CLASS-D%20AMPLIFIER#p1055max=Stereo&p2982=Digital%20Input&p89=Class-D
“[Warning: Tread lightly, we’re in “Audiophile Territory” now where Science and Engineering facts may be easily dismissed.]”
Is exactly why I reserved myself from saying that a class D amp has many favorable properties that can make it better than a class A amp both in terms of distortion, but also power efficiency.
But personal biases for the listener and expectations tends to make people way more finicky than what is actually on paper having the flattest response.
People get used to the character of their setup and consider everything else “worse”, unless they can be persuaded by things they consider to be inherently better. Measured performance isn’t as important for most people.
But just like food, one can have a taste for a specific type of sound, so one can think one setup sounds better than another. Even if it has worse specs.
I was going to say somewhat the same thing, you just beat me to the punch.
But to add to the conversation. Class A amps are often used because the component count is less and they cost less to build. This has nothing to do with linearity or distortion. And when you are stamping out thousands of widgets, 50 cents each gets pretty substantial.
Class A is indeed an amp that is relatively cheap to construct.
Though, class D amps is rapidly taking its place. Since this can sometimes be a single chip solution needing 1 filter cap for the power rail and an inductor on the output, so 3 components and one has an amp with rather high power efficiency, perfect for battery driven applications.
> But saying it doesn’t distort the waveform is a bit brave…
That’s very generous… saying class A doesn’t distort the waveform is simply untrue. As someone who has built class A amplifiers, it is actually difficult to build one that doesn’t distort the input waveform. This is due to the inherent non-linearities in the response of real transistors and/or tubes.
Yes, making a poor class A amp is rather easy.
If one takes a random transistor a few resistors from the closes drawer and potentially a cap or two, it isn’t going to be the best amp in the world, it might even let out the magic smoke.
Getting it to have little distortion is a fair bit finicky to say the least.
Class D amps are far more trivial in comparison. Since it is just a half bridge driven hard on or off, being filtered by a suitably large inductor. Here switching speed is the only real thing of importance, the rest is just PWMing the output to it, and measuring the output with a voltage divider, and simply compare it to the input voltage.
To be fair, a class D amp is a good first PWM oriented microcontroller project. Especially if the micro has a decent comparator in it, but one can always just use an external one if one wants something better.
I agree. I built a couple of amplifiers when I was a kid, and indeed transistors and tubes have a response that is not linear. What is interesting with Class A though, is that the harmonics are even, so they fall on the next octave and they sound musical (no, I am not a musician, ask them to explain why). With a Class AB, the harmonics are odd and sound horrible. One shouldn’t look only at the figures, but also at the psychological aspect of sound.
A properly designed class AB high power analog audio amplifier can provide 0.001% distortion 20 Hz – 20 kHz. at full power out. None of the Texas Instruments devices in your citations can do that.
Hello guys!
What software do you recommend for building and testing an A class amp? Sounds like fun!
I want to do a heaphone amp.
Exactly my thoughts. A properly well designed class AB amplifier (check the Douglas Self book about power amplifiers) is really hard to beat in terms of simplicity, low distortion, sound quality and output power. Not the most power efficient design (around 50 to 65% efficiency) but it certainly still has its place.
What makes you think at the levels concerned the difference is bearable?
A properly designed amp is one that meets your or your customer’s needs, and there are very few needs for an amplifier with an advertised 0.001% distortion across the entire audio range.
In the mid 1980’s I designed a tunable analog audio oscillator that was used for calibration and it’s distortion could be no more than 0.1% across the entire audio range. That’s 100X worse than you are quoting! It worked well, but to succeed I used all the same value high end Mylar capacitors across the entire circuit, some instrument rated OP Amps and high stability resistors. The oscillator was followed by a 0 gain buffer, technically it was essentially a class A amplifier. The amp had the same 0.1% distortion rating as the oscillator.