Class A vs AB -- Do They Really Sound Different?

[GXAlan]

My experience aligns with @DonH56I have a TOTL Marantz PM-95 (and a PM-94 class A/AB) MOSFET integrated from 1990 with switchable Class AB (120wpc) or Class A (30wpc) and apart from the increased physical transformer noise when switching to Class A and the massive increase in heat produced, the subjective difference is zero, that is until you run out of power in Class A and it is very obvious.

A huge amplifier, 27.5kg built on a diecast chassis, all copper plated and several kilogram heatsinks to get rid of the heat in class A.

Guess who sent @amirm a TOTL Marantz PM-90? That's one generation newer than the PM-95, and drops the on-board DAC resulting in better "advertised specs" of 111 dB vs 106 dB SNR and a THD of 0.0015% instead of 0.0080%. It's the last Marantz before HDAM was introduced in the PM-99SE (which resulted in a poorer advertised THD of 0.0050%)?

It's also interesting in that it used dual EL transformers but it is not a dual mono design. The use of two transformers is reportedly designed to allow the flux leakage to partially cancel out.

I have an Accuphase P-266 in addition to the Marantz. Accuphase's own data suggests that Class A decreases the 20 kHz distortion but in the 1st watt, it's actually worse for the 1 kHz and 20 Hz distortion compared to their AB mode. I'm really curious to see the 32-tone test on the PM-90 between Class A and AB. It may be interesting to test the PM-90 at 100mw, 1W, 5W, 10W in Class A vs. AB.

 

[Tom C]

Please excuse this elementary question.
There are different kinds of distortion. Intermodulation distortion is tested separately from total harmonic distortion, for example. What is the test for quantifying crossover distortion? It won’t show up in tests for IMD or THD, right? Or will it?

 

[DonH56]

Please excuse this elementary question.
There are different kinds of distortion. Intermodulation distortion is tested separately from total harmonic distortion, for example. What is the test for quantifying crossover distortion? It won’t show up in tests for IMD or THD, right? Or will it?

Crossover distortion happens in synch with the signal so will usually show up in normal THD/IMD tests, though it typically is not sinusoidal and has higher-frequency terms that show up in a frequency plot (FFT). It can cause higher distortion at lower signal levels and contribute to the rising SINAD plots at lower power levels, though these days a lot of that is from the noise since crossover distortion is so low, especially in a class AB design (that shifts into class A at lower levels). The noise output is usually fairly constant, so as signal level drops, noise level relative to the signal increases, and THD+N or SINAD plots rise at low power levels.

HTH - Don

 

[levimax]

Thank you for replying.

class A wins, it will have lower noise and lower distortion, I claim it is inherently superior, but bad design and or manufacture is another story.
Class A or class AB reffers to the output stage only, so let's keep that in mind.

There is more to a Class A amp than the output stage. The PS for Class A has completely different requirements (and should be designed differently) than a Class AB amp. When you are running a PS at full output all the time as in a Class A it is really hard to keep the noise down on the rails relative to a low level output. This is why Class A tend to be noisier at low output. For class AB the PS only puts out full power at high output power so the PS noise is at a maximum when the signal is at a maximum, for Class A PS noise is always at the maximum regardless of output.

Class A is not more or less "magic" than Class AB .... the design and use cases for each are different and in most audio power amp use cases class AB can equal or exceed Class A performance and be cheaper and more efficient.

 

[Tom C]

Crossover distortion happens in synch with the signal so will usually show up in normal THD/IMD tests, though it typically is not sinusoidal and has higher-frequency terms that show up in a frequency plot (FFT). It can cause higher distortion at lower signal levels and contribute to the rising SINAD plots at lower power levels though these days a lot of that is from the noise since crosover distortion is so low, especially in a class AB design (that shifts into class A at lower levels). The noise output is usually fairly constant, so as signal level drops, noise level relative to the signal increases, and THD+N or SINAD plots rise at low power levels.

HTH - Don

Thank you Don. Very helpful, as always!

 

[MakeMineVinyl]

Please excuse this elementary question.
There are different kinds of distortion. Intermodulation distortion is tested separately from total harmonic distortion, for example. What is the test for quantifying crossover distortion? It won’t show up in tests for IMD or THD, right? Or will it?

Class A cannot have crossover distortion since there's no crossover at the zero crossing between two sets of transistors. The easiest test for crossover distortion is to look at the distortion products on an oscilloscope since crossover distortion has a unique visual signature. Lacking an oscilloscope, its harder to judge the presence of crossover distortion since the distortion products are all lumped together, regardless of the mechanism for their generation.

 

[Pdxwayne]

I have a Parasound A21 that operates in class A for the first 10 watts. After that, it operates in class AB.

As compared to using class AB Yamaha amp, the Parasound sounded a bit more 3D. This is consistent when using my 3 ways floorstander and my Paradigm Persona B.

The extra 3d feels is due to class A or something else?

 

[MakeMineVinyl]

Class A is not more or less "magic" than Class AB .... the design and use cases for each are different and in most audio power amp use cases class AB can equal or exceed Class A performance and be cheaper and more efficient.

There's no reason class A/B cannot be as clean, or cleaner, than class A. As you noted, the power supply requirements are more rigid in class A. The most difficult working region with class A/B is right near the area where the signal transforms from class A to A/B. There are a lot of engineering challenges, compromises and market realities in choosing how to bias a class A/B amplifier. Engineers are never completely free to design to an 'ideal' solution because of these realities.

 

[Tom C]

Class A cannot have crossover distortion since there's no crossover at the zero crossing between two sets of transistors. The easiest test for crossover distortion is to look at the distortion products on an oscilloscope since crossover distortion has a unique visual signature. Lacking an oscilloscope, its harder to judge the presence of crossover distortion since the distortion products are all lumped together, regardless of the mechanism for their generation.

This much I understand. But if crossover distortion shows up in the customary tests, then comparing and choosing an amp becomes much simpler, and topology can considered incidental. Just pick the one with the specs that most closely match your needs, and stop worrying about A vs. B vs AB vs D, etc.

 

[MakeMineVinyl]

The extra 3d feels is due to class A or something else?

I'd imagine its something else, but if you're enjoying it, who am I to question it?

 

[MakeMineVinyl]

This much I understand. But if crossover distortion shows up in the customary tests, then comparing and choosing an amp becomes much simpler, and topology can considered incidental. Just pick the one with the specs that most closely match your needs, and stop worrying about A vs. B vs AB vs D, etc.

Yes, but like I said, the most straightforward, and in reality the only real way to see crossover distortion is to look at the distortion artifacts on an oscilloscope. No reviews that I've seen routinely do that, and even if they did, most people wouldn't be able to interpret it. All the other numerical data, and even FFT plots, cannot isolate crossover distortion from all the other forms of distortion in any meaningful way.

 

[levimax]

This much I understand. But if crossover distortion shows up in the customary tests, then comparing and choosing an amp becomes much simpler, and topology can considered incidental. Just pick the one with the specs that most closely match your needs, and stop worrying about A vs. B vs AB vs D, etc.

I agree with this.

One of the standard audiophile arguments for Class A is that feedback is not necessary to eliminate crossover distortion like on a Class AB amp. Traditionally audiophiles do not like feedback in my opinion because they don't understand it and it "sounds like a bad thing".

 

[Pdxwayne]

I'd imagine its something else, but if you're enjoying it, who am I to question it?

One thing I don't really like about the Parasound is that it produces slight hum. I can sense that light hum from my speakers like 10 ft away. I can't sense it once I play music, so not too bad. I wish it could be as quiet as my other class AB and class D amps.

Again, not sure if it is due to class A vs AB....

 

[MakeMineVinyl]

I agree with this.

One of the standard audiophile arguments for Class A is that feedback is not necessary to eliminate crossover distortion like on a Class AB amp. Traditionally audiophiles do not like feedback in my opinion because they don't understand it and it "sounds like a bad thing".

The problem is that feedback usually does not eliminate crossover distortion completely - it gets back to market realities and compromises. Crossover distortion can be reduced to the point of being a non-issue for essentially everyone, but then idiots like me who have horns come along and are bothered by it.

 

[Tom C]

Actually, when I first started to learn electronics (late 1970’s), we were taught class A has lowest distortion, specifically because of the presence of crossover distortion in AB and B. Class B sounds awful, like the kind of thing that comes out of Radar’s PA horn on MASH, and is just for when max output is needed, like addressing the public outdoors. Class AB was considered to be somewhere in the middle. But we were techs, not taught by engineers. We had to be able to build and repair, not design.
IC’s at that time were limited mostly to microprocessors and memory chips. There were a few differential amplifiers on a chip, but their advantages were limited to small and cheap. The specs weren’t that good, IIRC. We really had a hard time accepting that transistors could ever sound better than tubes, because transistors are for battery powered AM radios that you can take to the beach. Or put in a car. But not for serious listening at home.
Time marches on, and much progress has been made. Old habits die hard.

 

[Tom C]

I agree with this.

One of the standard audiophile arguments for Class A is that feedback is not necessary to eliminate crossover distortion like on a Class AB amp. Traditionally audiophiles do not like feedback in my opinion because they don't understand it and it "sounds like a bad thing".

When your first name is negative, you must be bad, right?

 

[levimax]

The problem is that feedback usually does not eliminate crossover distortion completely - it gets back to market realities and compromises. Crossover distortion can be reduced to the point of being a non-issue for essentially everyone, but then idiots like me who have horns come along and are bothered by it.

Is "crossover distortion" any more or less "bad" than any other type of distortion? My understanding is that it will show up in the standard THD and IMD measurements, is that correct? In your use case I am understanding that at very low output levels crossover distortion will show up at a relatively high level and audible, why doesn't feedback work to eliminate crossover distortion at these low levels? Not challenging just trying to learn.

 

[MakeMineVinyl]

Is "crossover distortion" any more or less "bad" than any other type of distortion? My understanding is that it will show up in the standard THD and IMD measurements, is that correct? In your use case I am understanding that at very low output levels crossover distortion will show up at a relatively high level and audible, why doesn't feedback work to eliminate crossover distortion at these low levels? Not challenging just trying to learn.

Yes crossover distortion components would be reflected in THD measurements and such but the problem is telling specifically what distortion is coming from crossover distortion or from other sources. That's where an oscilloscope comes in. Since I work on the engineering end I need to know the exact character of crossover distortion as distinct from all other forms of distortion. Since typical crossover distortion artifacts are at very high frequencies were the available feedback is at its lowest, it's hard to completely eliminate.

 

[levimax]

Since typical crossover distortion artifacts are at very high frequencies were the available feedback is at its lowest, it's hard to completely eliminate.

OK thanks. For modern AB amps I am not sure lack of feedback is an issue even at high frequencies. I switched out my 25 watt Class A amp I was running on my tweeters in a tri-amp system with a Neurochrome Mod-86 amp which is using something like 180 dB of negative feedback. The amps measure well even at low output and I can't tell a difference except the Neurochrome amps are quieter and especially much cooler running and smaller and lighter. For 2.5 watts like you need it is a different game. More than one way to skin a cat.

 

[Ken Tajalli]

I agree with this.

One of the standard audiophile arguments for Class A is that feedback is not necessary to eliminate crossover distortion like on a Class AB amp. Traditionally audiophiles do not like feedback in my opinion because they don't understand it and it "sounds like a bad thing".

I believe it is "global feedback" audiophiles don't like.
and yes, class A requires less GFB.