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

[watchnerd]

Thank you, I am in the same direction in my multichannel multi-amplifier project...

Good lord.

How do you have time to work on all that?

 

[tomchr]

My Parasound A51 claims something like this I think. It is a great sounding amp.

I don't doubt that. However, a high Class AB bias point does not provide the lowest distortion. That may not have been the point of the A51 design either. Many times marketing needs prevail over engineering optimizations in commercial products.

Push-pull class A easily transits to class AB. It is all about idle current and load impedance. No mysteries here. There may be variable bias as well, or class A+. Marketing x user's education, as always. Study, though it is not easy.

Actually, it is quite easy to study (but admittedly harder to fully understand). I highly recommend Douglas Self's article series in Wireless World about distortion in audio amplifiers. It's 7-8 articles throughout a year in the mid-1990s. The articles later became his book on Audio Power Amplifier Design. He walks you through the various distortion mechanism in a power amp, including the impact of output stage biasing. You get the lowest distortion in an emitter follower output stage when one thermal voltage (Vt = kT/q) is present across each emitter resistor.

Tom

 

[dualazmak]

Good lord.
How do you have time to work on all that?

Thank you for your attention on my project.

Well, I am a semi-retired medical-pharmaceutical Ph.D. profession now working as a consultant for various medical aspects including COVID-19 pandemic.

As summarized at here, I started to renovate my YAMAHA NS-1000 in July 2019, and moved on to multichannel multi-amplifier project in October 2019. I decided to move forward very carefully and steadily in the project as shared in the first post.

In October-December last year, I have seriously considered purchasing Accuphase DG-68 and/or DF-65. I abandoned that plan, however, when I "discovered" OKTO DAC8PRO and software crossover EKIO in February this year.

Now I could establish all digital 192 kHz 24 bit DSP processing within Windows 10 PC running JRiver (or Roon) very much cost-effectively with EKIO and DAC8PRO without unnecessary AD - DA steps. What I have achieved with EKIO and DAC8PRO is just same as the extraordinary expensive TRIONNV ALTITUDE32 is doing in 192 kHz 24 bit on its built-in Linux and software/firmware.

After reading and partially participated in this interesting thread, I would like to go back to this sentiment, I like the Halfway Tree's comment and the Linkwitz's words he refered;
But no matter how good the measurements are, at the end of the day our ears are the final arbiter as to whether we like the sound, and I do. As Siegfried Linkwitz used to say "What is important to the eye is not necessarily important to the ear...,"
even though I well know that Linkwitz has done so intesive objective measurements throughout his remarkable contribution to our audio community.

And, I really like your title of this thread "Class A vs AB -- Do They Really Sound Different?", where you did not say "Class A vs AB -- Do They Really Measure Different?"

I do hope all of you would please stay safe and enjoy music during the holiday season and (hopefully towards happy) new year.

 

[watchnerd]

And, I really like your title of this thread "Class A vs AB -- Do They Really Sound Different?", where you did not say "Class A vs AB -- Do They Really Measure Different?"

I do hope all of you would please stay safe and enjoy music during the holiday season and (hopefully towards happy) new year.

Ah..semi-retired..that helps a lot.

Some people seem to be going straight to the measurement aspect.

I assume that's because, if one can measure the right thing, it will tell us if any difference exists and is within the range of audibility.

 

[restorer-john]

If anyone wants to do a Foobar ABX, I can do a couple of level matched (electrical, at the speaker terminals, low level for crossover distortion detection- straight into the A/D of say a focusrite 2i2v2) short recordings of my big Marantz PM-95 Mosfet in Class A and then in Class AB.

This puppy:
https://audio-database.com/MARANTZ/amp/pm-95-e.html

http://www.thevintageknob.org/marantz-PM-95.html

 

[watchnerd]

If anyone wants to do a Foobar ABX, I can do a couple of level matched (electrical, at the speaker terminals, low level for crossover distortion detection- straight into the A/D of say a focusrite 2i2v2) short recordings of my big Marantz PM-95 Mosfet in Class A and then in Class AB.

This puppy:
https://audio-database.com/MARANTZ/amp/pm-95-e.html

View attachment 100584

http://www.thevintageknob.org/marantz-PM-95.html

Mac & Linux user here, no Foobar for me.

Otherwise, I'd love to.

 

[watchnerd]

If anyone wants to do a Foobar ABX, I can do a couple of level matched (electrical, at the speaker terminals, low level for crossover distortion detection- straight into the A/D of say a focusrite 2i2v2) short recordings of my big Marantz PM-95 Mosfet in Class A and then in Class AB.

This puppy:
https://audio-database.com/MARANTZ/amp/pm-95-e.html

View attachment 100584

http://www.thevintageknob.org/marantz-PM-95.html

I noticed in the specs that one of the digital sampling frequencies is....

32 kHz

Whaaaaaat?

What the heck was that used for, given it would have a ceiling of about 16 kHz in the audible range?

Radio or some low bandwidth medium?

 

[andreasmaaan]

If anyone wants to do a Foobar ABX, I can do a couple of level matched (electrical, at the speaker terminals, low level for crossover distortion detection- straight into the A/D of say a focusrite 2i2v2) short recordings of my big Marantz PM-95 Mosfet in Class A and then in Class AB.

This puppy:
https://audio-database.com/MARANTZ/amp/pm-95-e.html

View attachment 100584

http://www.thevintageknob.org/marantz-PM-95.html

Yes please (. ❛ ᴗ ❛.)

 

[restorer-john]

What the heck was that used for, given it would have a ceiling of about 16 kHz in the audible range?

Digital radio broadcasts. Also the LP (gave 240mins) mode on DAT recorders was 14bit non-linear @32KHz, topped out at 14.5KHz.

 

[watchnerd]

Digital radio broadcasts. Also the LP (gave 240mins) mode on DAT recorders was 14bit non-linear @32KHz, topped out at 14.5KHz.

What an interesting tidbit of audio history arcania.

 

[avanti1960]

What's to be sorry about?

After having done a lot of level-matched ABX testing, I'm just acknowledging my limitations and propensity for sighted bias.

I acknowledge that I have visual biases but it does not account for the subtle but obvious differences I hear with Class A amplifiers.

 

[SIY]

I acknowledge that I have visual biases but it does not account for the subtle but obvious differences I hear with Class A amplifiers.

You don't know that because you haven't done the basic experiment.

 

[avanti1960]

You don't know that because you haven't done the basic experiment.

please do not assume to understand what and how i hear.
i have been to many audio shows over the years and can usually spot a class A amp the minute i walk in to a room regardless if i am aware of the particular amplifier's topology or not.

 

[ahofer]

please do not assume to understand what and how i hear.
i have been to many audio shows over the years and can usually spot a class A amp the minute i walk in to a room regardless if i am aware of the particular amplifier's topology or not.

uh huh

 

[MakeMineVinyl]

I acknowledge that I have visual biases but it does not account for the subtle but obvious differences I hear with Class A amplifiers.

Undeniable fact: class A/B amplifiers can exhibit crossover distortion, and class A amplifiers cannot.

I know the scientists here will scream about correctly designed class A/B being 'perfect' and 'that problem was solved' yada, yada, yada.

 

[watchnerd]

I acknowledge that I have visual biases but it does not account for the subtle but obvious differences I hear with Class A amplifiers.

No offense, but unless you did a blind test, you don't really know.

I've had plenty of listening experiences that I thought were 'subtle but obvious' differences, but when I blind tested myself, I failed.

 

[watchnerd]

Undeniable fact: class A/B amplifiers can exhibit crossover distortion, and class A amplifiers cannot.

I know the scientists here will scream about correctly designed class A/B being 'perfect' and 'that problem was solved' yada, yada, yada.

Crossover distortion aside, I'm still waiting from the rips from @restorer-john to see if I can hear when his amp switches from A to AB.

Oh, and happy holidays!

 

[watchnerd]

please do not assume to understand what and how i hear.
i have been to many audio shows over the years and can usually spot a class A amp the minute i walk in to a room regardless if i am aware of the particular amplifier's topology or not.

If true, you should be able to pass an ABX test with ease, then.

 

[pma]

Undeniable fact: class A/B amplifiers can exhibit crossover distortion, and class A amplifiers cannot.

I know the scientists here will scream about correctly designed class A/B being 'perfect' and 'that problem was solved' yada, yada, yada.

You always find some traces of crossover distortion in class AB amplifier when measured with 4ohm load or smaller. The question is only the measurement resolution. The other question is audibility of very low levels of crossover distortion.

But class A is not automatically a cure to all distortion issues. In simple circuits, large signal distortion of loaded class A output stage exists, but has a different spectrum than class AB. To kill the distortions or to reduce them to negligible level, it is necessary to make a combination of local feedback in the class A output stage and global negative feedback. Then you can get some -130dBr distortion only H2 or H3. This is the difference against very good class AB, no higher harmonics if the design is excellent. These higher harmonics will remain even in the excellent class AB output stage.

One can see crossover distortion residuals even in the Benchmark AHB2 measurements
here:
https://www.audiosciencereview.com/...-ahb2-amplifier-audio-measurements-png.26575/

and even more here:
https://www.audiosciencereview.com/...tion-distortion-audio-measurements-png.26582/

 

[MakeMineVinyl]

But class A is not automatically a cure to all distortion issues.

I never wrote that class A is 'a cure for all distortion issues'; I specifically said crossover distortion, which clearly is not 'all distortion'.