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

[MakeMineVinyl]

/notes MakeMineVinyl's "Manufacturer" tag

/giggles a bit at the delightful irony

Chronicles From The Sausage Factory

 

[DonH56]

John Siau, Benchmark's VP/Dir of Engr, has discussed crossover distortion in the AHB-2 thread.

 

[MerlinGS]

...You haven't done the basic experiment. So seriously, your sonic claims have the evidentiary value of similar ones from Elsie Wright and Frances Griffiths.

In fairness, I could usually tell when there were a couple of Classe DR3 running in bridge mode in an audio system. The rooms were always warmer than other rooms giving the sound a nice warm feeling

I should note the DR3s were one of the few amps that could comfortably drive the Apogee Scintilla (a reminder, its impedance could drop below 1 ohm).

 

[watchnerd]

the Apogee Scintilla

Sweet Christmas, I haven't heard the Scintilla name since I was in college.

 

[Head_Unit]

Any reason, if assuming no Class D amps, to prefer Class A over Class AB at a given price point?

As I just posted elesewhere, I really hate the entire idea of "Class XXX amps sound YYY." It is a ridiculous, stupid idea. Within an operation class there is huge leeway, so you could design a great amp within any class of operation or a piece of excrement. To say "all class D suck" or whatever is nonsense, especially as I'd presume 96.4%+ of the people parroting those observations have never heard large quantities of amplifiers let alone direct comparison let alone blind comparison.

In a previous life I actually did listen to a lot of different amps, classes AB and D and even T-they just sometimes sounded different, more to do with the individual design team's skill than anything else I think (well, and price!).

 

[avanti1960]

As I just posted elesewhere, I really hate the entire idea of "Class XXX amps sound YYY." It is a ridiculous, stupid idea. Within an operation class there is huge leeway, so you could design a great amp within any class of operation or a piece of excrement. To say "all class D suck" or whatever is nonsense, especially as I'd presume 96.4%+ of the people parroting those observations have never heard large quantities of amplifiers let alone direct comparison let alone blind comparison.

In a previous life I actually did listen to a lot of different amps, classes AB and D and even T-they just sometimes sounded different, more to do with the individual design team's skill than anything else I think (well, and price!).

My experiences in listening tells me that there are tendencies (strong points and weaknesses) in the sound within the given classes. As you noted within a given class execution and sound quality can vary.

 

[Head_Unit]

My experiences in listening tells me that there are tendencies (strong points and weaknesses) in the sound within the given classes. As you noted within a given class execution and sound quality can vary.

Yes I would agree with that myself-BUT also believe that can get mired in historical experience: people saying for decades that all transistor amps must suck based on their experience in the early 1970s, or that all Class D amps suck based on early 90s experience. I also don't believe those "tendencies" apply at all well to individuals many newbies of whom would likely be perfectly happy with many amps or cables or disc players if their brains did not get pre-swayed by others' opinions

 

[Vespa_SS180]

Like with many other audiophile topics, there is a clear line between “how it was” [in the 20th century] vs what is the state of the art today. Yes, a couple of decades ago, a push-pull amplifier was cheaper, cooler, more powerful, and less-linear than a class A amplifier. However, this is no longer the case with todays power, control (and often integrated) components of a highly efficient and highly linear power amplifier. More and more becoming direct digital [power] conversion, erasing the difference between classical “transistor” architectures.

As a matter of fact, a push-pull amp can also be run in Class A, so there is no distinction between Class A and Push-Pull

 

[levimax]

As a matter of fact, a push-pull amp can also be run in Class A, so there is no distinction between Class A and Push-Pull

There is a difference between a single ended Class A amp and a Push Pull amp biased to Class A. In theory a single ended class A will have no crossover distortion at all even at the lowest output levels where as a Push Pull biased to Class A may have a small amount at very low output levels. There is a member here using super efficient horns (107 dB per watt) that says crossover distortion on anything besides a single ended Class A is audible on these horns.

 

[beren777]

I think a lot of these discussions would be simpler if we could agree to use the word "distort" rather than "sound." Depending on the skill of the design team, there may be distortions that are similar within classes of amps. That distortion could be characterized through measurements of signal in vs signal out. If the distortion isn't in that measurement and someone still says they can hear a difference, it's more likely to be speakers, room, or placebo effect causing a difference, not the amplifier. With current technology I feel the goal should be to make the reproduction chain as transparent as possible with regard to original signal, and if there is distortion you like, apply it via DSP. My opinion is that many "amplifiers" are also designed to impress a characteristic distortion on the input signal which some refer to as "house sound."

 

[MakeMineVinyl]

There is a member here using super efficient horns (107 dB per watt) that says crossover distortion on anything besides a single ended Class A is audible on these horns.

Yup, I did say that! With horns that efficient every little niggle and distortion has a 20 decibel magnifying glass over it. What wouldn't be noticed with a speaker of 85 DB efficiency is plainly obvious when amplified by 20 decibels.

 

[BrokenEnglishGuy]

is the crossover distortion a problem with normal speakers? 85 ~ 92dB ~, I have my purifi eigentakt and lost the interest in amplifiers

 

[Vespa_SS180]

There is a difference between a single ended Class A amp and a Push Pull amp biased to Class A. In theory a single ended class A will have no crossover distortion at all even at the lowest output levels where as a Push Pull biased to Class A may have a small amount at very low output levels. There is a member here using super efficient horns (107 dB per watt) that says crossover distortion on anything besides a single ended Class A is audible on these horns.

Right. I have no practical experience of Push-Pull amps run in Class A so far, only Class AB ( Line Magnetic LM-216iA ).
I would like to try my little single-ended class A amp, Reisong A10, ( EL34:s ) hooked up to an efficient horn speaker.
It drives my old Heybrook HB1 for now ( 90 dB/W m ) and even if it sounds very good, it is kind of lean in the lower register.

 

[Vespa_SS180]

Yup, I did say that! With horns that efficient every little niggle and distortion has a 20 decibel magnifying glass over it. What wouldn't be noticed with a speaker of 85 DB efficiency is plainly obvious when amplified by 20 decibels.

Hi! Your horns , have you built them yourself?

 

[MakeMineVinyl]

Hi! Your horns , have you built them yourself?

No. Aside from some added bracing in the low frequency horn cabinets they are close to factory stock. The high frequency horns have probably 15 lb of Aquaplas applied to their exterior to damp out any resonances. I was an engineer at the company who made them way back then.

 

[MakeMineVinyl]

is the crossover distortion a problem with normal speakers? 85 ~ 92dB ~, I have my purifi eigentakt and lost the interest in amplifiers

No. It shouldn't be a problem at all.

 

[tomchr]

There is a member here using super efficient horns (107 dB per watt) that says crossover distortion on anything besides a single ended Class A is audible on these horns.

Confirmation bias?

There's no scientific reason distortion should be more audible on a high-efficiency horn than on any other speaker. If the distortion is 120 dB down, it'll be 120 dB lower than the fundamental both on the horn and on the other speakers, thus will be equally audible (or inaudible as the case may be).

However, what could be going on is that the Class A designs were optimized for lower output power and delivered better performance at the uW to mW levels necessary for normal listening levels with high-efficiency speakers, whereas the Class AB/D designs were intended for higher power output and didn't have the same good performance at the uW/mW levels. If that is the case, it would have more to do with the amp design and less to do with whether it operates in Class A or any other class.

Tom

 

[Vespa_SS180]

No. Aside from some added bracing in the low frequency horn cabinets they are close to factory stock. The high frequency horns have probably 15 lb of Aquaplas applied to their exterior to damp out any resonances. I was an engineer at the company who made them way back then.

Any known "big" brand, or smaller niche builder?

 

[MakeMineVinyl]

Any known "big" brand, or smaller niche builder?

It was a very large and well-known company that did professional audio and consumer audio sort of like JBL but not JBL. I was an engineer at JBL also back in the day.

 

[MakeMineVinyl]

Confirmation bias?

There's no scientific reason distortion should be more audible on a high-efficiency horn than on any other speaker. If the distortion is 120 dB down, it'll be 120 dB lower than the fundamental both on the horn and on the other speakers, thus will be equally audible (or inaudible as the case may be).

However, what could be going on is that the Class A designs were optimized for lower output power and delivered better performance at the uW to mW levels necessary for normal listening levels with high-efficiency speakers, whereas the Class AB/D designs were intended for higher power output and didn't have the same good performance at the uW/mW levels. If that is the case, it would have more to do with the amp design and less to do with whether it operates in Class A or any other class.

Tom

Your sort of correct on the amplifier part. With extremely efficient horns we are listening constantly well well below a watt of power. A traditional class a b amplifier is biased in that region. With extremely efficient horns that discontinuity can be audible. Remember at 107 DB horn places a massive 20 decibel magnifying glass on the low level performance of the amplifier. Anything down there is magnified and in addition the very directional character of a horn amplifies this even more since everything noise floor and all are going directly to your ears and not so much the room. Another very practical advantage of a single ended triode amplifier in this application is the very limited two and a half watt output power. This protects the unobtainium diaphragms from accidental damage or damage from equipment fault. It's also a struggle to achieve a low enough noise level from these amplifiers so that hiss is not heard beyond perhaps a foot of the mouth of the horn. My particular amplifiers which use negative feedback have only two decibels voltage gain. That helps keep the noise inaudible with anything other than your ear right in the bell of the horn.