That's interesting. So there's no audible difference betwwen 90F and 100F, or between 100F and 110F, or between 90F and 110F. But there are audible differences between 90F and 95F, between 105F and 90F, between 105F and 100F? Funny.
Anyway, my point (which has been lost after several replies-to-replies), was as follows. The statement was: class-A is always better than class-AB. So I take a class-A and raise the supply voltage. Does that change the sound? I wouldn't think so, if the amp is reasonbly well designed, but if it changed it might well be for the better (less dynamic Vce modulation, for one thing). Yet that class-A amp is now transformed into a class-AB amp
jd
Increasing bias current decreases the inherent output impedance of the output stage. Imagine a class B EF: one device has a Vbe/offset of .7V at 0A, but an offset of .8V at 1A. This constitutes a resistance of 100 mohms. In class A, since both devices work at the same time, the resistance of both devices is parallel, making for 1/2 the output impedance of class B.
Even if a no-feedback OPS somehow has very low voltage error, there should still be some error due to the reactance of the speakers combining with the output impedance, which will modulate depending on the current draw and device characteristics. This will be worst for class AB (if I'm not mistaken), since Zout is constantly modulating between 100mohms, 50mohms (during crossover), and then 100mohms again. Class B will hump at crossover.
If you have only GNFB, the error of this output impedance transfers through the entire feedback loop before being corrected. This doesn't sound good especially if you consider EMI entering through the speaker cables. I'm not sure but if an amplifier is constantly responding to radio waves right up to the parts' Ft, is this very healthy for linearity?
Don't ask me what all this sounds like. I don't know.
Just to add something to this mighty thread.
- keantoken
Even if a no-feedback OPS somehow has very low voltage error, there should still be some error due to the reactance of the speakers combining with the output impedance, which will modulate depending on the current draw and device characteristics. This will be worst for class AB (if I'm not mistaken), since Zout is constantly modulating between 100mohms, 50mohms (during crossover), and then 100mohms again. Class B will hump at crossover.
If you have only GNFB, the error of this output impedance transfers through the entire feedback loop before being corrected. This doesn't sound good especially if you consider EMI entering through the speaker cables. I'm not sure but if an amplifier is constantly responding to radio waves right up to the parts' Ft, is this very healthy for linearity?
Don't ask me what all this sounds like. I don't know.
Just to add something to this mighty thread.
- keantoken
Last edited:
I am not 100% sure whether I'm absolutely right or not
I think a person's amp of choice would be that amp which presents the music to him the way he wants it. Unfortunately, most people do not understand the difference between what they want and what their ears want, or to say what their Primary Auditory Cortex portion of the brain wants.
About 2 years ago I made a Class-AB amp with global NFB for one of my relatives here. He likes various kind of music but is fond of classical and meditation related music and the amp runs almost all the time at their home; whenever I take a walk by their home I hear it playing some nice classical piece. He was quite happy with the amp.
Some months ago I made a Class-A amp with no global NFB. On a Sunday in the morning I was listening to various music tracks one after another randomly, just for testing the amp. Suddenly he arrived at our home and was talking to dad at the ground floor. I, at the first floor, was running the amp a bit louder than usual and the windows were left open. He heard the music and almost ran to my room and knocked the door. He stayed the whole day till 4 pm just listening to all the music I played. When leaving, he asked me - "Would you please make a pair of those for me?".
I said, "Why, you already have so good and powerful an amp already at home; and you play the machine all the time and if this class-a is left turned on 18 hours a day all it'll do is to increase your monthly electricity bill by 50% or more." It didn't work, he started coming our home once every two days just to listen to the new amp and to praise it! Well, seeing my hand-made amp being praised feels good but it was too much to listen to his "exclamations" about the "details" and "life" of the amp, over and over.
Fortunately both of us have the same kind of full range drivers fit in same sized boxes wired by same kind of wires with no passive xover. I found it to be a great chance to test the amp on him! One day he came and I started playing all the hard rock and pop and metals one after another; After he listened for about an hour I asked him to kindly tell what he feels about the reproduction of these tracks; and he said, "mmm... They don't sound good enough... Why is it so? At least the sound is not as good as it is on the previous amp you made." Life is full of wonders!
Anyway.
The above made me come to the conclusion-
Music which has hits, thumps, blasts, kicks, strikes sound better on Class-AB. Obviously those parts of music takes a small place in the whole track, but they can immediately force one's brain pay attention to whatever is being played; no matter whether he likes that music or not. Like he is made to listen to it. Class-A is lacking in this area. Although they reproduce the music theoritically better than the AB, they just don't sound as good in the above scenario. Most of my friends love this kind of music and everyone of them has a Class-AB in their home; many of them listened to their favorite tracks in my Class-A and most of them DIDN'T like the sound, naturally.
On the other hand, music which doesn't have those "hits, thumps, blasts, kicks, strikes" or have very little of them, sound better in a Class-A; Music which doesn't force one to listen, but which one wants to listen to is better played on Class-A. In this area Class-AB is lacking and my tests on the aforementioned relative(my uncle actually) says this clearly. He almost always went to sleep whenever I played some smooth, easy listening music on the Class-A. Fortunately I am a classical music lover and Class-A is my choice.
I know that the above is just subjective opinion, but I said what I saw by myself and what I think. Also there are many smooth music lovers who like AB over A, I think the sole reason behind it is that none of them is Me.
Inherent disadvantages in Class-A will never go off, still people will love this topology in same numbers as others who love AB, D or F or G or H,I,J,K,L etc....
The two amps I talked about and made are these
http://sound.westhost.com/project3a.htm
and
http://sound.westhost.com/project83.htm
I think a person's amp of choice would be that amp which presents the music to him the way he wants it. Unfortunately, most people do not understand the difference between what they want and what their ears want, or to say what their Primary Auditory Cortex portion of the brain wants.
About 2 years ago I made a Class-AB amp with global NFB for one of my relatives here. He likes various kind of music but is fond of classical and meditation related music and the amp runs almost all the time at their home; whenever I take a walk by their home I hear it playing some nice classical piece. He was quite happy with the amp.
Some months ago I made a Class-A amp with no global NFB. On a Sunday in the morning I was listening to various music tracks one after another randomly, just for testing the amp. Suddenly he arrived at our home and was talking to dad at the ground floor. I, at the first floor, was running the amp a bit louder than usual and the windows were left open. He heard the music and almost ran to my room and knocked the door. He stayed the whole day till 4 pm just listening to all the music I played. When leaving, he asked me - "Would you please make a pair of those for me?".
I said, "Why, you already have so good and powerful an amp already at home; and you play the machine all the time and if this class-a is left turned on 18 hours a day all it'll do is to increase your monthly electricity bill by 50% or more." It didn't work, he started coming our home once every two days just to listen to the new amp and to praise it! Well, seeing my hand-made amp being praised feels good but it was too much to listen to his "exclamations" about the "details" and "life" of the amp, over and over.
Fortunately both of us have the same kind of full range drivers fit in same sized boxes wired by same kind of wires with no passive xover. I found it to be a great chance to test the amp on him!
One day he came and I started playing all the hard rock and pop and metals one after another; After he listened for about an hour I asked him to kindly tell what he feels about the reproduction of these tracks; and he said, "mmm... They don't sound good enough... Why is it so? At least the sound is not as good as it is on the previous amp you made." Life is full of wonders!Anyway.
The above made me come to the conclusion-
Music which has hits, thumps, blasts, kicks, strikes sound better on Class-AB. Obviously those parts of music takes a small place in the whole track, but they can immediately force one's brain pay attention to whatever is being played; no matter whether he likes that music or not. Like he is made to listen to it. Class-A is lacking in this area. Although they reproduce the music theoritically better than the AB, they just don't sound as good in the above scenario. Most of my friends love this kind of music and everyone of them has a Class-AB in their home; many of them listened to their favorite tracks in my Class-A and most of them DIDN'T like the sound, naturally.
On the other hand, music which doesn't have those "hits, thumps, blasts, kicks, strikes" or have very little of them, sound better in a Class-A; Music which doesn't force one to listen, but which one wants to listen to is better played on Class-A. In this area Class-AB is lacking and my tests on the aforementioned relative(my uncle actually) says this clearly. He almost always went to sleep whenever I played some smooth, easy listening music on the Class-A. Fortunately I am a classical music lover and Class-A is my choice.
I know that the above is just subjective opinion, but I said what I saw by myself and what I think. Also there are many smooth music lovers who like AB over A, I think the sole reason behind it is that none of them is Me.
Inherent disadvantages in Class-A will never go off, still people will love this topology in same numbers as others who love AB, D or F or G or H,I,J,K,L etc....
The two amps I talked about and made are these
http://sound.westhost.com/project3a.htm
and
http://sound.westhost.com/project83.htm
Last edited:
No more "Inherent disadvantages" for so long?
You compared 2 different topologies: one that has a disadvantage of crossover distortions, but advantage of linear characteristic on high power due to feedback, it is good for rock music. Another has disadvantage of distortions on high power, and advantage of linear characteristic on low power due to single-ended nature, it is good for classical and other natural music.
It was not a comparison of class A and class AB, it was comparison of 2 different topologies.
The single "inherent disadvantage" of class A still remain: relatively high power dissipation, and it is the single one.
Yes.
And I was wondering why some people still ignore such an obvious thing.
That's why my Tower amps are excellent for any genre of music.
Jan Darth,
a lot of the Class A classics were more like high bias level Class AB, so there's not even a need to raise the voltage.
Example is the AudioAnalyse A9, or Duson A9 after AA went belly-up.
Claimed to have been All Class A but in reality Class A up to 25W continuous and 65W in AB.
(i'm sure you'd look smashing in a black robe)
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.