what is the difference between a low power class A amp and a moderate power class AB amp with a very high bias dissipating the same quiescent power ?
Hi,
there is a design situation where the ClassA amplifier is exactly the same as the ClassAB amplifier.
Take a 25W into 8r0 ClassA amplifier as the reference.
The peak output voltage for a continuous sinewave output signal is 20Vpk.
The maximum output current when driving a resistive test load is 2.5A.
The bias of the ClassA output stage is just >1.25A, say ~1.3A.
Now look at a ClassAB amplifier with a 4pair output stage using 130W To247 devices and Re = 0r068
The optimum bias for ClassAB is Vre <26mV. Take Vre=22mV as the optimum determined by distortion/spectra analysis.
Ib~324mA/device.
The total bias current is 1.294A, it is biased identically to the reference amplifier.
Both amplifiers are 25W into 8r0 and both deliver 25W of ClassA into 8ohm.
Both will deliver substantially more ClassAB power into lower impedance loads.
there is a design situation where the ClassA amplifier is exactly the same as the ClassAB amplifier.
Take a 25W into 8r0 ClassA amplifier as the reference.
The peak output voltage for a continuous sinewave output signal is 20Vpk.
The maximum output current when driving a resistive test load is 2.5A.
The bias of the ClassA output stage is just >1.25A, say ~1.3A.
Now look at a ClassAB amplifier with a 4pair output stage using 130W To247 devices and Re = 0r068
The optimum bias for ClassAB is Vre <26mV. Take Vre=22mV as the optimum determined by distortion/spectra analysis.
Ib~324mA/device.
The total bias current is 1.294A, it is biased identically to the reference amplifier.
Both amplifiers are 25W into 8r0 and both deliver 25W of ClassA into 8ohm.
Both will deliver substantially more ClassAB power into lower impedance loads.
what is the difference between a low power class A amp and a moderate power class AB amp with a very high bias dissipating the same quiescent power
They will sound different. The sonic signature or summation of all the distortions and noise of each is different. Other than that, they are very similar.
They will sound different. The sonic signature or summation of all the distortions and noise of each is different. Other than that, they are very similar.
It is difficult to offer a general answer, as much depends on
the devices and the conditions.
🙂
My 2 eurocents worth: Assuming that the AB bias is the same as the class-A bias, this implies that the class AB has higher supply/higher max output.
They will both work in class A up to that bias point. After that, the class A will start to clip, while the AB goes into, well, AB. As long as they stay below the class A clip point, difference depends on topology and devices.
jd
if the two amps are mostly the same except one clips transients and another doesn't - then why not build all class A amps as heavily overbiased class AB amps ?
have transition from A to AB occur lets say 10db below clipping. if used to actively drive a tweeter such an amp would spend 99% of the time in class A.
?
have transition from A to AB occur lets say 10db below clipping. if used to actively drive a tweeter such an amp would spend 99% of the time in class A.
?
Last edited:
Hola Compadres,
Is not the point of class a bias to get the transistor to a more linear part of the curve and to avoid crossover distortion. Taking into consideration my simple mind and how things work once you get past the working area of class a most designs go to ab as a sort of soft clipping. Just a disclaimer I am not an engineer and I defer to those that I revere.
Regards, Elwood
Is not the point of class a bias to get the transistor to a more linear part of the curve and to avoid crossover distortion. Taking into consideration my simple mind and how things work once you get past the working area of class a most designs go to ab as a sort of soft clipping. Just a disclaimer I am not an engineer and I defer to those that I revere.
Regards, Elwood
I would also like to know that.....
Not taken into concideration that class A and AB amps often uses different topologies to achieve the best from the class they work in, then to my knowledge you can say that a class A is a overbiased class AB amp and a class AB is a underbiased class A.......
🙂
but not in the case where Re is chosen to make the optimal ClassAB bias equal to the ClassA bias current.
the two amps have the same ClassA current limit,
have the same voltage clip limit
have the same low impedance ClassAB current ability.
They are the same.
Neither is an overbiased or underbiased anything.
All that is different is that the ClassA emitter resistor can be selected from a range of values that suits the designer's philosophy. Aksa for instance is quite sure that Re=0r47 gives a very good sound quality performance.
Now, there's an experiment.
Build a 25W into 8r0 ClassAB with Re=0r068 and a ClassA with Re=0r47, both with the same PSU, the same 4pair output stage, biased to the same 1.29A and listen for the sound differences.
I wonder if they would measure the same. The slight loss in maximum output voltage using the higher Re should be easily measureable.
Hello
Bias current can not tell us if one audio amplifier works in class A or AB or B. It is not enough just that. I think yours make some mistake here.
To answer this, I posted the following two image files. Class AB is Class B with a higher bias current. So all.
Topological, Class B differs from Class A, as shown in pictures.
Best Regards
Bias current can not tell us if one audio amplifier works in class A or AB or B. It is not enough just that. I think yours make some mistake here.
To answer this, I posted the following two image files. Class AB is Class B with a higher bias current. So all.
Topological, Class B differs from Class A, as shown in pictures.
Best Regards
Attachments
Last edited:
I think no one would argue a difference between a class ab and a single ended class a amp. Is this discussion not about the sound difference between a class AB and a class A amp with complementary output pairs? The limited number of double ended class A amps I have seen look almost identical to class AB amps, thats why the originall question is there. I am new to class A designs so I end with a question. Other than bias level (and SOAs exc. that come with it), how are the topologies different between these latter2 amps?
cbdb,
How did you think to achieve diagram audio amplifier in Class A?
It's wrong when a diagram amplifier in Class B or AB increases bias current and said that the diagram works in class A. It 's all wrong.
In terms of SOA, you can not generalize. The best post a diagram amplifier and discussion of them, referring not only to the characteristics of the transistors used.
The sound set of amplifiers in class AB, A or B depends and how we analyze the question: workbench (or practical) or only in terms of auditory perception (which is different from person to person). It may be that a person will not notice the difference sound reproduction amplifiers that work in different operating classes.
Nice day and please excuse my english.
How did you think to achieve diagram audio amplifier in Class A?
It's wrong when a diagram amplifier in Class B or AB increases bias current and said that the diagram works in class A. It 's all wrong.
In terms of SOA, you can not generalize. The best post a diagram amplifier and discussion of them, referring not only to the characteristics of the transistors used.
The sound set of amplifiers in class AB, A or B depends and how we analyze the question: workbench (or practical) or only in terms of auditory perception (which is different from person to person). It may be that a person will not notice the difference sound reproduction amplifiers that work in different operating classes.
Nice day and please excuse my english.
bias current tells us all.
If we know the maximum output voltage and we know the ClassA resistance lower limit, then we know the maximum current that the load resistance will draw.
If all the output devices continue to actively supply that output current then the amplifier remains in ClassA.
The overriding requirement to achieve ClassA into a particular load resistance is the stage bias current.
If we were discussing single ended then Iout<Ibias. But we are not looking at single ended.
For push pull circuits Iout<2times Ibias.
There is some leeway in how much less than Ibias we need to be to maintain good linearity in the output. Too near the transistor Ic vs Vbe knee and linearity collapses.
Should that leeway (2*Ibias -Iout) be 1mA or 5mA or 50mA? Or does it depend on the topology adopted?
If we know the maximum output voltage and we know the ClassA resistance lower limit, then we know the maximum current that the load resistance will draw.
If all the output devices continue to actively supply that output current then the amplifier remains in ClassA.
The overriding requirement to achieve ClassA into a particular load resistance is the stage bias current.
If we were discussing single ended then Iout<Ibias. But we are not looking at single ended.
For push pull circuits Iout<2times Ibias.
There is some leeway in how much less than Ibias we need to be to maintain good linearity in the output. Too near the transistor Ic vs Vbe knee and linearity collapses.
Should that leeway (2*Ibias -Iout) be 1mA or 5mA or 50mA? Or does it depend on the topology adopted?
Last edited:
we should not need to clarify the ClassA definition.
But, some here consider that ensuring a device does not switch off and passes a near constant current meets the ClassA definition, (try that with a single ended ClassA amplifier and tell us it is still ClassA and sounds nice).
My slightly extended definition is my attempt to ensure that those miscreants can't screw up the discussion of ClassA vs ClassAB sound quality by entering the ClassAB enhancements (superA, AA, A+, sliding bias, etc) that are constantly being re-invented to improve ClassAB cheaply.
Last edited:
- Status
- Not open for further replies.