Which class A output stage is preferred?

[diarmuid]

Hello All,

Say I want to drive a large capacitive load with a buffer.

Say the buffer is a 2 stage op-amp and I am restricted to using only a class A output stage (for arguments sake).

Being that this leaves me with the choice of either a common-source amplifier or source-follower output stage, which would be the most effective at driving a high capacitive load?

Would both be useless or is there any argument for choosing one (or both) over a class AB output stage?

Thanks,

Diarmuid

 

[BradtheRad]

Class A operation has a resistor in the transistor's opposite leg from where the load is. This impedance is in the way during one half of a cycle. Its net effect is to limit current through the output stage.

The AB (push-pull) method has transistors turning on and off, to allow low impedance to the supply rails. More current is available to a heavy load.

 

[Junus2012]

use common source because it has better output swing than the source follower, although the C.S has more output impedance than S.F but this is not your interest since you are driving only capacitive load

 

[diarmuid]

Thanks Senan. Thats what I was thinking. I just saw such a situation where a source follower was used as an output stage to drive a capacitive load. Right enough the output swing is limited which does cause problems for the application.

Cant think of any good reason to use a source follower!!!

 

[Junus2012]

you are welcome

the good reason for using the S.F is it has a lower output resistance which will be useful if you drive a small resistive load, the S.F also can be used as a D.C level shifter
Cant think of any good reason to use a source follower!!![/QUOTE]

 

[diarmuid]

Would this perhaps be an argument in favour of the source follower:

Charging up a large capacitive load would require a large peak current. With the common source amp, its peak current will be its bias current (what makes it class A). However, with the source follower, its peak current is only limited by the size of the output transistor.

Therefore, in theory you could supply a larger peak current for a given bias current with the source follower than with the common source amp. This would mean you would charge your output capacitive load quicker and with less power using the source follower over the common source amp.

Does this make sense?

If so, is it correct to view a source follower as a class A amplifier? By definition a class A amp is one whose peak current is equal to its bias current which (from above I think) need not necessarily be the case for a source follower.

 

[Junus2012]

S.F can work as class A or AB, what zou was talking about is class AB mode where the current is limited bz the size of the transistor, but if you connect the S.F as class A then it will perform exactly like C.S class A where the maximum current is defined by the biasing current

Would this perhaps be an argument in favour of the source follower:

Charging up a large capacitive load would require a large peak current. With the common source amp, its peak current will be its bias current (what makes it class A). However, with the source follower, its peak current is only limited by the size of the output transistor.

Therefore, in theory you could supply a larger peak current for a given bias current with the source follower than with the common source amp. This would mean you would charge your output capacitive load quicker and with less power using the source follower over the common source amp.

Does this make sense?

If so, is it correct to view a source follower as a class A amplifier? By definition a class A amp is one whose peak current is equal to its bias current which (from above I think) need not necessarily be the case for a source follower.

 

[FvM]

The information to decide about a suitable output stage hasn't been yet given, I think. Besides high capacitive load, voltage swing, source and sink current and bandwidth requirements should be considered.

My first guess would be a common source class A stage with the output capacitor forming the dominant pole.

 

[diarmuid]

I see. Given that a wide output voltage swing is required, the cs amp seems the preferred option.

Thanks guys for your help.

All the best,

Diarmuid

- - - Updated - - -

Just one other thing thats bothering me:

Say I have a simple cs amp: driver is the NFET and bias is the PFET. Applied to the driver is a sine wave.

For the positive half of the sine wave the NFET sinks current from the load and for the negative half of the sine
wave the PFET sources current to the load.

The current sourced to the load will be defined as it comes from the biasing PFET. However the current which is
sinked from the load depends on the current carrying capability of the NFET. Thus source and sink currents will
not be equal.

Does this lead to distortion of the output signal? For example: positive half of the output signal is more slugish
than the negative half?

Thanks,

 

[customclothingNZ]

In this case, the source and sink currents should be equal. The output signal might be affected if not.