What is the role of biasing in a Class A BJT amplifier?

[b-rett]

Can someone clarify the overall purpose and implications of biasing a BJT Class A amplifier? I just can't get my head wrapped around the entire purpose.

As far as I know, biasing the circuit will place it in a certain operating region, but I don't know the limits/implications of doing this.

Example from book: "Consider the operation of the emitter follower circuit for sine-wave input. Neglecting V_CEsat we see that if the bias current I is properly selected, the output voltage can swing from -V_CC to +V_CC with the quiescent value being zero."

What constitues a "properly selected" bias current I and why does this gave way to the output being able to swing between -V_CC and +V_CC?

 

[vk6kro]

Have a look at this circuit and graph:
http://dl.dropbox.com/u/4222062/Voltage%20amplifier.PNG

You can see that if the input voltage is gradually increased, the output stays the same (12 V) for a while then starts to drop until the output is zero.

Quite a small change in voltage (0.6 V to 1.1 V) caused the change in output from 12 volts to zero. So, it works as an amplifier.

But what if you had a sinewave you wanted to amplify?
Sinewaves swing positive and negative.
If you just used a sinewave at the input, you would get output variation only for the small input range of +0.6 volts to +1.1 volts. All the negative part of the sinewave would be lost and a lot of the positive half. The output would be very distorted.

The best way to do this would be to set the zero point of the sinewave at about 0.85 volts and then vary the sinewave around this level. This is biasing.

Suppose you did this and varied the sinewave from 0.7 v to 1 v. You can see that the output would vary from about 10 volts to about 2 volts and it would be an undistorted sinewave but amplified to be bigger than the input.

This is all that biasing does. It tries to make sure that the output will be an undistorted version of the input.

 

[b-rett]

makes sense now.

i see biasing often done with current sources usually at the emitter of the npn. is there a way to easily choose this bias current to design for a specific transfer function with respect to changing input voltage, or does it require a small signal model and calculating the overall gain and manipulating impedance for pole/zero?

 

[vk6kro]

i see biasing often done with current sources usually at the emitter of the npn. is there a way to easily choose this bias current to design for a specific transfer function with respect to changing input voltage, or does it require a small signal model and calculating the overall gain and manipulating impedance for pole/zero?

The gain of an amplifier is not normally set by the bias current. It is a function of the characteristics of the transistor (which you have little control over) and the load resistance.

The bias current is usually dictated by these factors.

For example, if the load resistor has already been chosen, the bias current is selected to give a collector to emitter voltage of about half of the supply voltage. This way, the output voltage can swing equally in each direction.

i see biasing often done with current sources usually at the emitter of the npn.
I don't think I have seen this. Could you give an example?

 

[b-rett]

ex this amplifier which is biased by I from Q2.

 

[vk6kro]

Yes, I have seen that circuit, but it is very rarely used.

Usually a resistor works OK.

 

[Adjuster]

Yes, I have seen that circuit, but it is very rarely used.

Usually a resistor works OK.

That's certainly true for discrete circuits, where resistors are usually cheaper than transistors, and possibly more reliable.

By contrast, current source biasing is more commonly used within Integrated Circuits. This is partly because moderate to large value integrated resistors are relatively bulky compared to transistors. It is also possible to obtain a number of related bias currents by a technique called current mirroring, which can allow useful features like adjustable speed/power trade-offs.

 

[vk6kro]

This circuit is really more of a substitute load than a biasing component.

You still have to arrange biasing to Q1's base.