How to design a Power Amplifier


I have several electronics books from which I am presently reading the subject of power amplifiers. I noticed two different approaches.

In the first book, the author claims that a power amplifier is one that delivers the most possible power that the transistor is capable of to a load. So, he said the Q point of the amplifier should be set on the power curve. Given that the transistor cannot handle currents above \$I_{C(max)}\$ or collector emitter voltage above \$V_{CE(max)}\$, tangents can be drawn on the power curve from the these two extremes. These tangents are the ac load lines and all the points between the intersection of the two tangents with the power curve are the permissible Q points such that the transistor is driven to its utmost limit. Also these Q points will always be at the center of the ac load lines (tangents). This seems very clear and intuitive.

In another book, the approach is different. Here, the author begins with the dc load line and shows that the ac load line is different. Then he goes on to explain that to maximize the voltage and current swing the Q point has to be on the center of the ac load line, which again seems logical.

In the first approach the amplifier can be designed with the load in mind. By this I mean, if you know \$V_{CE(max)}\$ on the characteristic curve, you can draw the ac load line and immediately get the Q point. From the slope of the ac load line we can easily find the ac collector resistance and from there proceed with finding the biasing resistors.
In the second approach, locating the Q point seems to be very difficult because even if the Q point is at the center of the dc load line, it is very possible that the Q point is not at the center of the ac load line !

So here are my questions. Which method should I adopt in designing a power amplifier ? Also, suppose I chose the second approach how can I be sure that even if I manage to set the Q point at the center of the ac load line, it will be on the power curve ?

Best Answer

You're on the right track. I've only used the load line method for PAs with common source FETs, but the concept should be the same. It can be shown that the same load is valid for class A, class B, and others. Design for class A and mess with the bias later. Here are the basic steps...

Step 1: Select a transistor that can provide the desired output power

Step 2: Draw the load line using the max voltage and current of the transistor

Step 3: Calculate the necessary load from the load line

Step 4: While including the output parasitics of the transistor, add components to achieve the desired load