Why is the following simulation not working?

Question

The following is a multistage amplifier. It consists of 3 parts:

1. Waveform generator and inverting amplifier
2. A small signal amplifier with gain 20 and a small signal amplifier with gain 10
3. A power amplifier where efficiency = 0.6 * maximum efficiency

When I do not connect the power amplifier, the output waveform on the second small signal amplifier is as expected, that is, 10 V. But, when I connect the power amplifier to its output, the output waveform on the second small signal amplifier decreases to mV. Why is it so? How can I correct it?

Answer 1

First of all, the 0.2µF capacitor at the output of the second gain stage (is that Co2? There's a line running through the ref des, making it hard to read) is completely redundant, and should be removed.

Second, you have a huge bias current running through the output transistors as a result of how their bases are biased at ±1/3 the supply voltage (three 1KΩ resistors in series). This reduces the input impedance of that stage even more than you might expect from the bias network alone.

You need a driver for the output stage that can supply the current that it needs; the second JFET isn't up to the task.

Answer 2

The input impedance of the power-amp is no more than 500 ohms due to the 1k resistors biasing the final output transistors. The input impedance could be a little less given that the power o/p transistors will also draw current through their respective bases in order to drive the 60R load.

The gain of the stage that drives the PA is largely determined by the 6k8 (or is it 8k8) in the drain of Q2 - this gain is being dramatically killed by the PA's input impedance - probably by a factor of at least 6.8/0.5 = 13.6.

This is why your 10V signal is now in the milli-volt region - you need an emitter follower buffer amp - that's what I'd try anyway.