Six Amplifier Schematics

Amplifiers tend to use the following six configurations.

These schematics do not show coupling capacitors or the resistors used for biasing. They are simplified schematics to show where the inputs and outputs are. Each configuration features high or low impedance and voltage or current gain.

Three configurations are popular.

Above are the common emitter and emitter follower.

Above is the differential amplifier. It has amazing properties for precision and bandwidth.

Three configurations are special purpose.

Above are the common base and cascode.

Above is the Darlington circuit for super current gain, but mainly for supplies and audio.

To flesh out the basic schematics above, we have to get into the complex topic of biasing the amplifier.

The drawing below develops this. But first, some notes about biasing.

• The name, bias, dates from the 1920s when it was used for vacuum tubes. Amplifier bias has nothing to do with opinions or attitudes.

• Bias is hard to talk about because it can only be seen on a meter or oscilloscope. Bias current is extra hard to talk about because we can scarcely measure currents.

• Bias is a DC thing that affects AC waveforms, and this bridge between DC to AC makes it more complex to think about.

• The best way to start learning about bias is to take a working audio amplifier (one or two transistors) with speaker output and adjust the bias so it distorts the output. This is the approach in the following drawing.

• Understanding bias comes a little at a time. Full understanding comes from seeing bias for other types of amplifiers (other than common emitter) and when you can bias your own circuits.

• Bias is accomplished by adding "bias parts," resistors and capacitors.

• Bias is only needed with "linear" amplifiers. It is not needed for logic circuits.

• Bias of a common-emitter amplifier (following diagram) is complicated by the inversion on the collector. To make full sense, you have to have two things working in your mind, the bias feature and the collector inversion.

This presentation leads the reader to recognize bad bias but doesn't really give a tutorial about optimum biasing.

Let's look at the drawing and see three regions. 1) Circuit in upper left. 2) Distortion in upper right. 3) Logical thinking about voltages, steps A through D.

Let's start with the circuit at upper left. This is the most basic audio amplifier, drawn with biasing parts. Note where input goes in and output comes out. Note that the input is weak and the output is strong, and they are the same frequency. Cb, the input cap, lets AC come in to the transistor base without disrupting bias.

The bias point of the base is 4.5V in this example, as noted on the schematic.

The upper right shows what an oscilloscope shows when bias is wrong, such as when the base voltage is 7V or 2V rather than 4.5V. Don't worry about saturation or cutoff at this point. Just know that the four distorted waveforms, which have flat spots, sound bad in a speaker. You need to know that putting too much AC voltage into any amplifier will cause flat spots (distortion) in even the best-biased amplifier. "Overdriving" is when too much AC is going in, and you need to turn down the volume control.

The best bias is so that the output waveform can be driven to the largest amplitude, with flat spots developing at top and bottom at the same moment. This sentence is the key to this presentation.

If bias is bad but not too bad, you can use a volume control to get a small, undistorted output, but when you dial up the input AC amplitude, the output will become distorted at a disappointingly low output amplitude.

Now, let's talk through steps A through D in the drawing. A shows a "voltage divider" delivering 4.5V to the base when R1 and R2 are the same, 4.5V being half of 9V.

B shows the base-to-emitter, "Vbe" drop of .7V, namely 4.5V - .7 = 3.8V. All silicon NPN or PNP transistors have this .7V drop.

C shows the 3.8V applied to the 1k resistor, and the calculation of the current, I, that flows in the emitter. 3.8/1000 = 3.8 milliamp.

D shows Rc and how the 3.8mA drops collector voltage from 9V to 7.2V. If you put a meter onto the collector, it reads 7.2V.

A fact: D at 7.2 means that the peak of the output can get to 9-7.2=1.8V. That makes peak-to-peak able to get to 3.6V. Any stronger signal will get flat spots. It is possible to redesign the circuit to get, like, 75% of 9V to be the peak-to-peak output, but that is beyond the scope of this presentation. (But the idea is to adjust R1 higher so the emitter is 1V, leaving an 8V peak-to-peak swing available on output.)

An extension of this presentation, not written up here, is how amplifier inputs are high impedance, like 33k here, and outputs are low impendance, like 470 ohms here. This is necessary so that lower-impedance loads (speakers) can be driven, and to avoid loss of gain due to a voltage-divider effect on the amplifier output. Successive stages of gain will have lower and lower resistors as you get close to the speaker.

Another extension is "gain budget," where you might need to amplify .5mV from a mic to 30V on a speaker. You decide how many gain stages there need to be and how much voltage gain you need in each stage.

Another extension is the function of Ce and R4. R4 is to give gain. Without C3 and R4, gain is 470/1000, just .47, which is loss, not gain. With Ce and R4, gain=470/R4, and R4 might be 24 ohms.

Biasing turns out to be complex. If you get in the saddle with biasing, input and output impedance, and the six amplifier topologies at the top of this page, you are most of the way through audio amplifier design. But these three items are a lot! After these three, you have to get into controlling feedback through the supply, managing emitter-follower oscillation, choosing supplies or designing to particular, "given," supplies, using power transistors and heat sinks to give high output power, designing in fuses and overload protection to prevent fire, and using common test equipment to diagnose problems.

An electrical engineering student who is pointed toward logic design and computers won't learn much of this at all. But EEs who want general EE knowledge will get to where they understand most of this. A good grounding in linear amplifiers amounts to learning a new language. Once you know the language, you can actually think with this language, the same that mathematicians think in terms of math language and accountants think in terms of law, databases, and spreadsheets.