|
|
| Elliott Sound Products | Amplifier Basics - How Amps Work (Intro) |
 Main Index
Articles Index
|
The term 'amplifier' is somewhat 'all-encompassing', and is often thought (by many users in particular) to mean a power amplifier for driving loudspeakers. This is not the case (well, it is, but it is not the only case), and this article will attempt to explain some of the basics of amplification - what it means and how it is achieved. This article is not intended for the designer (although designers are more than welcome to read it if they wish), and is not meant to cover all possibilities. It is a primer, and gives fairly basic explanations (although some will no doubt dispute this) of each of the major points.
I will explain the basic amplifying elements, namely valves (vacuum tubes), bipolar transistors and FETs, all of which work towards the same end, but do it differently. This article is based on the principles of audio amplification - radio frequency (RF) amplifiers are designed differently because of the special requirements when working with high frequencies.
Not to be left out, the opamp is also featured, because although it is not a single 'component' in the strict sense, it is now accepted as a building block in its own right.
This article is not intended for the complete novice (although they, too, are more than welcome), but for the intermediate electronics or audio enthusiast, who will have the most to gain from the explanations given.
Before we continue, I must explain some of the terms that are used. Without knowledge of these, you will be unable to follow the discussion that follows.
| Electrical Units | |||
| Name | Measurement of | aka | Symbol |
| Volt | electrical 'pressure' | voltage | V, U, E (EMF) |
| Ampere | the flow of electrons | current | A, I |
| Watt | power | W, P | |
| Ohm | resistance to current flow | Ω, R | |
| Ohm | impedance, reactance | Ω, Z, X | |
| Farad | capacitance | F, C | |
| Henry | inductance | H, L | |
| Hertz | frequency | Hz | |
Note: 'aka' means 'Also Known As'. Although the Greek letter omega (Ω) is the symbol for Ohms, I often use the word Ohm or the letter 'R' to denote Ohms. Any resistance of greater than 1,000 Ohms will be shown as (for example) 1k5, meaning 1,500 Ohms, or 1M for 1,000,000 Ohms. The second symbol shown in the table is that commonly used in a formula.
When it comes to Volts and Amperes (Amps), we have alternating current and direct current (AC and DC respectively). The power from a wall outlet is AC, as is the output from a CD or tape machine. The mains from the wall outlet is at a high voltage and is capable of high current, and is used to power the amplifying circuits. The signal from your audio source is at a low voltage and can supply only a small current, and must be amplified so that it can drive a loudspeaker.
Impedance
A derived unit of resistance, capacitance and inductance in combination is called impedance, although it is not a requirement that all three be included. Impedance is also measured in Ohms, but is a complex figure, and often fails completely to give you any useful information. The impedance of a speaker is a case in point. Although the brochure may state that a speaker has an impedance of 8Ω, in reality it will vary depending on frequency, the type of enclosure, and even nearby walls or furnishings.
Units
In all areas of electronics, there are smaller and larger amounts of many things that would be very inconvenient to have to write in full. For example, a capacitor might have a value of 0.000001F or a resistor a value of 150,000Ω. Because of this, there are conventional units that are applied to make our lives easier (well, once we are used to using them, anyway). It is much easier to say 1uF or 150k (the same as above, but using standard units). These units are described below.
| Conventional Metric Units | ||
| Symbol | Name | Multiplication |
| p | pico | 1 x 10-12 |
| n | nano | 1 x 10-9 |
| μ | micro | 1 x 10-6 |
| m | milli | 1 x 10-3 |
| k | kilo | 1 x 103 |
| M | Mega | 1 x 106 |
| G | Giga | 1 x 109 |
| T | Tera | 1 x 1012 |
Although commonly written as the letter 'u', the symbol for micro is actually the Greek letter mu (μ) as shown. In audio, Giga and Tera are not commonly found (not at all so far - except for specifying the input impedance of some opamps!). There are also others (such as femto - 1x10-15) that are extremely rare and were not included. Of the standard electrical units, only the Farad is so large that the defacto standard is the microfarad (µF). Most of the others are reasonably sensible in their basic form.
It is important to understand that the symbol for microfarad is µF (or uF), not mF - that's a millifarad, and is 1,000 µF.
The term 'amplify' basically means to make stronger. The strength of a signal (in terms of voltage) is referred to as amplitude, but there is no equivalent for current (curritude?, nah, sounds silly). This in itself is confusing, because although 'amplitude' refers to voltage, it contains the word 'amp', as in ampere. Maybe we should introduce 'voltitude' - No? Just live with it.
To understand how any amplifier works, you need to understand the two major types of amplification, and a third 'derived' type:
In the case of a voltage amplifier, a small input voltage will be increased, so that for example a 10mV (0.01V) input signal might be amplified so that the output is 1 Volt. This represents a 'gain' of 100 - the output voltage is 100 times as great as the input voltage. This is called the voltage gain of the amplifier.
In the case of a current amplifier, an input current of 10mA (0.01A) might be amplified to give an output of 1A. Again, this is a gain of 100, and is the current gain of the amplifier.
If we now combine the two amplifiers, then calculate the input power and the output power, we will measure the power gain:
| P = V × I | (where I = current, note that the symbol changes in a formula) |
The input and output power can now be calculated:
| Pin = 0.01 × 0.01 | (0.01V and 0.01A, or 10mV and 10mA) | |
| Pin = 100µW | ||
| Pout = 1 × 1 | (1V and 1A) | |
| Pout = 1W |
The power gain is therefore 10,000, which is the voltage gain multiplied by the current gain. Somewhat surprisingly perhaps, we are not interested in power gain with audio amplifiers. There are good reasons for this, as shall be explained in the remainder of this page. Having said this, in reality all amplifiers are power amplifiers, since a voltage cannot exist without power unless the impedance is infinite or zero. This is never achieved, so some power is always present. It is convenient to classify amplifiers as above, and no harm is done by the small error of terminology.
Note that a voltage or current gain of 100 is 40dB, and a power gain of 10,000 is also 40dB.
Input Impedance
Amplifiers will be quoted as having a specific input impedance. This only tells us the load it will place on preceding equipment, such as a preamplifier. It is neither practical nor useful to match the impedance of a preamp to a power amp, or a power amp to a speaker. This will be discussed in more detail later in this article.
The load is that resistance or impedance placed on the output of an amplifier. In the case of a power amplifier, the load is most commonly a loudspeaker. Any load will require that the source (the preceding amplifier) is capable of providing it with sufficient voltage and current to be able to perform its task. In the case of a speaker, the power amplifier must be capable of providing a voltage and current sufficient to cause the speaker cone(s) to move the distance required. This movement is converted to sound by the speaker.
Even though an amplifier might be able to make the voltage great enough to drive a speaker cone, it will be unable to do so if it cannot provide enough current. This has nothing to do with its output impedance. An amplifier can have a very low output impedance, but only be capable of a small current (an operational amplifier, or opamp is a case in point). This is very important, and needs to be fully understood before you will be able to fully appreciate the complexity of the amplification process.
Output Impedance
The output impedance (Zout) of an amplifier is a measure of the impedance or resistance 'looking' back into the amplifier. It has nothing to do with the actual loading that may be placed at the output.
For example, an amplifier has an output impedance of 10Ω. This is verified by placing a load of 10Ω across the output, and the voltage can be seen to decrease to ½ that with no load. However, unless this amplifier is capable of substantial output current, we might have to make this measurement at a very low output voltage, or the amplifier will be unable to drive the load. If the output clips (distorts) the measurement is invalid.
Another amplifier might have an output impedance of 100Ω, but be capable of driving 10A into the load. Output impedance and current are completely separate, and must not be seen to be in any way equivalent. Both of these possibilities will be demonstrated later in this series.
It is very rare that you will ever be able to perform a direct measurement of output impedance. An opamp configured for a gain of 10 (20dB) will usually have such a low Zout that it's almost impossible to measure it directly, other than by using an input level of a few microvolts. Most power amps will be stressed badly by attempting to drive close to a short circuit, and will show their displeasure by blowing up or triggering their protection circuits (if fitted).
The output impedance is also independent of the power supply impedance. This causes the maximum undistorted power to fall with lower impedance loads, so an amp may be able to deliver 50W into 8Ω but only 80W into 4Ω (continuous power - peak power can be higher for short transients). Failure to understand that all of these factors are independent from each other will lead to false conclusions. It's easy to fall into the traps, and some manufacturers make this worse by claiming that their 'XyZ-5000' 50W amplifier can deliver 100 amps to the load, but fail to tell buyers that no sensible (or even non-sensible) load can ever draw that much current.
The output impedance is (roughly) equal to the open-loop (zero feedback) output impedance, divided by the feedback ratio. An amplifier may have an open-loop Zout of 5Ω, with 46dB of feedback (a factor of 200). Closed-loop Zout is then 5 / 200, or 25mΩ. However, the feedback ratio is almost always frequency dependent, so unless the frequency is specified, the Zout figure may not be meaningful.
Feedback
Feedback is a term that creates more and bloodier battles between audio enthusiasts than almost any other. Without it, we would not have the levels of performance we enjoy today, and many amplifier types would be unlistenable without it.
Feedback in its broadest sense means that a certain amount of the output signal is 'fed back' into the input. An amplifier - or an element of an amplifying device - is presented with the input signal, and compares it to a 'small scale replica' of the output signal. If there is any difference, the amp corrects this, and ideally ensures that the output is an exact replica of the input, but with a greater amplitude. Feedback may be as a voltage or current, and has a similar effect in either case.
In many designs, one part of the complete amplifier circuit (usually the input stage) acts as an error amplifier, and supplies exactly the right amount of signal (with correction as needed) to the rest of the amp to ensure that there is no difference between the input and output signals, other than amplitude. This is (of course) an ideal state, and is never achieved in practice. There will always be some difference, however slight. Note that any amplifier that suffers from crossover (aka notch) distortion cannot be made linear with feedback, because at zero output (where this distortion occurs) there is also (almost) zero gain. You can't have feedback unless there is some 'excess' gain!
Signal Inversion
When used as voltage amplifiers, all the standard active devices invert the signal. This means that if a positive-going signal goes in, it emerges as a larger - but now negative-going - signal. This does not actually matter for the most part, but it is convenient (and conventional) to try to make amplifiers non-inverting. To achieve this, two stages must be used (or a transformer) to make the phase of the amplified signal the same as the input signal.
The exact mechanism as to how and why this happens will be explained as we go along.
Design Phase
The design phase of an amplifier is not remarkably different, regardless of the type of components used in the design itself. There is a sequence that will be used most of the time to finalise the design, and this will (or should) follow a pattern.
200 / 50 = 4mA.Since the Class-A driver must operate in Class-A (what a surprise), it will need to operate with a current of 1.5 to 5 times the expected maximum driver current, to ensure that it never turns off. The same applies with a MOSFET amp that will expect (for example) a maximum gate capacitance charge (or discharge) current of 4mA at the highest amplitudes and frequencies.
These are only guidelines (of course), and there are many cases where currents are greater (or smaller) than suggested. The end result is in the performance of the amp, and the textbook approach is not always going to give the expected result. Note that there are some essential simplifications in the above - it is an overview, and is only intended to give you the basic idea.
For the purposes of this article, there are three different types of amplifying devices, and each will be discussed in turn. Each has its strengths and weaknesses, but all have one common failing - they are not perfect.
A perfect amplifier or other device (known generally as 'ideal') will perform its task within certain set limits, without adding or subtracting anything from the original signal. No ideal amplifying device exists, and as a result, no ideal amplifier exists, since all must be built with real-life (non-ideal) devices.
The amplifying devices currently available are:
There are also some derivatives of the above, such as Insulated Gate Bipolar Transistors (IGBT), and Metal Oxide Semiconductor Field Effect Transistors (MOSFET). Of these, the MOSFET is a popular choice among many designers due to some desirable characteristics, and these will be covered in their own section.
All of these devices are reliant on other non-amplifying ('support') components, commonly known as passive components. The passive devices are resistors, capacitors and inductors, and without these, we would be unable to build amplifiers at all.
All the devices we use for amplification have a variable current output, and it is only the way that they are used that allows us to create a voltage amplifier. Valves and FETs are voltage controlled devices, meaning that the output current is determined by a voltage, and no current is drawn from the signal source (in theory). Bipolar transistors are current controlled, so the output current is determined by the input current. This means that no voltage is required from the signal source, only current. Again, this is in theory, and it is not realisable in practice.
Only by using the support components can we convert the current output of any of these amplifying devices into a voltage. The most commonly used for this purpose is a resistor.
All active devices have certain parameters in common (although they will have different naming conventions depending on the device). Essentially these are ...
This is by no means all of the ratings, there are many more, and vary from device to device. Some MOSFETs for example will have Peak Current ratings, which will be many times the continuous rating, but only for very limited time. Bipolar transistors have a Safe Operating Area (SOA) graph, which indicates that in some circumstances you must not operate the device anywhere near its maximum power dissipation, or it will fail due to a phenomenon called second breakdown (described later).
With most semiconductors, in many cases it will not be possible to operate them at anywhere near the maximum power dissipation, because thermal resistance is such that the heat simply cannot be removed from the junction and into the heatsink fast enough. In these cases, it might be necessary to use multiple devices to achieve the performance that can (theoretically) be obtained from a single component. This is very common in audio amplifiers.
There are some things that you just can't get away from, and maths is one of them. (Sorry.) I will only include the essentials here, but will describe any others that are needed as we go. I am not about to give a lesson in algebra, but the best reason for ever doing the subject is to learn how to transpose electronics formulae ! Transposition is up to you (unless I am forced to do it for a calculation here or there).
Ohm's Law
The first of these is Ohm's Law, which states that a voltage of 1V across a resistance of 1 Ohm will cause a current of 1 Amp to flow. The formula is ...
R = V / I (where R = resistance in Ohms, V = Voltage in Volts, and I = current in Amps)
Like all such formulae, this can be transposed (oops, I said I wasn't going to do this, didn't I).
V = R × I (× means multiplied by), and
I = V / R
Reactance
Then there is the impedance (reactance) of a capacitor, which varies inversely with frequency (as frequency is increased, the reactance falls and vice versa).
Xc = 1 / ( 2π × f × C )
where Xc is capacitive reactance in Ohms, π (pi) is 3.14159, f is frequency in Hz, and C is capacitance in Farads.
Inductive reactance, being the reactance of an inductor. This is proportional to frequency.
Xl = 2π × f × L
where Xl is inductive reactance in Ohms, and L is inductance in Henrys (others as above).
Frequency
There are many different calculations for this, depending on the combination of components. The -3dB frequency for resistance and capacitance (the most common in amplifier design) is determined by ...
fo = 1 / ( 2π × R × C ) where fo is the -3dB frequency
When resistance and inductance are combined, the formula is
fo = R / (2π × L)
Power
Power is a measure of work, which can be either physical work (moving a speaker cone) or thermal work - heat. Power in any form where voltage, current and resistance are present can be calculated by a number of means:
P = V × I
P = V² / R
P = I² × R
where P is power in watts, V is voltage in Volts, and I is current in Amps.
Decibels (dB)
It has been known for a very long time that human ears cannot resolve very small differences in sound pressure. Originally, it was determined that the smallest variation that is audible is 1dB - 1 decibel, or 1/10 of 1 Bel. It seems fairly commonly accepted that the actual limit is about 0.5dB, but it is not uncommon to hear that some people can (or genuinely believe they can) resolve much smaller variations. I shall not be distracted by this!
dB = 20 × log ( V1 / V2 )
dB = 20 × log ( I1 / I2 )
dB = 10 × log ( P1 / P2 )
As can be seen, dB calculations for voltage and current use 20 times the log (base 10) of the larger unit divided by the smaller unit. With power, a multiplication of 10 is used. Either way, a drop of 3dB represents half the power and vice versa.
There are many others, but these will be sufficient for now. I do not intend this to be a complete electronics course, so I will give you that which is needed to understand the remainder of the article - for the rest, there are lots of excellent books on electronics, and these will have every formula you ever wanted.
| Main Index
Articles Index
|