AMPLIFIERS
Block Diagram
Figure 3-8. Block diagram of an amplifier.
This next device, shown in Figure 3-8, is called an amplifier, which makes signals bigger. RF signals constantly need to be made bigger as they move from place to place. It's just like driving a car. You drive around from place to place using up gas, and when you run low, you fill up. An amplifier is a filling station for RF signals. They move around from place to place (either in the air or along a conductor) and when they need a boost, hopefully there is an amplifier around. Visually, a small signal enters the large end of the block diagram (left side) and leaves as a large signal from the pointy end.
Fundamental Properties of Amplifiers
Gain
There are three fundamental properties of all amplifiers: gain, noise or power, and linearity. Gain is a measure of how much bigger the output signal is than the input signal, and is measured in, as you may have guessed, dB. Some places in an RF system need a lot of gain (40 or 50 dB), and some places only need a little (5-10 dB).
Amplifiers fall into three main categories: low noise, high power, and "other." A low noise amplifier is the very first amplifier a signal encounters after it comes through the antenna in a receiver. High power amplifiers are the last amplifier a signals goes through before it flies out the antenna in a transmitter. Every other amplifier is an "other."
Noise Figure
Low noise amplifiers (LNA) listen for very small RF signals so they must be vewy, vewy, qwiet. A measure of an LNA's quietness is called noise figure (NF) and is measured in, of course, dB. The fundamental property of an LNA is noise figure. The lower the NF of an LNA, the better. Some RF engineers pay big bucks for an LNA with super low NF. (The rest of us just invest in mutual funds.) What is a good NF? It depends. Think of it this way: the lower the noise figure, the smaller the signal which the LNA can hear, the further away the LNA can be, and thus the greater the range, of say, a cellular phone.
Output Power
High power amplifiers (HPA) boost the RF signal as big as possible (or allowable) just before it is shot out of the antenna. The bigger the signal, the farther it travels and the greater the range, of say, a cellular phone. The second fundamental property of an HPA is output power, measured in watts. Generally speaking, the higher the power, the better.
Unfortunately, RF engineers insist on making things difficult and tend to express output power in dBm. What the heck is dBm you ask? It literally means "dB above one milliwatt." For instance, 10 dBm is a signal 10 dB above (or bigger than) 1 mW, 30 dBm is 30 dB bigger than 1 mW, and so on. A 30 dBm signal, which is 30 dB (or 1000 times) bigger than 1 mW, translates to one watt (1000 x 1 mWatt). So 30 dBm is the same as 1 watt. Damn engineers. See Table 3-1 for some common conversions.
| Power in Watts | Power in dBm |
|---|---|
| 0.1 mW | -10 dBm |
| 1 mW | 0 dBm |
| 1 watt | 30 dBm |
| 1000 watts | 60 dBm |
Figure 3-9a shows an example of an LNA and Figure 3-9b shows an exaggerated example of an HPA. The two were chosen to demonstrate how much bigger HPAs can be compared to LNAs. This situation arises out of the need to remove heat from the HPA. Because amplifiers are not 100% efficient, some of the energy that goes into the amplifier comes out as an RF signal, but the remainder comes out as heat, which is why many HPAs contain fans to provide internal cooling, like the fan in a car's engine. If the HPA's fans stop working, you might as well get out the marshmallows.
When discussing the output power of an HPA you need to know that everything is relative. For instance, while the output amplifier in most cellular phones puts out less than 1 watt, the output amplifier at the basestation end puts out 50 watts and both amplifiers are referred to as HPAs. Just because it is referred to as an HPA does not mean it puts out a lot of power.
Figure 3-9a. An LNA. Courtesy of Mini-Circuits
Figure 3-9b. An HPA. Courtesy of Amplifier Research
Linearity
In this age of digital communications there is a third fundamental property of amplifiers called linearity. One of the implications of digital wireless communications is that when a digital signal rides on top of an RF carrier, any amplifier which the signal goes through must be really linear. Linearity is a measure of how much the amplifier distorts the shape of the signal. As you will learn in the future section on modulation, tiny changes to the shape of the RF carrier (sine wave) actually contain information; therefore, unwanted changes in the signal's shape serve only to distort the information. What RF engineers want at the output of their amplifiers is a signal which is bigger than, but identical in shape to, the input signal. As a way of visual introduction to the concept of linearity, I present you with the single most important piece of information about any amplifier, the transfer curve (see Figure 3-10). A transfer curve is a graph of the output power versus the input power of an amplifier. All amplifiers display this type of behavior.
Figure 3-10. Output power versus input power of an amplifier.
Referring to Figure 3-10, as the input power to an amplifier increases (moving to the right on the horizontal axis), the output power from the amplifier increases by a like amount, at least until the point marked "A." Everything up to point A is known as the linear region of an amplifier. It is in this region which the amplifier must operate if it is to avoid distorting the RF signal. The output power at point A is referred to as the P1dB (pronounced p wun' d b) point or the P 1dB power or the one dB compression point. P1dB has an exact definition, but it will only cause confusion. It is much simpler to think of the P1dB point of an amplifier as the highest power the amplifier can put out and still be in the linear region. Or another way of viewing it, P1dB is the highest linear power an amplifier can put out.
Beyond point A on the horizontal axis (increasing the input power further), notice that the output power no longer rises, but stays flat. In other words, an increase in input power no longer results in an increase in output power. The amplifier stops amplifying. After point A, the amplifier is said to be in saturation and enters the non-linear region. It is in the non-linear region where all the signal distortion occurs and messes up a cellular phone call. Output power greater than P1dB is referred to as saturated output power. In some instances, saturated output power is useful, but not in digital wireless communications. Just as a note: if the input power is increased further, eventually a point is reached where the output power of the amplifier increases rapidly, as shown. Of course at this point the output power is in the form of flames shooting out of the amplifier, which is why it is referred to (jokingly) as the 911 point.
One method of measuring an amplifier's linearity is by its intercept point (which is often referred to as the third order intercept point). The higher the intercept point, the more linear the amplifier. The intercept point is represented by the symbol Ip3 (pronounced i' pee three). (Sometimes an amplifier's Ip3 is referred to as its dynamic range.) Like power, the intercept point is also measured in dBm. An amplifier with a 40 dBm intercept point is more linear than an amplifier with a 30 dBm intercept point. A rule of thumb used by all RF engineers is that an amplifier's Ip3 is 10 dB greater than its P1dB point. There is no need to go any further.
Did You Know?
The third order intercept point is actually a mythical point that does not really exist and cannot be measured directly. Instead, a bunch of other measurements are made and then the Ip3 of an amplifier is calculated using these other measurements. Leave it to RF engineers to base the performance of this critical component on something that doesn't even exist.
How Amplifiers Work
How an amplifier supplies gain to an input signal is an interesting process. The input signal itself does not actually get bigger per se as it moves through the amplifier. Instead, the input RF signal acts to control another type of power called DC power, where DC stands for direct current. (DC power or DC voltage, unlike a sine wave, does not vary with time, but is constant, just as the voltage from the battery in a flashlight.) Similar to turning the tires of an automobile, as you drive you aren't really turning the tires themselves, you are controlling the tires' movement with a controller called the steering wheel. If the automobile were an amplifier, the driver would be the input RF signal, the tires would be the DC power, and the steering wheel would be the controller called a transistor, which you will learn about in a later chapter. In an amplifier, the RF input signal tells the transistor to "shape" the DC power to exactly reflect the shape of the input signal. In this way, the output signal has the same exact shape as the input signal, only bigger. How much bigger? It depends on how much DC input power there is. The main difference between HPAs and every other amplifier is that HPAs have greater DC input power.
Special Amplifiers
Limiting Amplifiers
There are two special cases of amplifiers which you should be aware of if you really want to impress an RF engineer. The first type is called a limiting amplifier and, as the name implies, it limits the output power. This type of amplifier is used in places where the component which follows it will be damaged if its input power is too high. The limiting amplifier provides a sort of protection for the next component. You may recall from Figure 3-10 that all amplifiers behave somewhat as limiting amplifiers. At a certain input power, the output power levels off. Basically, the only difference between a limiting amplifier and any other amplifier is that limiting amplifiers do not blow up at the 911 point—in theory.
Balanced Amplifiers
The other type of amplifier is called a balanced amplifier. It is not so much a different amplifier as it is a different amplifier design. In a balanced amplifier, there are two amplifiers in parallel (see Figure 3-11).
Figure 3-11. Block diagram of a balanced amplifier.
Referring to Figure 3-11, in a balanced amplifier design the RF signal enters at the left side as usual. Once inside, the signal gets split in two, with half going to one amplifier and half going to the other amplifier. Once inside the amplifiers, both "half" signals get amplified and are then added back together before leaving at the output on the right side. From the outside, a balanced amplifier looks and behaves just like a regular amplifier.
At this point you must be wondering why go through all that if one amplifier will do. Well, there are two advantages to the balanced amplifier design which just cannot be realized from a single amplifier. The first advantage is that there are two amplifiers. If one fails, there is still one working, albeit with reduced performance. Balanced amplifiers are often used in circumstances requiring a high degree of reliability or fault tolerance.
The second advantage-and you are going to have to take my word for it because it is too complicated to explain-is that balanced amplifiers provide a better match (lower VSWR) than regular amplifiers. They leak less, and less leaking is premium performance which some RF systems simply must have to function properly.
Did You Know?
Amplifier prices vary dramatically. In general, the higher the frequency or the wider the bandwidth or the greater the output power, the more the amplifier costs. Also, super low noise figure amplifiers are expensive. Just as an example, some low cost surface-mount amplifiers sell for as little as 50¢, while some high power, wideband amplifiers sell for as much as $50,000. Just be glad you don't have an RF engineer on your Christmas shopping list.
Variable Gain Amplifiers
There is one last amplifier type you should know about, the variable gain amplifier or VGA. Most amplifiers have fixed gain (i.e., the gain has one single value). A fixed gain amplifier with 10 dB of gain will make all input signals ten times bigger. Variable gain amplifiers have an external control which allows the user to vary the gain over some predefined range. All VGAs come specified with a "gain range" like 10-20 dB, for example. A VGA is like a gas range, where the heat (gain) varies from simmer to boil and the external control is the knob on the stove. The block diagram of a VGA is shown in Figure 3-12 and, as you can see, it is nothing more than a regular amplifier with an arrow through it.
Figure 3-12. Block diagram of a variable gain amplifier.
In practice, rarely is a VGA's external control connected to the outside world, like the knob on a stove. More often than not, the VGA's control is connected to some other point in the RF system that senses what is going on and changes the VGA's gain accordingly. Whenever a component changes its own performance based on something happening somewhere else in the system, it is known as feedback. VGAs frequently sense an RF signal "farther down the line." If the signal it is sensing is too big, it lowers its own gain, and if the signal isn't big enough, it cranks it up. It continues changing its own gain until the signal it is sensing is just right, and then it keeps its gain right at that point.
Did You Know?
There is an entire branch of electronics called feedback theory, which predicts the behavior of feedback circuits. The thermostat controlling a home's temperature is a classic example of feedback theory. All feedback circuits work the same way, by sensing an output (room temperature) and varying an input (furnace flames) to achieve a specific outcome (desired temperature). RF systems, like all electrical systems, make extensive use of feedback circuits. I guess the components don't like it when it gets too cold.