FILTERS
Block Diagram
Figure 3-13. Block diagram of a filter.
This next device in the lineup is called a filter (see Figure 3-13). The filter does just what you think it does—it filters out all the signals which are not wanted. Visually it can be viewed as a bunch of signals at different frequencies entering the filter on the left (f1, f2, and f3) and only the desired signal (at the desired frequency) coming out the right side (f2). Where does all this "stuff" which is unwanted come from? Everywhere.
The Filter's Function
There are tons (actually, they don't weigh anything) of these invisible RF waves cruising around in the air at any moment in time at all different frequencies. These waves come from all of the RF transmissions taking place, like cellular phones, satellite communications, radar, even sunspots. All of these waves cruising around try to get into the receiver by way of the receiver antenna, and many do. Some of the signals even get past the LNA, but that's where the filter comes in. It acts like a bouncer at a nightclub letting all of the "acceptable" signals in and turning away all of those other "loser" signals. Instead of selecting people by attractiveness or wealth, the filter selects signals by their frequency.
In an ideal (and simplified) world, when the filter gets done doing its thing, the only signal still standing is the exact signal desired. Of course the world is not ideal, which is why you will often see several filters in both the transmitter and receiver.
It is simple enough to understand what the filter's purpose is in a receiver, but why is one needed in a transmitter? As you will soon learn, there is an evil component in all RF systems known as a mixer. And what this evil mixer does, (among other things), is inject unwanted signals (at unwanted frequencies) into the signal to be transmitted. But alas, the transmitter gets the last laugh, because just before the signal gets amplified by the HPA on its way out the antenna, a filter is used to eliminate all of those unwanted signals the evil mixer injected into the original signal in the first place. Ha ha ha ha ha.
The Role of the FCC
In most cases, the filter on the transmit side is required by the Federal Communications Commission (FCC). When someone is given permission to transmit a signal at a particular frequency, they are prohibited from transmitting at any other frequency, for fear it will mess up someone else's signal. The transmit filter makes sure none of these "illegal" signals ever leave the transmitter.
Did You Know?
The FCC, which has the power to allocate all airborne transmission in the U.S., actually gave away-for free-the right to the cellular frequency bands by way of lottery many years ago. After discovering that this little ploy netted the U.S. Government exactly zero dollars, they wised up. All rights to wireless transmission since then have been auctioned off to the highest bidder. Cha ching.
Filter Types
There are so many different filters that someone could write an entire book on the subject, which probably explains why there are entire books on the subject. But all filters, no matter how they are constructed, fall into one of four categories, as detailed in Table 3-2.
| Filter Types | Explanation |
|---|---|
| Low pass | Allows all frequencies below a certain frequency to pass while rejecting all others. It is like letting only short people into the bar. |
| High pass | The opposite of a low pass. |
| Bandpass | Allows all frequencies between two specific frequencies to pass while rejecting all others. It is like letting only people between 5′10″ and 6′2″ into the bar. It is also called a notch filter. |
| Band reject | The opposite of bandpass. |
Several filters are shown in Figure 3-14. Notice all the different shapes and sizes (how pretty!). Filters share the same common frequency property as all other RF stuff: the higher the frequency, the smaller the filter.
Filter Performance
Frequency Response
Generally speaking, filters are passive devices and therefore do not require a power supply. In fact, filters operate by varying their insertion loss as a function of frequency (see Figure 3-15). Figure 3-15 is known as the frequency response of the filter. (Could it be called anything else?) Every filter has a frequency response, which is all that is required to describe the filter's performance. Figure 3-15 shows the frequency response of a bandpass filter. With frequencies less than point A and greater than point B, the insertion loss is high. High insertion loss for a filter might be 30 dB, which means at those frequencies, the filter wipes out 99.9% of the RF signal. Conversely, frequencies between points A and B experience a low insertion loss. Low insertion loss might be one dB, which only wipes out about 20% of the signal.
Referring once again to Figure 3-15, the frequencies between points A and B are known collectively as the passband, because the filter allows frequencies in this band to pass on through. Likewise, the frequencies outside of the points A and B are collectively known as the stopband.
Figure 3-14. Filters. Courtesy of K&L Microwave, Inc.
Figure 3-15. A band pass filter's frequency response.
Special Filters
Duplexers
You may hear the word duplexer (or diplexer), which is a fancy device that combines two filters into a single component. These are mostly used in conjunction with basestation antennas. By combining both the transmitter and receiver filters into the same device, the same antenna can be used for transmitting and receiving, thereby reducing the number of antennas required at the basestation. If a duplexer combines two filters in a single device, how many filters do you suppose get combined in a triplexer?
SAW Filters
There is one last type of filter you should be aware of, the saw filter, and no, it is not made out of the wood cutting tool. "Saw" stands for surface acoustic wave, which is a fancy way of saying sound wave. Remember that RF devices get bigger as the frequency gets smaller, and below a certain frequency standard RF filters become prohibitively large. One day some bright RF engineer discovered that if the RF signal is first converted into a sound signal, the components required to filter this "acoustic" wave are much smaller, which is how saw filters work. First, they convert the RF signal into a sound signal, then they filter the sound signal, and finally, they convert the sound signal back to an RF signal. Saw filters are used for very low frequency filtering. (Aren't you glad you know that?)
One last thing needs to be mentioned about filters in particular, and passive devices in general. Amplifiers aren't the only components which have a one dB compression point (P1dB) and an third order intercept point (Ip3). Passive components, like filters as well as others, also have these parameters. In the case of passive components, these two parameters do not measure how much power the components put out (they don't—they're passive), they measure how much power the components can handle without distorting the signal. You can imagine that if a very large signal is put into a small filter, something bad will happen. In fact the "bad" thing which happens is signal distortion, which is any unintended change in a signal's size or shape. Of course, if the signal becomes much too big for the filter, the filter might catch on fire, which is also a form of signal distortion.