Filters play an important role in signal processing. A filter is an electronic circuit or software algorithm that lets certain frequencies pass through it and restricts others. At a basic level, you can think of them like a water filter. But instead of filtering out impurities in water, they filter out impurities in your signals, improving signal quality along the way. Also, as we’ll see, the ability of a filter to separate sets of frequencies brings with it additional utility.
The earliest signal filters came long before digital circuitry and transistors could enhance their performance. These simple filters, such as those created by George Campbell, relied on the properties of inductors and capacitors to restrict or allow signals to pass through them. With the advent of transistors and digital circuitry, filters were able to get more precise in their filtering. Today, many digital signal processing functions, including filtering, can take place entirely in software.
Types of Filters
Due to the limitation of relying purely on capacitors and inductors, the earliest filters were limited to high-pass and low-pass filters. However, by combining these filters, the filtering can be made more complex. In this section, we’ll explore the various types of filters that are made available from these humble beginnings.
Low-Pass Filters
The low-pass filter does exactly what its name implies; it allows low frequencies to pass through. Each filter will have a specific frequency cut-off point. Any signals above that point are blocked, while those below it can pass. Some applications of low-pass filters include:
- Removing low-frequency noise from an audio signal, such as removing a 60 Hz hum from electrical interference
- Separating the bass and treble frequencies in a music signal
- Removing the AC ripple from a power supply
High-Pass Filters
This is the opposite of a low-pass filter. These filters also have a specific cut-off frequency, but the cut-off goes in the other direction. Frequencies higher than the chosen level are allowed to pass through, and lower ones are restricted. High-pass filters have been used for:
- Removing high-frequency noise from an audio signal, such as hisses or static
- Smoothing out a noisy signal, such as in temperature or pressure sensors where high-frequency variations are not relevant
- Detecting edges in images so effects such as image sharpening can be applied
Band-Pass Filters
The goal of the band-pass filter is to allow a specific band of frequencies to pass through. A band-pass filter can be made simply by stacking a low-pass filter and a high-pass filter together in series. Below are some uses of the band-pass filter:
- Separating specific frequency components in an audio signal, such as isolating a particular instrument or voice
- Filtering out specific frequency components in a sensor signal, such as filtering out noise from other sources
- Tuning in to specific frequencies, such as a radio receiver
Band-Stop Filters
Also called a notch filter, the band-stop filter does the opposite of a band-pass. It prevents a certain band of frequencies from passing through and allows all others. These filters are made by connecting a low-pass filter and a high-pass filter in parallel instead of in series. Band stop filters have many applications:
- Noise reduction in telecommunications systems
- Improving signal quality in guitar amplifiers and other audio systems
- Interference reduction from close frequencies in a radio receiver
All-Pass Filters
Finally, we come to all-pass filters. These let all of the frequencies in the signal pass through but also allow for manipulation of the signal’s phase. This property of all-pass filters has several uses:
- Correcting phase problems in a signal
- Creating reverberation or phasing effects in audio signal processing
- Implementing parametric equalizers in digital audio workstations
Passive Filters
Passive filters use passive components such as resistors, capacitors, and inductors to restrict or pass certain frequency components of a signal. These filters do not require an external power supply. They are also simpler to design and implement than active filters. However, passive filters are limited in their frequency response and cannot provide gain. Some pros and cons of passive filters include:
Pros
Passive filters:
- Are simpler and cheaper to design and implement than active filters
- Do not require an external power supply
- Do not introduce noise or distortion due to active components
- Have a low noise floor and high dynamic range
Cons
However, these filters:
- Cannot amplify signals
- Have a limited frequency response
- May introduce signal attenuation due to component losses
- May have a higher input impedance and lower output impedance
Active Filters
Active filters use active components such as operational amplifiers (op-amps) for improving signal quality by manipulating the frequency components of a signal. They require an external power supply, allowing them to provide more precision and flexibility than passive filters. The external power supply also allows active filters to be used for signal amplification. Some pros and cons of active filters include:
Pros
Active filters:
- Can provide high precision and accuracy
- Can amplify signals
- Provide a specific frequency response (if properly designed)
- Can have a high input impedance and low output impedance
Cons
However, active filters:
- Require an external power supply
- May introduce noise or distortion due to active components
- Have a more complex circuit design than passive filters
- Cost more due to the use of active components
Choosing the Right Filter
Filters come in many different forms, with various options and specifications in each category. Finding the right filter for your use case can be difficult. The tips below will help you narrow down your choices when selecting a filter for improving signal quality.
- Filter Type: The first thing to consider is which kind of filter you’ll need. We’ve seen some applications of the main filter types in this article. Choose the type of filtering that matches your use case. Also, consider whether you need active filtering or if passive filtering will be sufficient.
- Frequency Range: Next, consider the specific frequencies that need to be filtered. Find a filter that operates within the filter range that your application requires.
- Impedance: The impedance of a filter is the resistance that the filter presents to the flow of a signal. When designing electrical circuits using filters, the impedance of the filter needs to be matched with the input and output signals.
- Frequency Response: Filters have grown more advanced since they were first introduced. They can now filter the signal in very precise ways, allowing for several frequency response options beyond simply passing or restricting the signal. Choose a frequency response curve that matches your needs.
Final Thoughts on Improving Signal Quality
Electronic filters are simple but powerful tools for improving signal quality. They can help to boost the signals you want to keep and restrict the ones you don’t. They cover a wide variety of applications, from power supply regulation to aiding in advanced image processing functions. There’s a lot to learn about these versatile devices. We hope this guide has helped you to understand them a little better.
