The Fourier transform can be powerful in understanding everyday signals and troubleshooting errors in signals. Although the Fourier transform is a complicated mathematical function, it isn’t a complicated concept to understand and relate to your measured signals. Essentially, it takes a signal and breaks it down into sine waves of different amplitudes and frequencies. Let’s take a deeper look at what this means and why it is useful.
All Signals Are the Sum of Sines
When looking at real-world signals, you usually view them as a voltage changing over time. This is referred to as the time domain. Fourier’s theorem states that any waveform in the time domain can be represented by the weighted sum of sines and cosines. For example, take two sine waves, where one is three times as fast as the other–or the frequency is 1/3 the first signal. When you add them, you can see you get a different signal.
Figure 1: When you add two signals, you get a new signal.
Now imagine if that second wave was also 1/3 the amplitude. This time, just the peaks are affected.
Figure 2: Adjusting the amplitude when adding signals affects the peaks.
Imagine you added a third signal that was 1/5 the amplitude and frequency of the original signal. If you continued in this fashion until you hit the noise floor, you might recognize the resulting waveform.
Figure 3: A square wave is the sum of sines.
You have now created a square wave. In this way, all signals in the time domain can be represented by a series of sines.
Although it is pretty neat that you can construct signals in this fashion, why do you actually care? Because if you can construct a signal using sines, you can also deconstruct signals into sines. Once a signal is deconstructed, you can then see and analyze the different frequencies that are present in the original signal. Take a look at a few examples where being able to deconstruct a signal has proven useful:
- If you deconstruct radio waves, you can choose which particular frequency–or station–you want to listen to.
- If you deconstruct audio waves into different frequencies such as bass and treble, you can alter the tones or frequencies to boost certain sounds to remove unwanted noise.
- If you deconstruct earthquake vibrations of varying speeds and strengths, you can optimize building designs to avoid the strongest vibrations.
- If you deconstruct computer data, you can ignore the least important frequencies and lead to more compact representations in memory, otherwise known as file compression.
Deconstructing Signals Using the FFT
The Fourier transform deconstructs a time domain representation of a signal into the frequency domain representation. The frequency domain shows the voltages present at varying frequencies. It is a different way to look at the same signal.
A digitizer samples a waveform and transforms it into discrete values. Because of this transformation, the Fourier transform will not work on this data. Instead, the discrete Fourier transform (DFT) is used, which produces as its result the frequency domain components in discrete values, or bins. The fast Fourier (FFT) is an optimized implementation of a DFT that takes less computation to perform but essentially just deconstructs a signal.
Take a look at the signal from Figure 1 above. There are two signals at two different frequencies; in this case, the signal has two spikes in the frequency domain–one at each of the two frequencies of the sines that composed the signal in the first place.
Figure 4: When two sine waves of equal amplitude are added, they result in two spikes in the frequency domain.
The amplitude of the original signal is represented on the vertical axis. If you look at the signal from Figure 2 above where there are two different signals at different amplitudes, you can see that the most prominent spike corresponds to the frequency of the highest voltage sine signal. Looking at a signal in the time domain, you can get a good idea of the original signal by knowing at what frequencies the largest voltage signals occur.
Figure 5: The highest spike is the frequency of the largest amplitude.
It can also be helpful to look at the shape of the signal in the frequency domain. For instance, let’s take a look at the square wave in the frequency domain. We created the square wave using many sine waves at varying frequencies; as such, you would expect many spikes in the signal in the frequency domain—one for each signal added. If you see a nice ramp in the frequency domain, you know the original signal was a square wave.
Figure 6: The frequency domain of a sine wave looks like a ramp.
So what does this look like in the real world? Many mixed-signal oscilloscopes (MSO) have an FFT function. Below, you can see what an FFT of a square wave looks like on a mixed-signal graph. If you zoom in, you can actually see the individual spikes in the frequency domain.
Figure 7: The original sine wave and its corresponding FFT are displayed in A, while B is a zoomed in portion of the FFT where you can see the individual spikes.
Looking at signals in the frequency domain can help when validating and troubleshooting signals. For instance, say you have a circuit that is supposed to output a sine wave. You can view the output signal on the oscilloscope in the time domain in Figure 8 below. It looks pretty good!
Figure 8: If these two waves were added, they would look like a perfect sine wave because they are so similar.
However, when you view the signal in the frequency domain, you expect only one spike because you are expecting to output a single sine wave at only one frequency. However, you can see that there is a smaller spike at a higher frequency; this is telling you that the sine wave isn’t as good as you thought. You can work with the circuit to eliminate the cause of the noise added at that particular frequency. The frequency domain is great at showing if a clean signal in the time domain actually contains cross talk, noise, or jitter.
Figure 9: Looking at the seemingly perfect sine wave from Figure 8, you can see here that there is actually a glitch.