How is the oscilloscope FFT test implemented?

Internal Computing Spectrum Computing Unit.

The FFT operation of the oscilloscope is the fast Fourier transform, through which the conversion of the time-domain signal and the frequency-domain signal can be realized, and the frequency composition of the time-domain signal can be demonstrated. Each waveform can be decomposed into different frequencies and amplitude sine wave superposition, and the frequency points obtained by FFT operation are the frequencies of the harmonic components separated by square waves.

The purpose of the FFT arithmetic function is to test the impulse response of filters and systems; Distinguish and locate noise interference sources, and determine the true radiation; Analyze jitter, harmonic power, EMI; Due to the large amount of data processing required for FFT operation, many oscilloscopes are prone to jamming when performing FFT operation.

Most oscilloscopes have an FFT function, also known as a fast Fourier transform, but many people don't understand what this function is for.

The role of FFT fast Fourier transform:

FFT is the spectrum of analyzing signals, and has a wide range of applications in physics, electronics, number theory, combinatorics, signal processing, probability theory, statistics, cryptography, acoustics, optics, oceanography, structural dynamics and other fields.

Radio and television, which we are more familiar with, need to tune in to listen to different programs. The channel is the channel of the frequency, and different channels use different frequencies as a channel for information transmission.

The frequency domain analysis of the oscilloscope can also play a role in accelerating the debugging process in power supply debugging. In computers, the compression of images and files also uses the calculation of Fourier transform. There are also many tools in our commonly used PS software that use the Fourier transform algorithm.

Another example is the removal of some specific frequency components from a certain curve, that is, filtering, which is a very important concept in signal processing, and can only be easily done in the frequency domain. The noise-canceling headphones we use are the principle of filtering out the frequency of noisy sounds from the outside world.

Oscillograph. The FFT operation is the fast Fourier transform.

The Fourier transform can realize the conversion of time-domain signals and frequency-domain signals, and show the frequency composition of time-state domain signals. Each waveform can be broken down into sine waves of different frequencies and amplitudes.

Superposition, the frequency points obtained by the FFT operation are the frequencies of the harmonic components separated by the square wave.

The purpose of the FFT arithmetic function is to test the impulse response of filters and systems; Distinguish and locate noise interference sources, and determine the true radiation; Analyze jitter, harmonic power, EMI; Due to the large amount of data processing required for FFT operation, many oscilloscopes are prone to the phenomenon of covering the source card when performing FFT operation cavity calculation.

When an FFT is analyzed, the spectrogram measurement range is 1 2 (in the stopped state) or 1 4 (in the running state) of the FFT sampling frequency. The first thing to do is to make sure that the frequency point to be measured is within the measurement range. Using Zhiyuan Electronics ZDS2022 to find the frequency point of interest can be found in the following two ways.

2) If the frequency point of interest is not in the top 20, you can find it by zooming in on the spectrogram, zoom it out through the [M1] knob and move it through the [M2] knob, and locate the frequency point of interest in the center index to calculate its energy value.

The following points should be paid attention to when measuring the oscilloscope FFT:

1.Since FFT is a mathematical function, the more data it processes for a mathematical function, the more accurate it is. Therefore, when measuring, we need to increase the storage depth and time base as much as possible, so that the frequency resolution is higher.

In the following two figures, the comparison between the time base at 200 s and 2 ms respectively, it can be clearly seen that the FFT effect under the 2 ms time base is much better.

However, it should also be noted that the longer the signal length in the time domain, the better, because the oscilloscope has a limited storage depth, and the longer the waveform is recorded, the lower the sampling rate, which can lead to distortion of the source waveform. In general, a waveform duration of at least 4 to 8 waveform periods on a time-domain graph is appropriate.

2.Signals with DC components or deviations can cause errors or deviations in the formation of FFT waves, so AC coupling is an option to reduce the DC components.

3.When acquiring periodic signals, an average sampling mode should be used to reduce signal noise. It is recommended that the average number be no less than 16.

fft=fast fourior transform.This is a fast Fourier transform that is convenient for computer calculations.

There is also a type of DFT, but because of its complexity=n 2>>fft=nlogn, it is not adopted by computer algorithms.

The purple color is the spectrum of the above square wave in the frequency domain, and we know that the square wave can be expressed as the sum of an infinite number of sinusoidal functions by a Fourier series.

f(t) = a sin(k t) k (roughly not exact), so the frequency domain is also a superposition of sin(wt).

The pattern of the display is formed.

Because the FFT is sampled at a certain sampling rate, it should be a periodic function in the frequency domain.

However, due to the limitation of the number of points taken by DFT, and the number of points n*1 f (sampling time) is not equal to the period of the square wave, the function has a so-called sidelobe, so it is not a perfect impulse in the displayThe periodic feature is also not obvious, and taking the number of points in the FFT (DFT) affects the resolution in the frequency domain, and the more points you take, the higher the resolution in the frequency domain, and the more the display will look like an impulse