What is the bandwidth of an oscilloscope?

Without sufficient bandwidth, the oscilloscope will not be able to resolve high-frequency changes, bumps will be distorted, edges will disappear, and details will be lost. Without sufficient bandwidth, all the power and glamour of the oscilloscope will be meaningless.

Any signal can be decomposed into a superposition of many harmonics. From the frequency domain, the general principle of bandwidth selection is: the bandwidth is sufficient if it can cover 99.9% of the energy of each harmonic of the measured signal. The reason for bandwidth selection is that we cannot intuitively know what the bandwidth corresponding to 99.9% of the energy of the measured signal is.

The saying "the higher the bandwidth of an oscilloscope, the better" is correct in a sense: the higher the bandwidth, the higher the bandwidth that can accurately measure the measured signal, and the higher the accuracy of signal reproduction; the higher the bandwidth of the oscilloscope, the smaller the rise time of the oscilloscope, and the higher the accuracy of the rise time measurement. However, the higher the bandwidth, the greater the value and the more valuable it is. In addition, from the application point of view, the higher the bandwidth is not necessarily better.

When the signal is decomposed to the Nth harmonic, the energy is attenuated to 99.9%, which can leave a sufficient bandwidth margin when selecting and using the oscilloscope. However, too high a bandwidth will cause a serious problem: the introduced noise energy exceeds the signal's own energy within the equivalent bandwidth plan, which will also lead to inaccurate measurement results. This is the signal-to-noise ratio (SNR) issue that must be mentioned repeatedly in measurement.

Assuming that a 500MHz oscilloscope can cover 99.9% of the energy of the measured signal, the measurement accuracy can reach within 5%. However, if we insist on using a 1GHz oscilloscope, the noise energy introduced in the 500MHz~1GHz frequency range is much greater than the remaining 0.1% of the measured signal energy covered in the 500MHz~1GHz range. The measurement result in the time domain is reflected as a lot of random noise with high-frequency components superimposed on the waveform, which affects the measurement results of some parameters. Therefore, the measurement result of 500MHz is more accurate! This is why we have to bind the bandwidth to 20MHz when measuring power supply ripple.

In the digital world, rise time measurement is critical. When evaluating and measuring digital signals, such as pulses and steps, rise time may be a more appropriate performance consideration. The oscilloscope must have a sufficient rise time to accurately capture the details of the rapid transition. The faster the oscilloscope's rise time, the more accurately it can capture the key details of the rapid transition. Rise time describes the effective frequency resolution of the oscilloscope. In some applications, you may only know the rise time of the signal. There is a constant that can be used to relate the bandwidth of the oscilloscope to the rise time: Bandwidth = K/Rise Time

Among them, K is a value between 0.35 and 0.45, depending on the shape of the oscilloscope frequency response curve and the shape of the pulse rise time response. The K value of an oscilloscope with a bandwidth <1GHz is usually 0.35, and the K value of an oscilloscope with a bandwidth >1GHz is usually between 0.40 and 0.45.

To determine the oscilloscope bandwidth required to accurately detect the signal fluctuations in the application under consideration, the 5-fold rule should be used, that is, the oscilloscope bandwidth ≧ the highest frequency component of the signal X5

An oscilloscope selected using the 5x rule will provide less than ±2% error in measurements, which is generally sufficient for current applications. However, as signal speeds increase, this empirical rule may not be achieved.

Some oscilloscopes offer a way to increase bandwidth through digital signal processing. You can use a DSP arbitrary equalization filter to improve the oscilloscope channel response. This filter expands the bandwidth, flattens the oscilloscope channel frequency response, improves phase linearity, and achieves very good matching between channels. It also reduces rise time and improves time domain step response.