MATLAB: Analyzing PicoScope data using Signal Processing Toolbox

Introduction

The articles on this page demonstrate the use of Signal Processing Toolbox^TM features to perform analysis and filtering on data collected directly into the MATLAB^® Environment using a PicoScope 5000 Series Flexible Resolution Oscilloscope and its corresponding MATLAB Instrument Driver for Instrument Control Toolbox^TM.

In the examples below, we used a PicoScope 5244B oscilloscope.

Post-capture – inspecting data plots

From the resulting waveforms captured we can see that the 16-bit resolution capture provides a much smoother sine wave compared to the capture at 8-bit resolution. At 16-bit resolution we have 65536 quantization levels over the full-scale voltage range compared to 256 quantization levels at 8-bit resolution.

Given that the voltage range used was ±5 volts, each quantization level corresponds to the following:

• 8-bit: 39.1 mV
• 16-bit: ~0.153 mV

The trigger point is marked with a red 'X'.

Block data capture (8-bit)

Block data capture (16-bit)

Post-capture – obtaining measurements

Once the data is captured we can use functions provided by Signal Processing Toolbox to obtain the following measurements:

• Peak to peak (peak2peak)
• Rise time (risetime)
• Fall time (falltime)
• Root mean square (rms)
• Spurious free dynamic range (sfdr)
• Signal to noise ratio (snr)

The measurements can easily be tabulated using the table function:

Tabulated measurements

Comparing the data for both captures, the difference in the values obtained can be attributed to the number of quantization levels as mentioned above, which in turn has an impact on the data values used in the calculations.

Performing a fast Fourier transform on a signal and analyzing data

In this example, a fast Fourier transform (FFT) is applied to a square wave captured using the PicoScope followed by measurements on the square wave itself.

Setup

Our MATLAB script configured the signal generator to output a ±2 volt square wave with a frequency of 1 kHz and set a trigger for a rising edge through 1 V on Channel A. We connected the 'Gen' output on the PicoScope to the Channel A input.

The script set up the PicoScope to capture a block of data with 3500 pre-trigger and 3500 post-trigger samples at a sampling interval of 50 µs (20 MS/s) at 14-bit resolution.

Post-capture – applying a fast Fourier transform

Once the data has been captured, we can compute the discrete Fourier transform (DFT) of the data to view it in the frequency domain using the fft function. The figure shows the fundamental frequency and odd-harmonic components of the waveform in the bottom sub-plot.

PicoScope 5000 series data and DFT plot

Post-capture – analyzing waveform data

Signal Processing Toolbox also provides functions to find overshoot, undershoot, pulse width and duty cycle of a waveform.

Here, the overshoot function plots and annotates a sub-section of the waveform data that is passed as an argument to the function.

Plot generated by overshoot function

Calling the pulsewidth function on the same sub-section of the waveform data without assigning the output of the function results in a plot of the data with annotation.

Plot generated by pulsewidth function

Assigning the output of the pulsewidth function to a variable, pw, results in the following,

pw =

   4.9990e-04

This corresponds to the time for half a cycle of a square wave at 1 kHz frequency, which is 500 µs.

The dutycycle function can be applied to the sub-section of the waveform data and this results in:

dutyCycle =

    0.5001

This is close enough to the 50% that would be expected for this square wave.

Applying post-capture filtering to streaming data

In this example, we use the Filter Design and Analysis Tool available in the Signal Processing Toolbox to design a filter that can be applied as a post-capture operation on data containing random noise values.

Arbitrary waveform plot

Setup

Our MATLAB script configured the signal generator to output an arbitrary waveform defined as a sine wave with random added noise (shown on the right) with an amplitude of ±2 volts and frequency of 1 Hz.

The script set a trigger for a rising edge through 500 mV on Channel A. We connected the 'Gen' output on the PicoScope to the Channel A input.

The script set up the PicoScope to capture data in streaming mode at a sampling rate of 1 MS/s and to stop after capturing 10 million post-trigger samples (10 seconds).

Low pass Butterworth filter design using FDATool

Filter design

You can use the Filter Design and Analysis Tool (FDATool) to design a filter which can be applied to the data post-capture.

The image to the right shows the configuration used for a low-pass Butterworth Infinite Impulse Response (IIR) Filter. It shows the parameters as well as the magnitude and phase response.

Once the filter is designed, you can export it to a MATLAB function via the FDATool User Interface. The function when called returns a discrete-time filter object. You can pass some properties of this filter object to the filtfilt function as arguments along with the waveform data in order to apply zero-phase digital filtering.

MATLAB is a registered trademark of The MathWorks, Inc. and Instrument Control Toolbox and Signal Processing Toolbox are trademarks of The MathWorks, Inc.