When using an RTOS like FreeRTOS, sooner or later you have to ask the question: how much time is spent in each task? The Eclipse based MCUXpresso IDE has a nice view showing exactly this kind of information:
FreeRTOS Runtime Information
For FreeRTOS (or that Task List view) to show that very useful information, the developer has to provide a helping hand so the RTOS can collect this information. This article shows how this can be done on an ARM Cortex-M.
For this last blog in the series Investigating ARM Cortex® M33 core I decided to explore the expansion features of the LPC55S69-EVK. This board has three expansion ports (PMOD, Arduino Duo, Mikroe click) and I picked the Mikroe expansion port. Why? Only because I had good experience with these boards with the Hexiwear project.
And because I have been doing some work this month with AWS IOT I wanted to get my LPC55S69-EVK onto my office WiFi network for the Christmas holidays. I know that the MCUXpresso SDK for lpcxpresso55s69 version 2.6.3 has a built-in WiFi example named qca_demo, and so that is what I am investigating today.
That WiFi example supports three WiFi shield boards, and I picked the Mikroe WiFi 10 click board. It’s part number MIKROE-3432 and available from all of the usual catalogue distributors.
In last week’s blog I explained that the LPC55S69 microcontroller from NXP has two Cortex® M33 cores, named core0 and core1. There was a lot of theory, and so this week I put it all into practice and show you how to debug 2 cores with MCUXpresso IDE.
Throughout this series I’ve been using the LPC55S69 microcontroller from NXP as a platform to investigate the ARM Cortex® M33 core. NXP designed the LPC55S69 with two Cortex M33 cores and so this week I’m investigating these in more detail.
You’ll remember that when ARM launch a processor core it will have a number of optional features. This is shown very clearly on the LPC55S69. The 150 MHz primary core – cpu0 – is a full implementation of Cortex® M33 and includes the optional components FPU, MPU, DSP, ITM and the TrustZone® features.
I’ve always felt that the Fourier Transform (and in particular the embedded implementation Fast Fourier Transform) is the GOAT* of the DSP algorithms. The ability to convert a time-domain signal into a frequency-domain signal is invaluable in applications as diverse as audio processing, medical electrocardiographs (ECGs) and speech recognition.
So this week I’ll show you how to use the Transform engine in the PowerQuad on LPC55S69 to calculate a 512-point FFT. All of the difficult steps are very easily managed and the PowerQuad does all of the very heavy lifting.