Running and Debugging Python on the Pi Pico

I’m currently refreshing my decades-old understanding of c++ before working on a new project on the Pi Pico. As that’s gonna take a while, I thought I’d see how the Pi Pico is with Python.

CircuitPython

The sample code provided by PImoroni includes python files that are meant to be used with CircuitPython, so I wanted to try that first.

Initial Setup

Before you can use CircuitPython, you need to install the appropriate UF2 file on your Pi Pico. As I’m using a Pimoroni Pico Explorer Base, I needed to install their custom UF2 file that includes the libraries for the display, et cetera.

Editing and Deploying Code

To actually run the examples, I tried two popular editors, Thonny, and Mu.

Thonny

In order to get things working, I had to configure Thonny to use the right interpreter, i.e. the interpreter on my Pi Pico:

1. Using the hamburger icon in the bottom right of Thonny’s window, select “Configure Interpreter”.
2. Select “MicroPython (Raspberry Pi Pico)”
3. Hit “OK”
4. I then had to disconnect and reconnect the Pico before Thonny would detect it.

I was then able to load the examples onto my Pico without problems. Thonny got me excited with its debugger icon, but sadly this apparently doesn’t work with MicroPython.

Mu Editor

The setup for Mu Editor was very straightforward, I just had to change the “mode” to “RP2040”. It was then able to load the examples onto my Pico with no problems. Between the two, I like the look of and the built-in tools in Mu Editor a bit better.

Arduino IDE

I started out with the PI 2040 because the board I was using with the Arduino IDE (an Adafruit Neopixel Trinkey) did not support debugging. Now that I have a probe, I wanted to see if:

1. I could build on the code I wrote for the Trinkey using the Arduino IDE
2. I could get the debugger in the Arduino IDE working with the Pico.

First, as I had been running MicroPython on the Pico earlier, I rebooted the Pico with the “Bootsel” button held. You should only need to do this once.

I then needed to make the Arduino IDE aware of the board. There is a library to add the Pi Pico as a “board” in the Arduino IDE, which I installed using the “Board Manager”. As I have a Raspberry Pi Debug Probe which is also a Pico, I used “Get Board Info” under the “Tools” menu to figure out which port was the Pico + Explorer Base, and which was the probe. I selected the port for the (non probe) Pico, and changed the following settings:

1. Debug Level is set to Wire.
2. Upload Method is set to Picoprobe (CMSIS-DAP).

With that, I was able to upload sketches to the Pico, and also to use the Arduino IDE’s debugger.

VS Code

I did briefly try the MIcroPico extension for VS Code, which did allow me to at least run one of the pimoroni examples once. I was not overly impressed with the reliability, as I would often have to reset the pico to run a new file.

Lots of additional features (library management, etc) seem to depend on having wifi enabled/configured on your pico, which seems overly involved for the kinds of sketching I have in mind.