[image] Microcontrollers sit between the abstract world of bytes and the real world objects you can see and touch. Programming microcontrollers is a way to connect these two worlds. The satisfaction of getting a sensor reading flashing on a cheap LCD display is known only to those who have spent the hours fighting the electronics. Many low-cost (< $10) microcontrollers and development boards have significantly reduced the barrier to get started. One such example is the ESP8266 microcontroller chip and the NodeMCU development board. I recently got started with this system and the present article is a compilation of step-by-step notes I made while setting up my environment.
The first step is to install the USB-to-serial driver appropriate for the board we’re using. For the computer to recognize the microcontroller when plugged in, the USB-to-serial driver is necessary. Without it, your computer won’t recognize the microcontroller and programming it would be impossible. The name of the USB-to-serial converter is usually written on the back side of the development board [image]. The board I’m using has the CP2102 USB-to-serial converter whose driver can be downloaded from the vendor website.
Once you have downloaded and installed the USB-to-serial driver, you can check if the ESP is recognized by your PC
by plugging it to the USB port. On a MAC, the ESP, when recognized properly, shows up as a file item in the /dev.
[image]. At this point, we’re ready to take the next step.
The firmware can be informally thought of as the operating system (OS) of a microcontroller. The firmware provides
APIs for accessing the resources of the chip. For the ESP8266, there are multiple choices for the firmware depending
on the language.
Micropython
(or its close cousin CircuitPython) is the Python-based firmware and the one we will be using in this article.
Micropython is a compiled binary with a .bin extension. It can be downloaded from the Micropython
downloads section.
So this small 600 kB .bin file is the operating system for the ESP8266. Next we have to figure out a way to load
this OS into the the ESP8266.
When a chip is freshly powered on, it does not yet have the OS in its memory; some process must load software
into memory before it can be executed. This initial process is called the bootloader. For ESP8266 the most common
bootloader is the esptool which is fortunately pip installable:
$ pip install esptool
After installing esptool we first need to erase the flash memory and then load the Micropython firmware we downloaded
in the previous step. Flash can be erased using the command:
$ esptool.py --port /dev/cu.SLAB_USBtoUART erase_flash
The firmware can be loaded using:
$ esptool.py --port /dev/tty.SLAB_USBtoUART \
--baud 115200 write_flash \
--verify \
--flash_size=detect -fm dio 0 \
esp8266-20210902-v1.17.bin
The command for loading the firmware is a bit long. We can copy paste it on the commandline as is, or editing it into a single line by removing the backslashes. The crucial input here is the baud rate. The recommended baud rate is
--verify and -fm dio options can be skipped,
but I found that skipping them leads to corrupted loads.If everything went correctly, then the firmware should be loaded and the chip is ready for programming. The next step shows how to verify the correct loading of the firmware.
This part describes how to obtain a Python shell running on the ESP8266. This is exciting since different chip level operations (making a pin ‘hi’ or ‘low’, sensor voltage, list of WiFi access points etc) can be tested in an interactive manner without having to recompile and reload the code.
Once the firmware has been loaded, multiple application can be used to connect to the ESP8266. On a Mac, the native
terminal multiplexer program screen can be used to open a console terminal on
the ESP8266. While many people have had success using screen to communicate with the chip, I could not get it to
work. Even after successful upload of the firmware, I could not get a Python shell to open using screen.
This is when I switched to Minicom
Minicom is a serial port communication application. That is, it enables communication with devices connected to the computer’s serial port. Minicom can be installed via Homebrew:
$ brew install minicom
Minicom is a menu-driven application. One-time initial setup is required, which can be performed by bringing up the setup menu:
$ minicom -s
The main settings are the baud rate and the default serial port location under “Serial port settings”. To avoid having
to repeatedly type the name of the serial port, the default value can be provided under “Serial Device”. Replace the
empty or existing value by the full path of the serial port. Also replace the empty or the default baud rate with
the recommended value 115200. At this point you can press ESC to exit out of minicom.
Once the setup is complete you can simply type minicom in the shell; it should open up the Python terminal on
ESP8266 ! Remember this is a greatly stripped down version of the regular Python. Poke around to
learn more about the available modules, system information, files etc. If you have some LEDs and resistors, you can
do some cool things right in the Python shell at this point including:
All of these can be done by following the brief tutorial at Micropython.org.
To exit out of Minicom, bring up the menu by pressing the exit key combination (ESC-x on my Mac) and selecting
`‘yes’ on the confirmation menu.
It is cool to interactively execute code and watch it make changes in the real world like turning on lights. But
your code will vanish from the chip memory the moment you exit out of minicom. The cool thing about microcontrollers
like the ESP8266 is that they can be powered externally (i.e. without the USB connection) and allowed to run freely.
To allow that, the code has to be burned into its non-volatile memory. Once the code is burned in the non-volatile
memory, it says there even after the power is turned off. Also, if the code is named main.py it will be run
automatically after the chip is powered up again. To know how this happens read more about the
ESP8266 boot procedure.
To burn our locally developed code, we use the Adafruit Micropython Tool, aka
ampy.
You can install ampy using pip:
$ pip install adafruit-ampy
ampy has simple get and put commands for two-way communication between the computer and the ESP. To transfer a
Python file on your local machine to the ESP’s non-volatile memory, you can use:
$ ampy --port /dev/cu.SLAB_USBtoUART put path/to/localfile.py path/to/remotefile.py
Similarly to transfer a file on the ESP to current directory on your local machine you can use:
$ ampy --port /dev/cu.SLAB_USBtoUART get path/to/remotefile.py
ampy supports a whole range of operations in addition to the get and put, which can be read from the
documentation.
main.py