Micropython thermometer using MCP9808 and SSD1306 OLED display

This little post was triggered by reading Owen Duffy's post titled "Arduino thermometer using DS18B20 and OLED display". Even though I'm quite comfortable writing C or C++, I've been enjoying the clean MicroPython environment. These days I've pretty much switched from classic Atmel Arduinos to RP2040 based boards.

I implemented this little thermometer in just a few minutes with two easy to find and install libraries. Here's how it looks:

And here is the code:

import time

from machine import Pin, SoftI2C

import ssd1306

from micropython_mcp9808 import mcp9808

i2c = SoftI2C(sda=Pin(6), scl=Pin(7), freq=100_000)

display = ssd1306.SSD1306_I2C(128, 32, i2c)

mcp = mcp9808.MCP9808(i2c) # , address=0x18)

time.sleep(0.5)

while True:

    temp = mcp.temperature

    #print(temp)

    time.sleep(5.0)

    display.fill(0)

    display.text(f"{mcp.temperature:.2f}C", 30, 15, 1) # x,y,colour

    display.show()

The imported modules are mcp9808 for the i2c temperature board and ssd1306 for the OLED display.

I use Thonny for editing, uploading and debugging.

One mystery is that the hardware I2C isn't working for me on the RP2040 board but SoftI2C works fine.

When getting started I scan for I2C devices like this:

i2c = SoftI2C(sda=Pin(6), scl=Pin(7), freq=100_000)

devices = i2c.scan()

print("devices: ")

for device in devices:

    print("0x{0:02x}".format(device))

Obviously adjust for the pins you've used. The RP2040 is a wonderfully powerful chip. The board I used here is from Seeed studio.

Cheers to Owen for all his great insights and sharing.

ChatGPT is very good at Micropython

While I'm here, an observation about the utility of the Large Language Models. I have been asking ChatGPT to write code to demonstrate how to do things and it's really good!

Around the house I have little devices that use Wifi to poll my solar inverter and show power use and generation. I've been using a third party JSON parser but I asked ChatGPT how to do it and it mentioned a built-in JSON parser that I wasn't aware of!