Practical Raspberry Pi 3 Projects: Building an I2C Environmental MQTT Logger

While the Raspberry Pi 5 dominates current headlines, the Raspberry Pi 3 remains a workhorse for specific embedded applications. If you are scavenging parts from a drawer or buying surplus hardware, Raspberry Pi 3 projects still offer a compelling mix of low power draw, native 40-pin GPIO, and sufficient processing headroom for sensor polling and network telemetry. However, the BCM2837 SoC has distinct hardware quirks—particularly regarding I2C pull-up resistors and Wi-Fi chiplet behavior—that will break your code if you treat it exactly like a Pi 4 or 5.

This guide walks through building a robust I2C environmental data logger with an OLED display and MQTT uplink. We will cover the exact hardware constraints, provide fully compilable Python code with error handling, and detail the specific debugging steps for the most common I2C failures on this board.

Why Build Raspberry Pi 3 Projects in 2026? (And the Spec Reality)

The Raspberry Pi 3 Model B+ (released in 2018) is the definitive version of the Gen 3 lineup. It upgraded the Wi-Fi to a dual-band Cypress CYW43455 chip and added Gigabit Ethernet (though bottlenecked by the USB 2.0 bus). For headless sensor nodes, the Pi 3B+ draws roughly 2.5W to 4W at idle, compared to the Pi 4's 3.5W to 6W, making it superior for off-grid solar or PoE-powered deployments where every watt matters.

Before wiring up sensors, you must understand the electrical differences in the GPIO header across generations. The internal I2C pull-up resistors on the Pi 3 are notoriously weak, which dictates how you wire external sensors.

Source: Raspberry Pi Official Hardware Documentation

Parts List and Pin Mapping

This build targets the Raspberry Pi 3 Model B+ running Raspberry Pi OS Bookworm (64-bit). Do not use 5V logic sensors; the BCM2837 GPIO pins are strictly 3.3V tolerant and will suffer permanent damage if exposed to 5V.

Bill of Materials

• Board: Raspberry Pi 3 Model B+ (Element14 / RS Components variant)
• Sensor: BME280 Temperature/Humidity/Pressure Breakout (Adafruit 2652 or equivalent 3.3V variant)
• Display: SSD1306 0.96-inch 128x64 I2C OLED (Address 0x3C)
• Passives: 2x 4.7kΩ through-hole resistors (for I2C pull-ups)
• Wiring: 28 AWG silicone stranded wire or a 40-pin ribbon cable

Pin Mapping Table

Assembly and I2C Bus Configuration

1. Enable I2C: Run sudo raspi-config, navigate to Interface Options > I2C, and enable it. Reboot the Pi.
2. Install Tools: Run sudo apt update && sudo apt install -y i2c-tools python3-smbus python3-pip.
3. Wire the Pull-ups: Solder one 4.7kΩ resistor between Pin 1 (3.3V) and Pin 3 (SDA). Solder the second 4.7kΩ resistor between Pin 1 (3.3V) and Pin 5 (SCL). This ensures clean square-wave clock edges.
4. Connect Sensors: Wire the BME280 and OLED in parallel on the I2C bus. Ensure the BME280's I2C address jumper is set to 0x77 (default for Adafruit) or 0x76 (common for generic clones).
5. Verify Addresses: Run i2cdetect -y 1. You should see 3c (OLED) and 77 (BME280) in the grid. If you see UU, a kernel driver has already claimed the device, which will block user-space Python scripts.

The Python MQTT Logger Code

This script reads the BME280, updates the local OLED, and publishes the payload to an MQTT broker. It includes explicit pin definitions, I2C error handling, and network reconnection logic. You will need to install the dependencies first: pip3 install smbus2 RPi.bme280 paho-mqtt luma.oled luma.core.

Reference: Eclipse Paho MQTT Python Client Documentation

import time
import json
import smbus2
import bme280
import paho.mqtt.client as mqtt
from luma.core.interface.serial import i2c
from luma.oled.device import ssd1306
from luma.core.render import canvas
from PIL import ImageFont

# --- HARDWARE DEFINITIONS ---
# Target: Raspberry Pi 3 Model B+ (I2C Bus 1)
I2C_PORT = 1
BME280_ADDRESS = 0x77  # Change to 0x76 if using generic clone
OLED_ADDRESS = 0x3C

# --- MQTT CONFIGURATION ---
MQTT_BROKER = '192.168.1.100'
MQTT_PORT = 1883
MQTT_TOPIC = 'sensors/pi3_lab/environment'

# Initialize I2C Bus and Sensors
bus = smbus2.SMBus(I2C_PORT)
calibration_params = bme280.load_calibration_params(bus, BME280_ADDRESS)

# Initialize OLED Display
serial_interface = i2c(port=I2C_PORT, address=OLED_ADDRESS)
display = ssd1306(serial_interface, width=128, height=64)

# Use default font; fallback to basic if custom fonts fail
try:
    font = ImageFont.truetype('/usr/share/fonts/truetype/dejavu/DejaVuSans.ttf', 14)
except IOError:
    font = ImageFont.load_default()

def on_connect(client, userdata, flags, rc, properties=None):
    if rc == 0:
        print('Connected to MQTT Broker')
    else:
        print(f'MQTT Connection failed with code {rc}')

client = mqtt.Client(mqtt.CallbackAPIVersion.VERSION2)
client.on_connect = on_connect

try:
    client.connect(MQTT_BROKER, MQTT_PORT, 60)
    client.loop_start()
except Exception as e:
    print(f'MQTT Initial Connection Error: {e}')

def read_and_publish():
    while True:
        try:
            # Read BME280 Data
            data = bme280.sample(bus, BME280_ADDRESS, calibration_params)
            temp_c = round(data.temperature, 2)
            humidity = round(data.humidity, 2)
            pressure = round(data.pressure, 2)

            # Update OLED Display
            with canvas(display) as draw:
                draw.text((0, 0), f'Temp: {temp_c} C', font=font, fill='white')
                draw.text((0, 20), f'Hum:  {humidity} %', font=font, fill='white')
                draw.text((0, 40), f'Pres: {pressure} hPa', font=font, fill='white')

            # Publish to MQTT
            payload = json.dumps({
                'temperature': temp_c,
                'humidity': humidity,
                'pressure': pressure,
                'timestamp': time.time()
            })
            
            if client.is_connected():
                client.publish(MQTT_TOPIC, payload)
            else:
                print('MQTT disconnected. Attempting reconnect...')
                client.reconnect()

        except OSError as e:
            print(f'I2C Hardware Error: {e}')
            # Clear display to indicate fault
            display.clear()
        except Exception as e:
            print(f'Unexpected Error: {e}')
            
        time.sleep(10) # 10-second polling interval

if __name__ == '__main__':
    try:
        read_and_publish()
    except KeyboardInterrupt:
        print('Logger stopped.')
        display.clear()
        client.loop_stop()
        client.disconnect()

Debugging: 'OSError: [Errno 121] Remote I/O error'

When running I2C scripts on the Pi 3B+, the most frequent failure is the kernel throwing an I/O error. If your terminal outputs OSError: [Errno 121] Remote I/O error, the BCM2837 SoC failed to receive an ACKnowledge (ACK) bit from the sensor during the I2C transaction.

The First Three Things to Check When It Fails

1. Verify Bus Visibility via CLI: Stop your Python script and run i2cdetect -y 1. If the grid is entirely blank or shows dashes where your sensor should be, the hardware link is broken. If the address shows up here but fails in Python, another process (like a systemd service or gpio daemon) is hogging the bus.
2. Measure Voltage Under Load: The Pi 3B+ 3.3V rail is generated by an onboard LDO that is highly sensitive to input voltage drops. Use a multimeter to measure Pin 1 (3.3V) relative to Pin 6 (GND) while the script is running. If it reads below 3.2V, your power supply is browning out, causing the SoC to throttle and the I2C clock timing to skew. Upgrade to a 5V/2.5A+ supply with thick USB cables.
3. Check for Logic Level Mismatch: Did you accidentally wire the sensor VCC to Pin 2 (5V) while the SDA/SCL lines are on 3.3V pins? Feeding 5V into the SDA line of a 3.3V sensor can backfeed into the BCM2837, permanently damaging the I2C block. Always power I2C sensors from Pin 1 (3.3V) on the Pi 3.

Ranked Causes for Errno 121

Extending or Simplifying the Build

Depending on your deployment environment, you may need to alter the footprint of this Raspberry Pi 3 project.

How to Simplify (Headless Deployment)

If this node is going into a sealed NEMA enclosure in an attic or greenhouse, drop the SSD1306 OLED entirely. Remove the luma.oled dependencies from the code and the physical wiring. This reduces the 3.3V current draw by roughly 20mA, eliminates a point of failure, and prevents OLED burn-in from static text. Run the script as a systemd service to ensure it restarts automatically after power blips.

How to Extend (Industrial or Long-Range)

The Pi 3B+ lacks native RS485, which is the standard for industrial Modbus sensor networks. To extend this build for industrial environments, add a MAX485 TTL-to-RS485 module. Wire the MAX485 DI/RO pins to the Pi's hardware UART (GPIO 14/15), and use the pymodbus library to poll industrial temperature transmitters. For remote agricultural sites where Wi-Fi is unavailable, swap the MQTT-over-Wi-Fi approach for a Dragino LoRa/GPS HAT, which sits directly on the 40-pin header and transmits telemetry over LoRaWAN to a local gateway.