Firebeetle 2 Gas Sensor

Robin Mitchell did a Product of the Week video on the  FireBeetle 2 ESP32-E board from DFRobot about a month ago.  And I won the giveaway on Ian's  show.   I had mentioned that I would use it to build a portable gas sensor since I had just gotten a Multichannel Gas Sensor board from Seeed Studio.  So, here is that project.

The FireBeetle 2 ESP32-E board has some great features that make it ideal for building a portable IoT sensor - ESP32 processor with WiFi/Bluetooth, battery interface, GDI display port, and female GPIO headers.

Components

Here are the parts that I am using for my project (front and back):

Top: Fermion 1.8" 128x160 TFT Displaywith GDI cable

Middle: FireBeetle2 ESP32E

Bottom: Seeed MultiGas Sensor V2

LiPo battery

Arduino IDE

I normally use the Arduino IDE to program ESP32 boards and the DFRobot Beetle boards are included in the Board Manager esp32 boards.  I am using Arduino IDE 2.2.1.  I included the following Additional Boards Manager URL under Preferences:  "https://raw.githubusercontent.com/espressif/arduino-esp32/gh-pages/package_esp32_index.json".

Then selected the FireBeetle 2 ESP32-E:

And used the default settings for my programs:

Hardware Connections

GDI Display - 18 pin FPC cable

MultiGas Sensor - GPIO Header 3.3V, GND, SCL(22), SDA(21) to Grove 4 pin connector

Battery -  2 pin JST-PH connector VBat, GND; Voltage Monitor on A2(34) - divide by 2 

Optional momentary push button on D7(13) to GND

Pinout

3D Printed Case

For flexibility I decided to mount all the parts except for the sensor in a 3D printed case.  The sensor connects via a Grove connector.  That way it is easy to swap sensors.

Top - TFT Display with slot for SD card

Bottom - FireBeetle2, LiPo battery, Grove connector(I2C), hole for Reset button access, Push Button, cutout for USB-C connector access

Display Test

The first thing I wanted to verify was the operation of the TFT display.

Install display library DFRobot_GDL from github into Arduino IDE.

Upload the TFT sample code from FireBeetle_2_Board_ESP32_S3_Advanced_Tutorial.  The code also runs on the FireBeetle2 ESP32E.

Picture below shows code running on TFT with case open before battery and cover were installed. 

MS8607 Test

Since I had not used the MultiGas sensor I wanted to try an I2C sensor that I have used before as a functional test.  I decide to use the MS8607 PHT (Pressure/Humidity/Temperature) sensor.

I installed the Adafruit MS8607 and Sensor libraries.  I wanted to publish the sensor values using MQTT so that I could display them on a Node-Red UI.  I run a local MQTT server and Node-Red using a RPi4.  

Here are the libraries that I am using:

#include <Wire.h>
#include <WiFi.h>
#include <PubSubClient.h>
#include <Adafruit_MS8607.h>
#include <Adafruit_Sensor.h>
#include "DFRobot_GDL.h"

The picture below shows the readings on the TFT.  Case is complete but still running on USB power.

Results on Serial Monitor:

The Node-Red Flow

Very simple flow receives PHT data on the topic published by the FireBeetle2 using an MQTT input node and displays it on the dashboard using a chart node.

The Node-Red Dashboard

MultiGas Sensor V2

Now to get with the actual project - the gas sensor.

The Seeed MultiGas Sensor V2 board uses four individual gas sensors for each gas type:

1. GM102B NO2 sensor
2. GM302B C2H5OH sensor
3. GM502B VOC sensor
4. GM702B CO sensor

The wiki is here: Grove-Multichannel-Gas-Sensor-V2.

The library on github is here: Seeed_Multichannel_Gas_Sensor.

Here is a picture of the gas sensor operating on battery power:

Results on Serial Monitor:

The Node-Red Flow

The Node-Red Dashboard

Not PPM

The gas sensor interface and display is working as I had planned, but there is still some work to be done.  I am somewhat confused by the results that I am getting from the Seeed MultiGas Sensor library.  The results that I get are not gas readings in PPM but just conversion of raw sensor voltages, so I will need to figure out how to "calibrate" or adjust them.

Here are the library sensor values along with the sensor voltages:

I guess I was confused by a Hackster project that I saw that used the same gas sensor and library: Air Quality Monitoring Station.  

That project called the same library functions to read the sensors but somehow returned "sensible?" values in PPM (except the NO2).  

Something that I will need to figure out.  For now at least I can get a qualitative measure of gas readings, i.e. detect changes.

Test with Isopropyl Alcohol

A quick test using isopropyl alcohol surprised me.  I don't have any active air movement but the time it took for the sensor to clear after exposure to a small amount of spilled alcohol was much longer than I expected (hours).  And the CO sensor is somewhat noisier than before the test.  I'll need to do a lot more experimenting with the sensor responses to different stimuli...