I also want to make an air pollution meter. I read in the Firmware Flasher section that a microUSB of less than 1 meter is required for installation. I found one at home that is between 95 and 100 cm. Is this appropriate?
I would have another question, that I need to install the NodeMCU driver on my computer first, and then download the appropriate driver from the firmware flasher, which should be installed on the NodeMCU? Or exactly how is it? Maybe there is a video of the steps described on the page?
You only need to install the driver for the NodeMCU once. Then the NodeMCU (or better the com port) should be selectable in the Flashing tool. If the port isn’t shown in the Flashing Tool then you may need another driver or the USB connection has a problem.
The “driver” that is installed on the NodeMCU is our firmware. The firmware is automatically downloaded by the Flashing Tool. So you don’t need to download it manually.
Your USB cable should be okay, if it’s also able to transfer data.
So the steps are:
install driver for the NodeMCU
connect NodeMCU to your computer
start Flashing Tool, the com port should be preselected
Would I have another question that a NO2 and CO2 sensor can be connected to the base set? If so, how? Does the data from these sensors appear on the map?
There is actually no NO2 sensor known to us that could measure values low enough when used outdoor (20 ppb(!)).
We are testing the Sensorion SCD30 as a CO2 sensor. It has an I2C interface while most other CO2 sensor have UART. But the UART for the ESP8266 is implemented in software. This is very unstable when there is more than one UART (our first UART is for the PM sensors).
What about Alphasense NO2-B43F with proper shielding, dedicated low noise power supply and good 16-bit ADC? Of course in pair with AFE from Alphasense (000-0ISB-02).
All readings at sensor temperatures above 30 ◦C are discarded to avoid non-linear temperature dependence of the electrochemical NO2 sensor (see Sect. 4.4)
But highest NO2 values are in summer … So most of the interesting values are discarded.
Each device is mounted near to a calibrated station for several days. This is done again after 2 months.
Who should organize this (with probably much more sensors than the 16 in the cited project)?
I don’t agree with “Hey! It’s too hard to solve. Let’s do nothing!” approach.
I will try to figure it out by adding some kind of air conditioning. Peltier cooling and heating element, maybe chemical water vapor removal. Cross reference with Alphasense OX-B431 ozone sensor. Or maybe I will use machine learning to correct measured values. In this case poor data is way better than no data at all.
SDS011 isn’t perfect either, yet we still use this sensor.
It’s not do nothing. We had some samples of the Alphasense sensor for testing. And the dependency not only from the “absolute value” of rel. humidity but also from the speed of change isn’t really easy to solve. Your heating and cooling or chemical water removal (chemical? and how often must this be changed) won’t really solve this.
I’m highly intereted in measuring NO2 Immissions. Did you make any progress related this? Have you ever tried a nonlinear regression like Random Forrest to estimate the concentration?
