>>> Join Us in the Fight Against Air Pollution

AirGradient Forum

Using AirGradient Firmware and WeMos with a Ozone Sensor Module?

Has anyone else done this, is this possible to do with the firmware and board? If so, can it also be used with AirGradient’s Online Panel of the sensor diagnostics and can it also be used with Prometheus Monitoring?

I’ve looked into doing it with esphome. I only found one or two i2c sensors that do ozone and they are exorbitantly expensive.

Jeff Geerling has a prometheus exporter firmware that works with the air gradient sensors.

I just came across this guy for ~40€: Gravity: I2C Ozone Sensor (0-10ppm) — Arduino Official Store
Seems reasonable to me.

Cost is a relative aspect, related to performance and to lifetime …
Very interested in the Outdoor Monitor.
Just a ‘what-if’-question related to measurement of gasses:
would it be possible to include inside the housing a sensor for measuring Ozone?
For example DFRobot SEN0321 seems an affordable device with reasonable performance, but perhaps you know better candidates. Considering that not every user will be interested in ozone-info, perhaps a suitable setup should become fitted-for-but-without the sensor, incl. interface to outside world already prepared for handling such interface => plug&play for such sensor.
Another approach (with some appeal for people which already have a dust-sensor in other setup):
substitute the dust-sensor PMS5003 for an Ozone-sensor like SEN0321,
with default-info in the interface for the absent dust-components (as for fitted-for-but-without).
What 's your opinion on this idea?

Exchanging PMS5003 for SEN0321 would impy that instead of UART an I2C is required as interface.
Technically should not be a problem, because SPG41 also applies I2C for interfacing.
Interfacing of an SEN0321 simply is sharing of the I2C-interface with the SPG41.
Obviously with related adaptation of firmware for the package’s processor and for the external interface to ‘outside world’.

Costwise an exchange of PMS5003 for SEN0321 would be an affordable configuration for a gas-measuring cluster.
Because now PMS5003 includes the measurement of Temp & Hum, to get a new integrated package dedicated to measurement of gas-components,
then probably the inclusion of an SHT3x or SHT4x would be appropriate as matching, integrated T&H sensor.
:wink: Don’t rely on the ‘neighbours’, but measure in the airflow for the gassensors and integrate into the common software …

The PMS5003 in the indoor models doesn’t include Temp/Hum sensing, that is only in the PMS5003T models found in the outdoor version.

Achim has some research about the reliability of the PMS5003 and is a large part of why it is used as the default.

Since this is open source, you are free to do whatever you would like with adding/removing sensors, but the official firmware won’t have native support, you’ll have to handle that yourself.

True/understood.
:wink: Have only been looking at the outdoor set, and (too much) orientation towards my own configuration (which already has multiple dust-sensors and a simple gas-sensor MiCS6814).
Therefore more focussing at a ‘Gas-Adjunct’

How is your experience with the MiCS6814?

I saw some studies where the accuracy was extremely bad and I wonder if you made some tests and compared your data from the MiCS to reference instruments?

IMHO the main ‘problem’ with MiCS6814 is that the sensor as output has uncalibrated, analogue value per sensor, which means that the user must ‘invent’ his own kind of calibration.
Being ‘at distance’ from professional calibration references that is not an easy job …
Have myself coarsely done by comparison with available, public RIVM-info.
Implication is that in my setup the data-display from this sensor is just an indication.
On a quiet day like today the graphlines will be flat and ‘uninteresting’.
But at ‘hectic times’ interesting pictures may appear.
Example at this weblink is for OldYear>NewYear Fireworks some years ago on a foggy day,
on which gases and dust are not blown away, but stay in the air.
As stated then, probably both dustmeasurement and gasmeasurement have effects from the high humidity.
[Be aware that under ‘Gas2’ there is no sensordata, just constant lines as placeholders]