About brxnet.org air quality station

Design/Construction/Iteration Blog

Dabblings in dust detection: below is a record of some of my ideas, research, device construction, and iteration (totally devops) as I have looked into this fascinating area of atmospheric air analysis. Graphs here.

2018-08-04

Yesterday we had a few points where our nearby sewage works was upwind of us and the whole area stank. My MQ135 sensor picked up the atmospheric pollutants reporting values of 24-25 as compared to the usual 19-20. I don't know about the units these things report yet, but I am delighted that it responds to these sorts of concentrations. Also, given the range of gases the MQ135 responds to it is difficult to know exactly which pollutant was present, however it was there in detectable quantities.

There didn't appear to be much response on the readings from the SDS011, but strangely the PPD42NS showed a noticeable uptick in readings that corresponded directly with the periods of bad smell. I have been given to understand that the SDS011 is superior to the PPD42NS so this bears further consideration.

2018-08-01

I switched over to a NodeMCU version of the project in order to link in to Luftdaten.info.

Graphs etc. available here

2018-07-21

The Arduino was getting pretty hot with the 12V additional input. I gather that as the board changes the voltage down to 5V a lot of excess heat is generated.

I have removed the extra 12V power supply for now and will look into sourcing a lower voltage supply going forwards.

2018-07-20

Click on the pic for a higher resolution version. I got down to some building earlier. Note the additional supply (not just the computer's USB). This is to accommodate the power draw for the MQ135, which appeared to lower the 5V rail on the Arduino as it attempted to heat the sensor. With the additional power the rail is back to 4.9+V and the MQ135 sensor is 10°C warmer.

The current setup includes the Arduino Uno, an MQ135 sensor (currently being burnt in), and a PPD42NS particle sensor. Both sensors are sensing and logging data to the serial port where it can be read off. Code from here and here.

While setting this up, I came across the following useful pages:

Next step is to get some drain pipe pieces to form a weather proof enclosure.

2018-07-18

Thanks SP for the pointer to this NodeMCU Luftdaten-type project.

2018-07-17

Discovered an analysis of several low cost off the shelf sensors, Ben Tan (2017). Laboratory Evaluation of Low to Medium Cost Particle Sensors. UWSpace. I've had a skim read. He gets into some of the maths behind it all, and describes the underlying physics. The evaluation is a bit scathing of the field, but useful.

2018-06-16

Discovered Eastbourne Air, also on Twitter at Clean Air Eastbourne.

2018-06-04

So it looks like the SDS011 is the preferred way to go ATM for the particulate sensor. Reasons:

That still leaves the gas sensors to decide on. I have the MQ135 sensor, but I need to evaluate if the following are worth having on board specifically because of waste water plant exhaust gases (there is a strong unpleasant odour from the Eastbourne Waste Water Works when the wind is in the wrong direction), and because of a multitude of indeterminate environmental influences in a Jakarta gang. The broader the spectrum of sampling the better placed we will be to understand what is going on.

... and there are many others.

2018-06-03

On further reading discovered that Luftdaten.info are now recommending the (more expensive: c. £20) SDS011 particulate sensor. Need to investigate for the rationale, and to discover whether they are better than/as good as the laser-based PMS 7003.

Grove box at Seeedstudio.com

This evaluation of the PPD42NS at PLOS ONE. Note the concern over accuracy at high levels of particulate concentration.

This description of a PPD42NS-based air quality monitoring kit at howmuchsnow.com.

This blog post about calibrating the PPD42. In the comments (May 15, 2016 at 7:37:20 pm) A.J. recommends the SDS011 as being laser-based - which matches the Luftdaten.info change.

2018-06-02

Came across an article on air quality on the Grauniad which referenced Luftdaten.info and a Friends of the Earth testing kit (donate £15-£50 and receive an NO2 kit, "including a full specialist lab analysis of your air monitoring tube").

This might help with the calibration. Luftdaten.info has a standard setup they recommend (including a simple idea for weather proofing the sensor array), and they are a national project run by a German university. Encouraging to see they use the PPD42-type sensor. And FotE will perform a lab test on an NO2 sample from the site - although it costs minimum £15 and the test runs for 2 weeks.

2018-05-30

Bought some resistors, some 10k potentiometers, MQ135 sensor, PPD42 sensor, and a 16x2 display.

How will I calibrate the sensors? Do I need to buy (expensive) pre-calibrated sensors?

2018-05-29

Looking into this again. Investigating particulate sensors, espcially whether PPD42 is worth using. The aqicn.org website has some views on the matter. They seem to like the "Plantower" PMS series because it is a laser based sensor (not simply optical cf PPD42) and can properly distinguish between 2.5 and 10 micron particles. The Plantower PMS 7003 (they quote $20 USD) appears to be the latest.

2017-06-30 (ish)

Began investigating the idea of an air pollution monitor for use in UK and Indonesia. Motivation is understanding polution levels in a "gang" (small street) microclimate in Jakarta Indonesia, and understanding pollution levels close to a waster water processing plant in Eastbourne UK. Budget is a major factor. Considering something based on a Raspberry Pi or Arduino or both. Want to be able to read local values at least, but really want to automatically upload to the web. Currently looking at:

Circuitdigest.com's Arduino idea using

Contact

If you'd like to get in touch you can send an email to an address where the first part is airquality and the second part is the domain of this website.