Graphs produced by SkyPiAQ, Air Quality Monitor
Copyright (c) 2019 – 2023 PWR Software
Update: 7th November 2021
I designed a new external enclosure using openscad and printed on my own-built velleman k8200, has been up and running for a week or so and been through some downpours, still OK, so switched graph uploads to that device.
Still got 5 more devices running long term tests all looking pretty solid! Like about 99.x% OK. One more note, fundamentally the humidity here is high, quite possibly due to a high-pressure pressure washer “hand car wash” a few doors down the road.
Update: 19th May 2020
The Chinese BME680 has been replaced with a UK manufactured BME680, this particular device again has its own peculiar responses in terms of the raw ohms values reading quite low generally and I think it may be a manufacturing issue in that batch of the sensor.
This test SkyPiAQ device is located in a quite well ventilated outbuilding and so the door may be open or closed. The door is mainly open during waking hours.
The failed Chinese BME680 has been replaced with new (Chinese) one, my conclusion is the Chinese BME680 sensors do vary according to the air quality but their response is significantly different to UK manufactured BME680.
Obviously when doors are closed the readings will be much higher than when the doors are open, this can be seen as usually a rapid improvement in air quality readings when the door goes from closed to open, and a less rapid fall in quality readings when going from open to closed. Also the neighbours have some quite heavy smoking visitors who smoke outside and this can be seen as occasional spikes in both air quality and particulates during the summer months, as during the winter months the actual background particulate levels outside are quite high generally due to a lot of people using open fires or fuel burners for heating and the industrial estate across the road quite possibly burning oil for heating.
Additionally this SkyPiAQ test device is currently testing a Chinese manufactured bme680 sensor, and this is the second one of those that exhibits the property that at certain air compositions it reads massively high/good relative to normal readings which obviously throws off the IAQ calculations by the Bosch library.
I will get some more data on this and then switch back to a UK manufactured device, some of which also have their own anomalies to be ironed out.
The Bosch BME680 sensor I am using in the SkyPiAQ device uses a Bosch provided library to calculate an “IAQ” value. The IAQ/VOC-Ohms values from a BME680 sensor are only really a rough relative guide on how an assortment of VOC’s (Volatile Organic Compounds, which includes alcohol vapor from the wine you are drinking, or the carbon monoxide from your breath, or traffic fumes, or central heating outlet, or cooking fumes etc. ) around the sensor varies, not any absolute index of air quality. For that I need to add several gas specific sensors. Also this Bosch library has not really worked satisfactorily up until the previous version which has improved a few things. A newer version of the library has been released recently which will be tested soon.
The Particulate sensors I am using however do come with a calibration certificate +/- 20% (which is not as bad as it sounds, so at actual 10ug/m3 the reading may be 2ug/m3 greater or less than actual.) All the six particulate sensors of this model I am using have been within tolerance of one another and give consistent readings.