For the past few weeks I’ve had sensors in my studio (2nd floor) and kitchen (first floor) and data collection has been interesting.
Some things I’ve learned looking at the data I’ve collected. I’ll post some images to illustrate these in the future, but just wanted to share my notes with other people.
The best news is that my sample set of 2 sensor sets read almost the exact same values.
However, the Sparkfun Environmental Combo board has some issues.
- The temperature it reports is about 4C above ambient temperature due to the head shed from it and the CCS811.
- The CO2/VOC sensor resets itself every time I restart the script. I was reading a CO2 of around 1500, restarted the script, and it reported CO2 of 400 reading up to 700 an hour later. [EDIT: this reset on restart timeout is described in the datasheet for the Sciosense CCS811. In my experience it takes about 2 hours to go back to reading the correct levels.]
- Humidity matches my Stretto sensors, so I’m good with that.
- You can only read the sensors as often as once a minute, so we’re not talking “smoke alarm” useful.
I’m using the Adafruit PMSA003I particle sensor and the PM numbers are believable. I don’t have a good reference for these sensors, but I do see the PPM drop when the HVAC cycles through it’s hourly run-for-15-minutes. I also see them go up overnight while the HVAC isn’t doing much (we have it set for 55F) and down during the day when it’s running the heater.
The biggest thing I’ve been surprised by is the high VOC level after we shower. I would expect it while we’re cooking on the gas stove, but the peak lines up with the CO2 peak after a shower. CO2 goes up from 1300 to 3300 and VOC goes from 150 to 3600. Within an hour VOC drops back to 150 but CO2 only drops to 2100 or so in that same hour. I’m guessing (based on minimal research) that we’re seeing a VOC related to the water being chlorinated (equiv) before it’s sent to our house.
Another good thing is how much the act of making a cup of tea in the kitchen is read by a sensor 20ft away. A 2 minute brew (with water from a dispenser) drive the CO2 and VOC up a bit then they drop a few minutes later.
The humidity sensor reads changes in both of these two cases, but they aren’t nearly as significant as for CO2 and VOC.
I’ve got some new PM measurement tools ordered from Sparkfun, these are name-brand PM sensors and I’ll see how they compare with one another and the Adafruit tools.« The “sensor box” project | sensor box and particle detection »