#diyairqualitymonitor — Public Fediverse posts

My DIY air quality monitor system has been out of commission for a while and I couldn’t be bothered to investigate it

But

I did this weekend, though, and I think I fixed a few issues too

The raspberry pi zero in charge of storing the readings was struggling, apparently kernel panic

So I reinstalled the flash entirely from a recent image, added my tools and my app, all of which took an hour

Everything is back and I iterated a few times on the microcontroller app, and it hasn’t been smooth but it’s been better than last year

Issues as far as I could see were:

• too many WiFi thingies close together in a small room => RF interference
• the bridge node was spending too much time not reaching storage over HTTP => not getting mesh sync => nodes disconnected

So I spaced them out and I added a timed reset

Let’s see how long this lasts!

#rpizero #iot #diy_electronics #arduino #esp32 #esp8266 #sqlite #mesh #DiyAirQualityMonitor

My DIY air quality monitor system has been out of commission for a while and I couldn’t be bothered to investigate it

But

I did this weekend, though, and I think I fixed a few issues too

The raspberry pi zero in charge of storing the readings was struggling, apparently kernel panic

So I reinstalled the flash entirely from a recent image, added my tools and my app, all of which took an hour

Everything is back and I iterated a few times on the microcontroller app, and it hasn’t been smooth but it’s been better than last year

Issues as far as I could see were:

• too many WiFi thingies close together in a small room => RF interference
• the bridge node was spending too much time not reaching storage over HTTP => not getting mesh sync => nodes disconnected

So I spaced them out and I added a timed reset

Let’s see how long this lasts!

#rpizero #iot #diy_electronics #arduino #esp32 #esp8266 #sqlite #mesh #DiyAirQualityMonitor

My DIY air quality monitor system has been out of commission for a while and I couldn’t be bothered to investigate it

But

I did this weekend, though, and I think I fixed a few issues too

The raspberry pi zero in charge of storing the readings was struggling, apparently kernel panic

So I reinstalled the flash entirely from a recent image, added my tools and my app, all of which took an hour

Everything is back and I iterated a few times on the microcontroller app, and it hasn’t been smooth but it’s been better than last year

Issues as far as I could see were:

• too many WiFi thingies close together in a small room => RF interference
• the bridge node was spending too much time not reaching storage over HTTP => not getting mesh sync => nodes disconnected

So I spaced them out and I added a timed reset

Let’s see how long this lasts!

#rpizero #iot #diy_electronics #arduino #esp32 #esp8266 #sqlite #mesh #DiyAirQualityMonitor

My DIY air quality monitor system has been out of commission for a while and I couldn’t be bothered to investigate it

But

I did this weekend, though, and I think I fixed a few issues too

The raspberry pi zero in charge of storing the readings was struggling, apparently kernel panic

So I reinstalled the flash entirely from a recent image, added my tools and my app, all of which took an hour

Everything is back and I iterated a few times on the microcontroller app, and it hasn’t been smooth but it’s been better than last year

Issues as far as I could see were:

• too many WiFi thingies close together in a small room => RF interference
• the bridge node was spending too much time not reaching storage over HTTP => not getting mesh sync => nodes disconnected

So I spaced them out and I added a timed reset

Let’s see how long this lasts!

#rpizero #iot #diy_electronics #arduino #esp32 #esp8266 #sqlite #mesh #DiyAirQualityMonitor

My DIY air quality monitor system has been out of commission for a while and I couldn’t be bothered to investigate it

But

I did this weekend, though, and I think I fixed a few issues too

The raspberry pi zero in charge of storing the readings was struggling, apparently kernel panic

So I reinstalled the flash entirely from a recent image, added my tools and my app, all of which took an hour

Everything is back and I iterated a few times on the microcontroller app, and it hasn’t been smooth but it’s been better than last year

Issues as far as I could see were:

• too many WiFi thingies close together in a small room => RF interference
• the bridge node was spending too much time not reaching storage over HTTP => not getting mesh sync => nodes disconnected

So I spaced them out and I added a timed reset

Let’s see how long this lasts!

#rpizero #iot #diy_electronics #arduino #esp32 #esp8266 #sqlite #mesh #DiyAirQualityMonitor

After all this time fiddling with my DIY air quality sensor, I’m still not sure how to read the sensor readings

Here a little bit about today’s adventure

today’s context

I am sitting on the train and the car is mostly full (about 30 adults). We’ve been at full speed for a good twenty minutes, so anything that needed to settle surely has by now. The train itself was put in service last year or whereabouts, so it’s neither brand new nor crumbling old.

The BME680 sensor I’m using has a three minute self calibration sequence that simply won’t give any readings. Anything I get is after 300s.

sensor readings and doubts

My issue is that at 14 minutes into the ride, the readings were still unsettled. Is that due to the train moving? The sensor not having enough air flow during its startup sequence to calibrate correctly?

At 24 minutes into the ride, the values look suspiciously like default values. Temperature at a steady 25, VOC reading near 25.0 and eCO2 at 500 (give or take a few units).

pressure readings make me more confident

I noticed the pressure sensor reads somewhat different values, sometimes 80 Pa difference in a few seconds even though we are at a steady pace, apparently on even ground, and the weather doesn’t appear to change. But it is very cloudy outside so who knows, maybe our speed causes us to go through more local weather differences than I realize.

However, when the train slows to a crawl, the pressure readings stabilize a lot. They start fluctuating again after we picked up some speed.

the end

I don’t quite understand the sensor’s fondness for eCO2 values so close to the 500 mark. This is a train car full of people, possibly with a working ventilation and filtering system, but I didn’t expect near-perfect or steady readings (for half an hour now).

#bme680 #esp8266 #diyAirQualityMonitor