Airthings / waveplus-reader

MIT License
133 stars 61 forks source link

runs for 28 days and then stops - and no long term radon #10

Open snyggapa opened 5 years ago

snyggapa commented 5 years ago

So interesting, I am running my own fork of this script on a cron job where it connects, gets the values, uploads to thingspeak (IOT charting package) and then disconnects. It has been flawless for 4 weeks however after what looks like exactly 28 days it has stopped working.

2 things to note - the radon long term value never returns anything other than zero, even though the short term values fluctuate constantly above zero

Secondly after what looks like 28 days the script is returning exactly the same values on every read call for every parameter.

I am wondering if the wave plus has some internal memory history cache that need to be "emptied" in order to allow it to add new values in ? I know it stores the history as phones can connect and back-load the data so perhaps the lack of a phone pulling the history has filled up its buffer?

I can't test this theory as the wave is half the country away - all I have is my PI that I can remote into . restarting the PI makes no difference, and it is genuinely retrieving the same values on every read call

Any thoughts?

Steve

marcelm commented 5 years ago

Not sure if this is helpful, but I remember having read on the Airthings homepage that the normal Wave (not plus) can buffer readings up to one year.

steoj commented 5 years ago

Have done the same as you, modified my own fork and logging to my influx db (reading out on Grafana/Openhab). Have also noticed that the long-term Radon readout never returns anything. Has not bothered me as I'm more interested in the short term reading. Had an issue where all readings apart from Radon and CO2 halted. Temp went to zero and the others stopped reporting. Got the following advice from customer service:

  1. Unpair your device.
  2. Take out the batteries of your device and leave it for an hour. If the power indicator (small flashing light) does not show blue, white or green, please replace the batteries.
  3. Reinsert the batteries.
  4. Reboot your phone.
  5. Pair the device, again. My device had then been running for more than a month - more like three.

Do not know if this is a similar issue as mine but you can try it out next time you are near the device.

snyggapa commented 5 years ago

Actually looked more like 14 days than 28 - I had my dates mixed up. Anyhow I managed to talk a builder through popping the back of and batteries out, and it's back working happily again. It's not been paired to a phone so just the raspberry is talking to it, nothing else

mjs82 commented 5 years ago

I just saw all measurements except CO2 and Radon freeze, exactly as @steoj experienced, but after 10.5 days.

Preparing to power-cycle the device as recommended. However I must say it's a bit disappointing that such an expensive product is apparently going to need regular battery pulls like this.

I hope we can get to the bottom of the cause and come up with a solution (even if that means detecting this condition and sending a "soft reset" command to characteristic b42e2d06-ade7-11e4-89d3-123b93f75cba).

jjanderson commented 4 years ago

I am seeing that the sensors are becoming unavailable in HA (Native NUC build - Intel NUC) after around 1 week. Very dissapointing.... I will reach out to Airthings