its normal. it has to do with temperature and ground hydrological pressure and other things. they are not consistent I noticed quite a large variation.
Not sure what you are referring to - can you elaborate ?
What chip is the usbttl? When you scoped it, did you leave a trigger on it for the whole 24hours? The only thing that comes to mind is that it's sending something very infrequently (like every 24hrs). Can't imagine why... especially in the case where you didn't plugin the usb data pins.
Quite strange.
What chip is the usbttl?
https://auseparts.com.au/FT232RL-3.3V-5.5V-FTDI-USB-to-TTL-Serial-Adapter-Module When you scoped it, did you leave a trigger on it for the whole 24hours? The only thing that comes to mind is that it's sending something very infrequently (like every 24hrs). Can't imagine why... especially in the case where you didn't plugin the usb data pins.
Quite strange.
No, as I had nothing to trigger on - I did free-run at .2s/div at humspeed and at 4800baud speed - after that I scanned through all the speeds.
Why would a once/24h trigger affect the whole day of operation. I would expect triggers within each hour ?!
radon levels might average in a year. seasonal dependence too
OK, I wrapped it up in a proper box and my Radon-device is now sending Syslog messages every 15min - I will probably change the interval to 30 or 60 minutes as the meter only changes once per hour
I added rechargeable battery and some logic to monitor battery level. Now the syslog looks like
2024-01-28 18:15:05.001 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (1-0) [80 10 12 11 82 1e 0d 03 6d 00 00 50 00 00 80 a5] 1-day=56, 7-day=59, longterm=59, uptime=4:03:13:30, volt=4180 / USB
2024-01-28 18:30:01.967 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (2-4) [80 10 12 11 82 1b 1c 03 6d 00 00 50 00 00 80 a5] 1-day=56, 7-day=59, longterm=59, uptime=4:03:28:27, volt=4196 / USB
2024-01-28 18:44:59.585 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (3-4) [80 10 12 11 82 19 2b 03 6d 00 00 50 00 00 80 a5] 1-day=56, 7-day=59, longterm=59, uptime=4:03:43:25, volt=4176 / USB
2024-01-28 18:59:57.865 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (4-4) [80 10 12 11 82 17 3a 03 6d 00 00 50 00 00 80 a5] 1-day=56, 7-day=59, longterm=59, uptime=4:03:58:23, volt=4178 / USB
2024-01-28 19:14:57.956 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (5-4) [80 10 12 11 82 17 0d 04 6d 00 00 50 00 00 80 a5] 1-day=55, 7-day=59, longterm=59, uptime=4:04:13:23, volt=4172 / BAT
2024-01-28 19:29:56.999 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (6-4) [80 10 12 11 82 16 1c 04 6d 00 00 50 00 00 80 a5] 1-day=55, 7-day=59, longterm=59, uptime=4:04:28:22, volt=4174 / BAT
2024-01-28 19:44:55.638 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (7-4) [80 10 12 11 82 15 2b 04 6d 00 00 50 00 00 80 a5] 1-day=55, 7-day=59, longterm=59, uptime=4:04:43:21, volt=4156 / BAT
2024-01-28 19:59:53.597 +01:00 10.7.14.152:4444 local0.info [] RADONESP32: (8-4) [80 10 12 11 82 13 3a 04 6d 00 00 50 00 00 80 a5] 1-day=55, 7-day=59, longterm=59, uptime=4:04:58:19, volt=4168 / BAT
Just wanted to thank you all for your work on this project. I picked up a unit on eBay because it was cheap, and searched the model name + "hack" because I wanted better access to data, and this is pretty much the best kind of result one could hope to find from that. I'm just too early, I guess! Good luck.
Stupid question, but has anyone considered bypassing the computer and trying to read off the sensor directly?
Thanks, I'm happy you can use it.
The TTL 3.3V port in the Airthings device is very picky and even small leak currents in TX/RX makes the device stop counting and hence deliver to low numbers.
As you can see above my readings dropped to 0 whan I had a computer connected via a USB2TTL33 converter.
We never found out exactly what happened, but I found that the ESP32 serial port was "accepted" by the device and as I wanted to build a standalone WiFi device I followed that route.
My implementation has the serial port on ESP32 live/active approx. 20s 4 times an hour, and that may be why it does not affect the device.
PS: My battery lasts > 2months on one charge. If I only polled the device 2 or 1 times an hour the "lifespan" could probably be much longer. The device is only updated internally every minute, so polling once a minute should be fine.
Personally, I've lost interest in this and moved on to the next project. Glad it helped those that took interest.
bypassing the computer and trying to read off the sensor directly?
The sensor measurement had some concept of counting pulses (radioactive decay events?) on a GPIO pin. I suppose you could sample that signal directly, but without reversing the full firmware algorithm and calibration scheme, we don't know how to translate it into a meaningful metric.