Two seperate versions of Wave plus devices with different readings (one invalid?)

[kvapehe]

Two different W+ devices with different serial numbers, have different readings.

Some differences are ok, but the newest seems to have invalid readings.

One W+ device with a SNR dating back a year ago and one W+ device delivered to me in march 2019 with a newer SNR show different behaviour. According to Airthings support there should be no differences either in the FW or the HW

I have experienced multiple issues with android phone apps , the dashboard and bluetooth syncing so now I have been testing on a PI 3B+
Phones tested Samsung J6, Huawei 10 lite and Sony Xperia, all fails especially after the updates in march. Both the W+ devices were updated with new FW and the phone Apps, even to the beta version on one phone.

Anyway the Pi test follows and it shows something is wrong with the newest W+ sensor readings.

Old device:
Working like a charm
│ Humidity │ Radon ST avg │ Radon LT avg │ Temperature │ Pressure │ CO2 level │ VOC level │
│ 32.0 %rH │ 46 Bq/m3 │ 0 Bq/m3 │ 22.2 degC │ 1011.02 hPa │ 702.0 ppm │ 99.0 ppb │
Long term radiation is like others have stated = 0 so I guess there are some mistakes regarding this sensor, but the other values seem to be healthy.

New device with a much higher SNR, works regarding getting the data, but some sensors are not giving the correct values and the python script reports some default "maximum values"

Newer device: Only Radon ST and Co2 appear to have valid values
│ Humidity │ Radon ST avg │ Radon LT avg │ Temperature │ Pressure │ CO2 level │ VOC level │
│ 127.5 %rH │ 37 Bq/m3 │ 0 Bq/m3 │ 655.35 degC │ 1310.7 hPa │ 536.0 ppm │ 65535.0 ppb │

Some details:

Older W+ device:
rawdata = self.curr_val_char.read()
'\x01C\x05\x00:\x00\x00\x00\x9e\x08l\xc5\xc6\x02g\x00\x00\x00a\x06'
struct.unpack('BBBBHHHHHHHH', rawdata)
(1, 67, 5, 0, 58, 0, 2205, 50543, 706, 114, 0, 1626)
│ 33.5 %rH │ 58 Bq/m3 │ 0 Bq/m3 │ 22.05 degC │ 1010.86 hPa │ 706.0 ppm │ 114.0 ppb │

Newer W+ device delivered to me around march 2019
'\x01\xff:\x00%\x00\x00\x00\xff\xff\xff\xffZ\x02\xff\xff\x00\x00\xff\xff'
(1, 255, 58, 0, 37, 0, 65535, 65535, 602, 65535, 0, 65535)
│ 127.5 %rH │ 37 Bq/m3 │ 0 Bq/m3 │ 655.35 degC │ 1310.7 hPa │ 609.0 ppm │65535.0 ppb │

All 255 values and 65535 values appears to be wrong and the values are derived directly from
self.curr_val_char.read() which calls
/usr/local/lib/python2.7/dist-packages/bluepy/btle.py(196)read()

This tells me either the data is not read the same way with these two serial numbers or the sensors do not report the correct values.

The serial numbers if needed may be provided to Airthings representatives
but I guess an in house test on different W+-devices within Airthings will show the same results
if both new and and old serial numbers are tested.

Note both W+ devices in my test were updated
to the newest FW automatically when tested with the phone app.

[oletorv]

Hi !

Could you please contact Airthings at support@airthings.com ?
Add the S/N of the devices and link to this link and we will look into it as fast as possible.

BR,
Ole A.

[kvapehe]

Hi. Airthings already know about the two serial numbers as I have an ongoing case regarding the app aon android phones. If you represent Airthings I can gladly send the serial numbers to you. If you like to see the support link I can send you that link The only reason I left out the serial number on github is that it might stay there forever Nice hearing from you Ole Andreas. Petter Arild Fra: Ole Andreas Torvmark <notifications@github.com> Sendt: torsdag 25. april 2019 15:41 Til: Airthings/waveplus-reader <waveplus-reader@noreply.github.com> Kopi: Heitman, Petter Arild <pehe11@vaf.no>; Author <author@noreply.github.com> Emne: Re: [Airthings/waveplus-reader] Two seperate versions of Wave plus devices with different readings (one invalid?) (#12) Hi ! Could you please contact Airthings at support@airthings.com<mailto:support@airthings.com> ? Add the S/N of the devices and link to this link and we will look into it as fast as possible. BR, Ole A. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#12 (comment)>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AICRS6OPABDBO5SQ2YZ4W2DPSGYH5ANCNFSM4HIGLNEQ>.

[kvapehe]

Followup on this report.
Airthings have been working on new FWs, and troubleshooting BLE-issues the latest year and
is now in a stage to launch a new FW that may solve many of the reported issues.

What has happened since early spring is that radon-measurements now report both long term and short term values. This part was fixed late summer / early autumn and may be visible if you have updated your FW in august.

When all issues in this report are fixed a summary will be written here.

Until later.

[snyggapa]

I can confirm that long term radon was fixed for me round about 7th August with what I assume was a FW update - at least my PI started to collect it and send to thingspeak on that date.

Looking forward to having BLE fixed because that is still really rubbish when syncing to my phone

[kvapehe]

The long term issue sound right, I first tested it fully around 14. august, and I know Airthings was working on that FW in late July. I also know they are working determined on the BLE-issues for the moment. It seems there were severals issues in the BLE-communication early on, leading to some strange results and it was more prominent on some phone models than others, leading them to think some Android phones had more issues than others.

The BLE is used in a quite advanced mode when collecting earlier data and the collecting of data with
the raspberry pi is done differently

My phone was one of the phones that hade strange results, and I started to make my own waveplus-script to se if the BLE with the pi was more stable. Soon after I ran the script it failed, so I modified it to report everytime it failed a readout, then continue. some try catches, and failure counters.

In some situations I had to disconnect and reconnect, other times I simply had to reread the sensors to get the results, and in early spring I had invalid data comming from the W+ (one of the devices at least)

Since my system had many issues I got more involved in digging into this, and some fixes are son to be tested.

As said earlier. I report back when I know more, and there definitely will be more FW released to fix the discovered issues.