BIOWIN / 2measure No. 270208 - weather station

[pinolec]

Hi,

I try to crack this device:
https://www.mediaexpert.pl/dom-i-ogrod/do-domu/stacje-pogodowe-i-termometry/stacja-pogody-bioterm-270208#reviews
also, this seems to be the same weather station:
https://it.aliexpress.com/item/32830480338.html
This is probably available under many different brands and model numbers.

By default it is recognised as:

time      : 2020-08-17 15:45:48
model     : Smoke detector GS 558                  id        : 19754
unit      : 21           learn     : 0             Raw Code  : 19a555
pulse_demod_pwm(): Wireless Smoke and Heat Detector GS 558 
bitbuffer:: Number of rows: 7 
[00] { 7} 00                                  : 0000000
[01] {96} 55 a5 19 5c 01 27 1f 09 f0 c6 8d 00 
[02] {96} aa a5 98 5c 01 27 1f 09 f0 c6 8d 00 
[03] {96} aa a5 98 5c 01 27 1f 09 f0 c6 8d 00 
[04] {96} aa a5 98 5c 01 27 1f 09 f0 c6 8d 00 
[05] {96} aa a5 98 5c 01 27 1f 09 f0 c6 8d 00 
[06] {89} aa a5 98 5c 01 27 1f 09 f0 c6 8d 00 

cu8.zip

I used
rtl_433 -R -86 -X 'n=Smoke-GS558,m=OOK_PWM,s=450,l=980,g=1499,r=9000'

to collected some data available here: BitBench

This is what I determined:
ID?:hhhhhh hhh TEMP_C:~12d b HUMID:~7d RAIN_TOTAL:~12d 3d WIND_KM:~5d 20b

There probably should be somewhere sign bit for temperature but I had no chance to test it yet. I couldn't find the wind direction and Pressure.

Please advise what to do next.

Kind regards,
Piotr

[zuckschwerdt]

Your BitBench looks good. I've updated the link to invert all bits and use PRE:hhh ID?:hh hh hh TEMP_C:12d HUMID:8d RAIN:12d WIND_KM:8d 4d 4d 4d 1b 7b
All fields align nicely to bytes. For wind direction I would expect 4 bit for 16 cardinal directions or 9 (12?) bits for a full degrees value. But also 8 bits for a checksum at the end. We have 13 bits left, so maybe 5+8? But I can't spot the wind dir.

Can you get readings for at least the 16 cardinal directions? Or even better slowly turn the direction indicator to see if there is a field at the end that changes slowly?
Once we figure that out I'll add the device support.

[zuckschwerdt]

Or perhaps keep the direction fixed (due north should give all 0's) and let other values change so we can see which bits at the end stay the same.

[gdt]

@pinolec Are you still working on this?