Add support for Auriol

[merbanan]

Add support for Auriol

[pm-cz]

Hello,
I have implemented decoder for Temperateure/Humidity sensor I already have (they are labeled SilverCrest and are from year 2008, but have the same protocol). I found your question after I spent most of last Sunday morning deciphering the protocol but the sensor data is the same.

Could you record some data for rainfall/wind sensors with -a -t switch? I have the code for temperature/humidity sensors working and pushed to https://github.com/pm-cz/rtl_433/

However, from the pastebin, are you sure you have the correct sensor (unless this is the wind speed)? The output for temperature/humidity should look like:

Iteration 1. t: 105    min: 104 (20)    max: 107 (264)    delta 10
Iteration 2. t: 105    min: 103 (4)    max: 107 (280)    delta 1
Iteration 3. t: 105    min: 103 (4)    max: 107 (280)    delta 0
Distance coding: Pulse length 105

Short distance: 503, long distance: 1024, packet distance: 2243

p_limit: 105

[00] {0} 00 : 00000000 
[01] {36} bb 4a 30 02 70 : 10111011 01001010 00110000 00000010 01110000 
[02] {36} bb 4a 30 02 70 : 10111011 01001010 00110000 00000010 01110000 
[03] {36} bb 4a 30 02 70 : 10111011 01001010 00110000 00000010 01110000 
[04] {36} bb 4a 30 02 70 : 10111011 01001010 00110000 00000010 01110000 
[05] {36} bb 4a 30 02 70 : 10111011 01001010 00110000 00000010 01110000 
[06] {36} bb 4a 30 02 70 : 10111011 01001010 00110000 00000010 01110000 
[07] {36} bb 4a 30 02 70 : 10111011 01001010 00110000 00000010 01110000 
[08] {23} bb 4a 30 : 10111011 01001010 00110000 

Anyway, I have implemented more Auriol sensors in my code (there are still overlaps between them I have to solve before issuing a pull request here), so you may try and maybe it will be decoded for you.

[marki555]

Hi,
In fact I have the wind sensor. You can see output of "-a -t" in pastebin (I have also one old Silvercrest temp sensor, so maybe some of data is from that one). This data was not recognized by your protocol.

This wind sensor reports temperature, relative humidity, wind direction and wind speed (average + gust). Details of the packets are in the PDF I mentioned in previous post.

http://pastebin.com/hfvpTmge
http://tablo.marki-online.net/gfile001.data
http://tablo.marki-online.net/gfile002.data
http://tablo.marki-online.net/gfile003.data

[pm-cz]

Hi,
please try my modified code in my github repository. I removed some safety checks which prevented decoding of your data (temperature sensor transmission was too short, the packet spacing for wind sensor is different) and changed the processing of wind data. Do you also have the rainfall sensor?

Just a side note - I have problems with decoding (esp. in analyze mode) on RPI. Some parts of packets are lost sometimes, maybe because of interrupt delay issue. But you may be using fixed kernel since I don't see those problems in the output (but if the issue is present, some measurements may get lost).

[marki555]

Thanks, it seems to be working OK. I get wind data every ~30 seconds, temp data every ~3 minutes. Data are correct, only battery warning (in wind data) is bogus. I am running raspbian on RPi with default kernel.

I have also the rainfall sensor, but I am not using it because I have no place where it could be positioned. But I will try to get some data from it this week.

Sensor wind event :
protocol      = Silvercrest Weather (2008)/Auriol (2011)
button        = 0
battery       = Low
wind speed    = 1.40
wind gust     = 1.60
direction     = 45
rid           = 183
hrid          = b7
b7 68 00 e0 60
b7 6f 68 10 00
Sensor wind event :
protocol      = Silvercrest Weather (2008)/Auriol (2011)
button        = 0
battery       = Low
wind speed    = 1.20
wind gust     = 1.00
direction     = 45
rid           = 183
hrid          = b7
b7 68 00 60 e0
b7 6f 68 a0 c0
Sensor temperature event:
protocol      = Silvercrest Weather (2008)/Auriol (2011)
trend         = decreasing
button        = 0
battery       = OK
temp          = 24.9
humidity      = 50
channel       = 1
id            = 179
rid           = 183
hrid          = b7
b7 49 f0 0a a0
Sensor wind event :
protocol      = Silvercrest Weather (2008)/Auriol (2011)
button        = 0
battery       = Low
wind speed    = 1.20
wind gust     = 1.80
direction     = 00
rid           = 183
hrid          = b7
b7 68 00 60 e0
b7 6e 00 90 70

[marki555]

I have now tried to parse the output with perl, I opened rtl_433 as pipe in perl. However when my perl program end, it closes the pipe which causes rtl_433 to get SIGPIPE and it cannot handle it and goes into endless loop. Can you fix it in your repository? This fix is very easy, just add this line at the beginning of sighandler() function:

if (signum == SIGPIPE) signal(SIGPIPE, SIG_IGN);

The issue is that the fprintf function causes another SIGPIPE signal, which calls again the function and again tries to fprintf...

[pm-cz]

Fixed in my repo. It would be interesting to receive at least some precipitation data (even the 0 variant to check if the decoding is correct.

Also fixed the bogus battery warnings

[marki555]

Rain data is not recognized.
I guess it is this:

*** signal_start = 62559269, signal_end = 62819299
signal_len = 260030,  pulses = 298
Iteration 1. t: 101    min: 98 (21)    max: 104 (277)    delta 32
Iteration 2. t: 101    min: 98 (21)    max: 104 (277)    delta 0
Distance coding: Pulse length 101

Short distance: 519, long distance: 1027, packet distance: 2282

p_limit: 101

[00] {1} 00 : 00000000
[01] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000
[02] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000
[03] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000
[04] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000
[05] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000
[06] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000
[07] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000
[08] {36} 21 6c 80 00 b0 : 00100001 01101100 10000000 00000000 10110000

signal_bszie = 524288  -      sg_index = 0
start_pos    = 2430916  -   buffer_size = 3145728
*** Saving signal to file gfile015.data
*** Writing data from 2430916, len 524288

http://tablo.marki-online.net/gfile015.data
LCD shows 0.2 mm, so according to specs I guess it is 0.25mm.

[pm-cz]

I missed that CRC is calculated differently (just) for the rainfall sensor, it should be OK and working now, try the new version

[marki555]

Thanks, good work! Rainfall is working OK now. But the issue with SIGPIPE is not fixed, the sighandler() function is the same as before.

Sensor rainfall event :
protocol      = Silvercrest Weather (2008)/Auriol (2011)
button        = 0
battery       = OK
rainfall      = 0.750000
rid           = 33
hrid          = 21
21 6c c0 00 f0

[pm-cz]

You are right, it got somehow "unedited". It was a part of separate commit to make integration with this repo easier and the commit disappeared. I did probably something wrong on local side and caused a revert to previous state. Hopefully this is the last fix.

[deennoo]

Thank you PM-CZ your code work for my 3 OTIO sensor (sht-10 & sht - 20) and for my Auriol Z32171A-TX

[merbanan]

The referenced gfile015.data gave this with current code.

@0.524288s : AlectoV1 Rain Sensor
House Code: 132
Channel: 0
Battery: OK
Total Rain: 0.25 mm
Integrity: CHECKSUM

[merbanan]

And the linked gfile001-002 does also work. It seems this is implemented. Closing issue.