I want to build an autonomous GPRS weather station

Great news!
Reading around on other forums, I found out that we do not even need a receiver! We can just connect the "official" receiver to PC through the audio board to be able to detect the signals and reverse engineer it!

This is the result for an Oregon Scientific sensor (TH138) I have (for another weather station):

I used this sensor because it only transmits temperature values and because somebody else already examined a sensor similar to this one:
http://alyer.frihost.net/thn128decoding.htm

Examining the Auriol sensors will be a bit more tricky as they transmit quite complex data:

Uploaded with ImageShack.us

I don't know which sensor(s) transmitted these data, as all of them were powered; it will be easy to grab just the udometer (by un-powering all the others), but it will be quite tricky to separate data for Temperature, Wind and umidity! :-?

I tried with another sensor I have, a simple thermometer: it should be easier to decode... but I can't!
Can anybody help?
I highlighted values represented by the waveforms.

Uploaded with ImageShack.us

Ok, I got it:

Uploaded with ImageShack.us

Now I'll try with the station.

How is it working out for you?

@jumpjack: it looks like you're making great progress!

Would you mind providing some detail about how you connected the receiver to the sound card?

Jon
Arduino Shield List: http://shieldlist.org

I opened the receiving station, and I located the receiver part: it's a small circuit physically separated from the main; it is connected through 3 wires, labeled as GND, PWR and (unreadable); of course I guessed (unreadable) was the signal. :wink:

So, I connected GND to GND of the earphone cable connected to PC, and SIGNAL to one channel of the cable.
Then I used Audacity to record "sound", using a MONO track: I had to determine by 2 random attempts which wire to connect to to get the signal recorded.
Then it was just a matter of sitting and waiting for signal bursts to come from the thermometer. :wink:

How is it working out for you?

Didn't yet have enough time to examine weather station data (they're very complex, and I seeone short signal every 30 seconds and one long and complex signal every 4 minutes, it's quite hard to figure out what's going on).
Besides, I'm not able to use the RX I bought to get the signals, I'm missing something... Isn't it enough to connect GND, VCC and RX oin these small devices?

Very interesting stuff! I will try to replicate what you did and see if i can come up with anything. I noticed the LED on my anemometer assembly flashes red every 30 seconds, but i didn't know there is a signal sent every 4 minutes.

Weird thing is that not each led flash corresponds to a radio burst sent! :-? Maybe it also indicates "data stored for further processing"? :-?

The "4 minutes stuff" can also be guessed from the instruction manual, which says that first synchronization among sensor and base can require up to 4 minutes. :wink:

Yes, the 4 minutes ringed a bell, i immediately looked it up, the manual says the base station needs approx. 4 minutes to sync with the sensors, but only for the first time you set it up, or when you need to change the batteries. It doesn't say it syncs every 4 minutes on its own. Are you suggesting that it is? Because if that signal is complex as you say it is, than that's got to be it.

"Data stored for further processing" could mean only one thing, it is stored in temporary memory and used to calculate the average values of wind speed, wind chill and possibly everything else measured.

Yes, the 4 minutes ringed a bell, i immediately looked it up, the manual says the base station needs approx. 4 minutes to sync with the sensors, but only for the first time you set it up, or when you need to change the batteries. It doesn't say it syncs every 4 minutes on its own. Are you suggesting that it is? Because if that signal is complex as you say it is, than that's got to be it.

I only saw that the 4-mins signal is 3-4 times longer than the 30-secs signal, didn't examine it.

"Data stored for further processing" could mean only one thing, it is stored in temporary memory and used to calculate the average values of wind speed, wind chill and possibly everything else measured.

It's just my idea, I didn't read it anywhere. Don't know if there is onboard processing in sensor.

About the gsm maybe u should check

http://code.google.com/p/sserial2mobile/
and if it has to be a module, i think this 1 is 1 of cheapest, i might be wrong
http://www.cooking-hacks.com/index.php/arduino-gprs-module.html

Thanks for that post from HackADay, it will help. Sserial2mobile i found a month ago and i'm looking into it. And the Arduino GRPS module is indeed one of the cheapest i have seen yet, i think i will get that one.

GPRS modems for 25$ is a very cool news, thanks! But I see I should connect the modem to the arduino through TX and RX pins of a pl2303 chip, which is just an USB-serial converter; as such a converter costs around 10$, should we guess that some GPRS serial modem exist and are sold for 15$ ?!?

GPRS serial modems do not come cheap, they cost around 130$+ or even more.

http://www.smsco.it/tomcat/en/buy/show.do#modem

Funny. :-?

Here's some cheap GSM/GPRS modem modules:

WaveCom q2403a 1800/900mhz GSM Module GSM Modem GPRS Modem dev kit
http://www.no1electric.com/index.php?main_page=product_info&cPath=11&products_id=262

RIM 1802G 1800 MHZ 900 MH GSM GPRS Modem Module
http://www.no1electric.com/index.php?main_page=product_info&cPath=11&products_id=261
http://cgi.ebay.com/RIM-1802G-1800-MHZ-900-MH-GSM-GPRS-OEM-Modem-Module-/290276292642?pt=LH_DefaultDomain_0&hash=item4395d0b822

LinkSprite GSM/GPRS SM5100B Modem Module
http://cgi.ebay.com/LinkSprite-GSM-GPRS-SM5100B-Modem-Module-/270604215104?pt=LH_DefaultDomain_0&hash=item3f0144a340

TMAS GSM/GPRS Wireless Modem Terminal TMA-M55i 1900MHz
http://cgi.ebay.com/TMAS-GSM-GPRS-Wireless-Modem-Terminal-TMA-M55i-1900MHz-/150534453298?pt=PCC_Modems&hash=item230c8d7832

ENFORA GSM2218-01 MT-GL GPS/GSM/GPRS QUAD-BAND MODEM
http://cgi.ebay.com/11-ENFORA-GSM2218-01-MT-GL-GPS-GSM-GPRS-QUAD-BAND-MODEM-/370468649136?pt=GPS_Devices&hash=item5641a6f8b0

What about the wavecom at the end of this page?

Any datasheet available?

The last being the Wavecom INTEGRA M2103A - WM12816, correct?

In 2009 Wavecom was acquired by Sierra Wireless, they have a link on their website for Legacy products:

http://www.sierrawireless.com/productsandservices/AirPrime/Legacy_Products.aspx

under legacy Integra modems there is a link to support and documentation pdf:

http://www.wavecom.com/modules/movie/scenes/products/index.php?fuseAction=download&key=fe9844deede292248eb870ee348ccff4&file=020_Hardware%20documentation/030_Specifications%20and%20User%20Guides/WA_DEV_M2106_PTS_003-001.pdf

Just capture signal isnt good enought without knowing what exactly system send.

"Wind" device basicly have 4 sensors :

  1. Wind Speed
  2. Wind orientation
  3. Outside temperatura
  4. Outside humidity

I did some research becouse I need that device to send data via 1-wire network, but finaly I decide to use both information (I like wide LED screen :slight_smile: )

What sensor is used :

  1. Wind speed (reed switch)
    Half period on / half period off (not so usual but O.K.)
  2. Wind orientation (not sure at the moment)
    From outside device come 7 wire, 2 of them are for Wind Speed , so
    we have 5 other , 16 orientation of wind , that mean 2^4, 1 wire in 4
    out maybe digital/or analog
  3. Outside temperature (Normal thermo resistor)
    Analog value
  4. Humidity sensor (Low lewel)
    Measuring resistance beatween two cooper field

Rain gouge
Standard dual reed switch system

If I am on your place I will use rein gouge for testing and alayzing signals.

When I finish 1-wire implementation I will send more info on forum.