[lug] decoding weather data via radio Was: Time sync w/GPS or radio

[Nate Duehr]

On May 13, 2008, at 9:34 PM, Mike Stanczyk wrote:

> On Mon, 12 May 2008, Nate Duehr wrote:
>
>> You'll have to either find someone who has completely reverse- 
>> engineered the product, or ask Oregon Scientific if they'll share  
>> the modulation type and the protocol with you.  (Doubt it.)
>>
>> Which bit in the bitstream represents what if it's binary?
>>
>> Some "hints" might be who makes whatever RF chipset Oregon uses...  
>> many manufacturers making low-end/low-priced products are going to  
>> often pick a standard off-the-shelf RF module and then put their  
>> own data packets or stream of information over those... so you can  
>> get the modulation type, speed, etc... from the datasheets for the  
>> chipset, but you still won't know which bits are temperature,  
>> humidity, wind speed... nor how they encoded them... that will  
>> require bench time, and a lot of experimentation to reverse- 
>> engineer it.
>>
>> Unless... the magic of Google shows you where someone else has  
>> already done all of that work and published it.  Or if Oregon is an  
>> open company about such proprietary things as their little weather  
>> stations.  Most companies making products that are priced anywhere  
>> from 3x to 10x the worth of the components (most Oregon Scientific  
>> circuit boards are pretty simple) aren't going to make information  
>> easily available about how to intercept their protocol... they'd  
>> rather sell you another receiver, and another for the other room,  
>> and another for the garage... you get the idea.
>>
>> --
>> Nate Duehr, (Since you asked for hams... WY0X)
>> nate at natetech.com
>
> You're spot on in all of this.  So with my weak Google-Fu I've got  
> this:
> A decode of the old digital format
> http://www.tinaja.com/glib/muse151.pdf
>
> And a claimed compatable DIY weather station:
> http://www.techdesign.be/projects/012/012.htm
>
> So my first step is go look at what's being broadcast and compare.

I'd open up the transmitter and look at the board and find the  
transmitter chip (it'll have to be connected to the antenna, easy to  
find because of that) and see if you can find a datasheet on the  
transmitter in the OreSci device online.  If it's a standard  
transmitter, you'll find the manufacturer, and the "mate" receiver  
chipset on whatever manufacturer's website that leads you to.

Then just buy it's "mate" and engineer something like the second  
example above.  (He did a nice job, by the way.  You could probably  
build that one to replace the entire OreSci for a similar price as  
just building a new base-station for the OreSci.

Anyway... receiving their transmitted signal with their already-chosen  
RF chipset is far easier than figuring it out with a spectrum analyzer  
and buying various chipsets until you find a compatible one, or  
building your own receiver (even harder).

If you look at the schematic for the "base station" he only using four  
pins of that device, that's how simple it is... power, ground, digital  
data output, and a pin to connect the antenna.  The receiver is a  
"black box" chip, but the linked datasheet shows that it's Amplitude  
Shift Keying modulated, about as "cheap and dirty" as it can get.

http://en.wikipedia.org/wiki/Amplitude-shift_keying

I would assume the OreSci are similar, but the best way to know is to  
find their transmitter chip, if they didn't "pot" it (covering it in  
epoxy black goo so you can't see the chip markings) or "scrape" it,  
scraping all the markings off.  (Since they're a "fairly big"  
manufacturer, they may also be using a custom made chip, but at their  
price points, I seriously doubt it.)

> Is there software in the ham community that can frequency analyize via
> the sound card?  I've got the scanner of course... ;-)

Yes, but frequency analysis can only take place after you've received  
the signal.  You need a laboratory grade device called a spectrum  
analyzer that has a sensitive receiver and can display frequency vs.  
time, to do the level of analysis you're attempting.

It can be done that way, but if you can find their transmitter chip,  
and then buy the "mate" receiver chip and design around that, it'll be  
a lot easier.  I might say that doing all that might be "fun" but even  
with a bunch of RF experience, I wouldn't go that route... I simply  
wouldn't have the time...

By the way, by the time you have a custom board made, buy all the  
components, spend a month or two learning to program a  
microcontroller, buy all the necessary "workbench" tools including a  
programmer for the microcontroller, etc etc etc... the list goes on...  
you'll have invested FAR more than the amount of buying an OreSci  
replacement.  (Mass production always beats home-brewing in price.)   
You're probably looking at a minimum of $300 to get set up properly to  
even do the project, and months of time.  If you're not planning on  
building anything else, prototyping things, etc... and that's not your  
passion -- it's probably a lot of money to sink into such an endeavour.

If you're doing it just for the fun of it, and it's not breaking the  
bank... then it's worth it.  Sadly, with changes in tech, not even  
most Radio Hams have a properly equipped workbench to work on such  
things anymore... you're headed off into EE (Electrical Engineering)  
stuff that most just aren't willing to learn anymore.  Rewarding and  
interesting if you like such things (I don't want to discourage you  
from trying) but expensive and frustrating if you're just trying to  
replace an "appliance".

To make a Linux analogy... you just walked in and asked how to re- 
write [insert any good working Linux application here] from scratch.   
Sure, you can do it, but do you really want to?  :-)

--
Nate Duehr, WY0X
nate at natetech.com