I wanted a wireless, battery operated LCD display to keep track of rainwater tank volume, and decided to repurpose a cheap 434 MHz weather station. The RF protocols have been worked out for Oregon Scientific, Accurite and some LaCrosse units. The latter two protocols are very simple and similar to IR transmission protocols.
The display unit used here is a (discontinued) LaCrosse 9611, which I picked up for $2 with its companion remote temperature sensor TX6U at a thrift shop. See product page here. I had to spend some time working out the TX6U transmission protocol, which is still in use with other LaCrosse weather stations. The protocol is very similar but not identical to the TX3TH protocol described here. The message data is intended to be degrees Celsius*10, with an offset of 500 so in principle, 0-999 can be transmitted. The result is displayed as degrees C or F.
I used a $3 eBay Pro Mini connected to a cheap 434MHz transmitter via pin A3 to send data packets, and it works great! The data rate is very low, about 1 sample per minute, but that is fine for tank volume.
In the attached photo, there are currently 1090 gallons of water in the tank. A demo program to format and send arbitrary numbers in the range -499 to 500 is attached (displayed as -49.9 to 50.0 in lower right corner). I have not exhaustively examined the entire range of data transfer, so let me know if problems arise.
As a bonus, the weather unit tells atomic clock controlled time, predicts the weather and the little guy in the display tells me how to dress for out of doors!
Finally, the Oregon Scientific displays are very nice but the bit timing is absolutely critical, and I had a hard time making a reliable RF connection.
TX6U_send.ino (3.29 KB)
Clever...
Would it not be better to reuse the indoor temp (for tank volume) as presumably the little man is wrong if the temp isn't 10.9 deg outside 
Or even toggle the RWT volume in an out with actual outdoor temp every x minutes and get the best of both worlds....
That is a pretty cool idea!!
I had not thought of creating a new sensor and then having it mimic an existing protocol for completely different sensor so the console can be re-utilized. Maybe alternating that as suggested is also good. It would be funny to see the little man dressing up or down every minute as the temp and tank levels alternated (maybe temp +ve and RWT -ve?).
I will have to rethink (expand?) some of my ideas at Oregon Scientific hacking
Thanks.
Probably best to 'toggle' the indoor temp, as the system probably does averaging over many readings from the outdoor temp?
- & - would be good for indicators indoors as well. (hopefully indoors in rearely if ever <0
Thanks for the suggestions, folks!
The indoor temperature is measured from a thermistor inside the display unit, so there is no obvious way to control that portion of the display remotely.
The display doesn't average over the outdoor temperature readings, it just immediately displays whatever comes over the RF channel.
The outdoor section of this particular base station can apparently accommodate a remote sensor that measures and transmits both temperature and humidity readings, so it should be possible to alternately display more than one quantity. The protocol for the humidity sensor has been posted, so I will experiment with that option and report back.
Edit: @robwlakes: I did successfully create sensors to mimic the Oregon Scientific V1.0 and V2.1 protocols, but the display units I tested were extremely finicky. Even very small timing errors in bit length or intermessage gap would cause the display to stop listening to the remote sensor. I was using Audacity to measure the timing, which is just not good enough. I eventually gave up, but with the use of a good logic analyzer, it should be possible to get them to work.
Cheers, Jim
