Analysing the data with a 433 MHz receiver unit and a logic analyser showed a burst of twelve identical signals, 3.9 ms apart, for each reading of outdoor status. For channel 1 these bursts were located 56.95 seconds apart. For channel 2, 67.05 seconds and for channel 3 84.97 seconds apart.
Each pulse is 0.47 ms wide, with 1.98 ms separating 'ones' and 0.99 ms separating 'zeros'. The last pulse in a 'burst' is narrower, 0.25 ms wide.
Using this information and by varying the environmental parameters the following table could be compiled:
bit 1 - 10 Sensor ID and battery status(?) bit 11 - 12 Channel ID bit 13 - 24 Temperature * 10 in two's complement notation bit 25 - 28 Always 1 bit 29 - 36 Humidity
Something also happens in bit 3 - 7 when the channel changes on the outdoor sensor.