The new sensor SNZB-04P just arrived.
What I have found is this (compared to the previous sensor SNZB-04):
- The housing of the sensor is very tight. I couldn't open it without destroying some of the snap jonts -- but the housing can easily be fixed to reuse it.
- The battery of the sensor is much bigger (CR2477 with diameter x height: 24 x 7.7 mm) compared to the previous sensor (CR2032 with diameter x height: 20 x 3.2 mm).
- The dimension of the sensors PCB is much bigger (length x width x height: 42.2 x 27.2 x 11.5 mm) compared to the previous sensor (length x width x height: 41.4 x 20.7 x 7.5).
- The new sensor is using an SMD hall sensor to detect the magnet, the old one used a simple Reed contact, which can be reused like described in the design.
- The new sensor has a temper switch which would be very interesting to be used as a second contact oi destinguish the three positions of the window handle. I had a look at the MQTT logs of Zigbee2MQTT. The tamper switch doesn't really work independent as its state always returns to false when the contact triggers -- even if the temper switch is open.
The current approach of the 3D design can not be used because there is no reed contact to unsolder to place it near the rotary pin magnet. Unsoldering the SMD hall sensor is not really an option. So, the 3D design of the model would have to be changed so that the PCB of the sensor has to be move beside the handle (instead to the lower/upper part), so that the hall sensor sees the rotary pin magnet. But this would make the 3D model to wide for the window frame.
Togther with the unexpected behaviour of the tamper switch, I come to the conclusion, that I will not support the new SNZB-04P sensor for the Window Rotary Handle Adapter.