I don't know if you guys have done much with Wyze, but they offer a wide selection of automation and security products at a cheap price. I like to tinker with them to make them more useful. What I'm working on now is converting their door entry sensor into a general purpose sensor that can take a input from wires.

Used simplysafe before. Their product is easy enough to convert over since it just uses a reed switch. However it hardly ever gets the signal back at the base station so it's not reliable enough for me to consider using. Wyze has this 3 legged hall sensor instead. Three legs on it are gnd, +3.3vdc and an output terminal. Output terminal changes from - to + depending on if a magnet is near it. The state of the sensor changes from secure to unsecure depending on the last voltage it received. In other words just cutting wire from either voltage will not change the state it records in the app. It need a new input of either gnd or + to report a change in state. This can be done easily enough by soldering on an external relay with both NC and NO terminals and letting the external wiring fire off the relay coil. This however isn't practical because it would be too large and draw too much power.

As a stop gap I'm going to try and use TMUX1219 from TI. It should get the job done, but I was wondering if there was an IC out there that anyone knows about that only has 4 legs. Legs would be +, -, output, and "trigger". When the trigger port received + it would switch the output terminal between either + or - like the hall switch on the board does now.

 

Take a look at this youtube video arount 6:20. He's using an opto-isolator to mod the sensor...

WYZE Contact Sensor Microcontroller Interface
0033me

 

I'll have to check that video out. That's the older V1 sensor. Thought it had a reed switch so I didn't even bother watching that video when it came up in search. Reed switch you can just bypass with a wire. If the guy uses an opto-isolator this might be what I'm looking for. Thanks.

 

Took a look at the video. Pretty cool what he did but it's not applicable to the new sensor with the Hall sensor on it. On that one the sensor was "dumb" when the reed is closed it sends the closed signal to the controller, when it's open it sends the open signal. On the V2 it's logical. It goes to gnd when closed and 3.3V when opened so you can't just open and close a wire to trigger it. Need to send either gnd or +3.3v to the pin for it to change states in the app.

Here is a datasheet for the Hall sensor. Pretty sure the one they use is the one with the CMOS output.

 

I would think it's applicable. If you remove the hall sensor you have 3 wires: power, ground, and signal.

The output of an opto-isolator is typically an npn transistor. The base is driven by the lignt. I'd connect
the emitter to ground, the collector to power through some resistor, perhaps 3.3V/100uA -> 33Kohms
or so and connect the collector/resistor point to the signal line.

When I looked at the video I couldn't determine what opto-isolator he used. It's also important to
get the polarity correct (power is +, ground is -), the opto won't work with reversed polarity.

The only other item to think about is power consumption. The hall sensor has cmost outputs
and is optimized for low power usage. The opto-isolator + resistor load will take more power,
which depending on your power source may or may not matter.