Tuesday May 23rd 2017


Z-Wave Doorbell – a hardware hack

While waiting for a real Z-Wave doorbell to come to market later this year (from Philio and Aeon Labs), I’ve been searching the forums and have found several examples of Z-Wave doorbell hacks. This forum thread at SmartThings was especially interesting because the author had discovered that the magnetic field of a doorbell solenoid is strong enough to close the built-in reed switch of a Z-Wave door/window sensor. Then he posted a photo of a big, honkin’ Z-Wave door sensor strapped to his doorbell.

Being a fan of the tiny Fibaro door/window sensor, and finding plenty of room inside the Heath/Zenith doorbell that I had picked up at Lowes a couple of years ago, I estimated that I could fit everything together nicely. Z-Wave FibaroDoorbell-e1426227221201.jpg
Unfortunately, that nice, big space where I had planned to mount the Fibaro sensor, where the sensor’s reed switch could be next to the solenoid – that space turned out to be next to the solenoid for the back door, not the front door solenoid. Also, the metal bracket that holds the solenoids effectively blocks the solenoid’s magnetic field, and there was no way that I could get the Fibaro sensor close enough to the front door solenoid to get it to trip the Fibaro sensor. BackDoorSolenoid-e1426227517834.jpg

Thinking that I might have to wire a separate relay in series with the doorbell, with its own contact closure that I could connect to the Fibaro sensor, I ran to Radio Shack where I found something that I hadn’t expected.

RadioShackReedRelay-e1426227444461.jpg Radio Shack part number 275-0233 is a reed relay with a 12-Volt DC coil. My doorbell has a 20-Volt AC coil, but I thought maybe with a diode and a resistor I could get the Radio Shack relay to work. Then I got to looking more closely at the Radio Shack reed relay, and I thought: I’ll bet there’s a reed switch inside that DC coil that looks exactly like the reed switch inside the Fibaro sensor. I briefly considered the possibility of just de-soldering the reed switch from the Fibaro and threading that reed switch inside the Solenoid bracket on my doorbell, but I thought the Radio Shack relay would be easier to work with. If the forum poster over at SmartThings was correct, I wouldn’t need to connect the Radio Shack relay’s DC coil to anything; I could rely on the magnetic field from the doorbell solenoid to close the reed switch. And now that I had the reed switch in a smaller package, there was a better possibility of mounting it close enough to the front door solenoid.
Following the schematic on the Radio Shack package, I soldered a lead to each end of the the reed switch:
I removed the unused back door solenoid from the mounting bracket, and I hot-glued the Radio Shack relay inside the bracket, close to the front door solenoid. I also coated the exposed leads of the relay with hot glue (heat-shrink tubing would have been neater, but I was feeling lazy). RelayHotGlue-e1426227317560.jpg
The Fibaro door/window sensor was also tacked into place with hot glue. There’s a tamper switch on the back of the Fibaro sensor, and I have had problems with tamper switches on another brand of sensor: when the tamper switch was not compressed, the sensor would not go to sleep, and it would run down the battery. So with the Fibaro sensor I slipped a thin piece of plastic in between the back of the sensor housing and the tamper switch in order to keep the tamper switch compressed. FibaroMounted-e1426268708859.jpg

I hooked it all up, connecting my reed switch leads to the Fibaro’s external input terminals. Pressing the front doorbell button, I got a response from the Fibaro sensor, but it wasn’t quite what I expected. I realized that I was dealing with an AC solenoid, which meant that the north/south polarity of the magnetic field was changing constantly. The reed switch doesn’t care about the polarity of the magnetic field; either way the switch will close, but at the zero-crossing of the AC current, there is no north or south, and the switch will open momentarily. This results in rapid-fire events from the Fibaro sensor. Fortunately the Fibaro sensor has a configurable delay that will maintain the alarm state. I found that setting parameter 1 to 32 seconds filtered out the unwanted chatter. I also had to change the “input type” on parameter 3 from its default setting of “normally closed” to “normally open.”

Now when somebody rings the doorbell, Z-Wave can trigger an event that will turn down the music, and I can run to the door to see who left a plate of cookies.

Twitter Feed