I wish I had added wire(s) for the light before I mounted it on the GDO, but there still is the issue that the best you can do is "toggle" the light rather than explicit on/off.
By the way, I strongly suspect you don't really need to solder two wires to the light pushbutton - just one. One of the two is almost certainly common with one of the two wires for the garage door toggle button. You can look at the PCB routes to determine or ohm it out. This point might be useful if you intend to run wire to a relay not at the GDO itself.
Very true but my plan is to disable all "automatic" light options such as motion and the safety sensor turning it on. I will then use a Hue motion for Lux and motion along with a Aeotec Motion Sensor 6 in the ceiling to turn it on. I could also put in a rule to check lux of the room at night to ensure its off. Definitely trickier than explicitly knowing its on but it is manageable with the equipment I have in the garage.
It does work but there are some differences you will work through. FWIW I have been on HE since March 2018 so my versions of his code are obviously very old as a result. Just looked and I am using v01.02.00 dated 2018-03-02
The cancel and pushover are just there for testing.
Then using the GV I use RM to determine current status and if "open" is requested it will open if closed or stopped. If "close" is requested it will close if it is open or stopped. And I was able to add in an audible warning (using sonos in garage) and plan to add visual for closing (waiting for Zigbee colored bulb to arrive).
Then my arrival/departure/bedtime routines call the actions from these two vice calling garage open/close to the GDO.
Now I just have to depart/arrive a few times and see if it is stable. The multisensors are still being flakey on open/close status, not sure why, but it definitely messes with Zooz GDO (because Zooz relies on single sensor to determine status). Oh and I did switch back to Zooz GDO vice LGK - meh.
As always, I'm very open to any feedback - esp RM stuff - if there is better logic, I'm all for it.
Interestingly - i had to wire them NC/C to the Zen17 - would expect NO/C not sure why.
I nipped the ends off (where screws go) and used hot glue. Mounted a magnet on either side of the door puller then mounted the sensors at the top of the T-track where each lands at open/close. Then I used hot glue to hold the wires to the top of the T-track heading back to the GDO.
I still have to hook up the wireless transmitter to the relays. I am using these: Sensors Zigbee relay board Wireless transmitter (paid more then I should have for someone else to solder 6 wires)
As it happens, my wife bought a wireless remote off Amazon for my daughter. She bought these: After market remote (knock off) which come to about $9.50 each.
Following the tips and advice from others in this discussion, I created my own Zigbee Garage Door Opener Device. I started by using a spare remote control for my Chamberlain MyQ garage door opener. I pulled out the circuit board as shown below:
This board paired very easily with my HE hub. I configured this relay board to be in "inching" mode, which acts as a momentary relay switch (you turn it on, and 1 second later it turns off automatically). So the plan is to activate the push button on the garage remote control circuit board.
Next, I soldered two wires on the garage remote circuit board across the leads to the push button. Then, I inserted those leads in the Zigbee board's relay COM (Common) and NO (Normally Open) connections.
Finally, I plugged in a micro USB cable into the relay board and put everything in a small electronic kit box, with the USB cable sticking out a hole in the box. I then velcroed this box under my office desk (where I've hidden a lot of my electrical devices) and plugged it into a USB adapter for power.
I can then turn on the switch for the relay device from my Hubitat Dashboard to activate the garage remote control. To make it even easier for me to trigger, I use the Android app called Macrodroid to trigger this device thru an HTTP request (using HE's Maker API) from my Galaxy Watch on my wrist. Really cool setup, thanks to all the great ideas and advice in this forum!
I made a 3D printed housing for the ZG-2001 Zigbee Relay Switch. I don't think I can post the STL files in the forum, but if you're interested in printing your own just PM me. I could also print one and mail it to you if you want.
My contact switches have Common, Normally Open, and Normally Closed connections. When and why would we use normally open vs normally closed? I know - obvious, right?
I ask because when I was initially wiring my open and closed sensors to Zen17 I was only able to get them to report using NO. NOW I'm seeing they are failing to report on initial change (closed to open) which is hosing my rules because they don't report until after a full open full close cycle THEN report correctly.
Is the circuit only "tested" for open/closed at certain points?
These are used in different scenarios but in your case you want to use NO, normally open. Think of the button you are pressing, normally it’s not pressed meaning it’s “open”/aka not connecting a circuit. When you press it, it completes/closes the circuit until you release the button.
NC can be useful with security contact sensors. I have a zone on my security system that monitors 4 windows in a room. They are “normally” closed but when one opens (not all), the security system senses that one of the windows is open. It doesn’t know which one but the fact that one is open, it breaks the circuit and thus the security panel knows one is open.
Right, for the GDO contact it is definitely NO and closes to push the button. But my reed sensors would not report at all when I tried NO - they only report when used on NC. ZEN17 doesn't have a "choice" for NO/NC on the sensor side, just S1/C or S2/C.
The real question is why do my closed sensors fail to report open when the GDO opens the first time. Feels like they are sleeping and don't wake up the first time the GDO cycles. And it only seems to happen if GDO has been closed for > 24 hours. Weird.
Inspired with the solution here, I roll up my own garage door switch without any relay. I have been developing this for a couple weeks and document them in the Environment sensor thread. I am really close to install it on my garage. I need to buy one more reed switch. I think it is time for me to share my solution here. Perhaps, it could help someone here.
I notice the remote (at least chamberlain, liftmaster and craftman) is a 3V device. I think it should be safe to connect the buttons to a 3.3V MCU and toggle GPIO to simulate a button press. I have a drawing on what I think about how the button works.
I used a clone arduino UNO which operate on 3.3V. I hook it up to Zigbee Environment Sensor which also act like a Thingshield. Here is how it look like. I transplant the GD remote on the second mezzanine.
Since, it is over kill to use arduino to just push a button, I added a few sensors on the board. I have a couple contact sensors. I have motion sensor and vibration sensor hook up as well. Here is how they look in hubitat.
As pictured, I am able to click buttons on my remote. I can monitor temperature, humidity, pressure, light of the garage and detect power outage in the garage. I am hoping that I can use the vibration to tell me the door is moving. I am able to detect motion inside the garage. At this point, I am able to detect whether the door is fully open or closed with 2 contact sensor.
I am sorry if this is too long. I primarily just want to share that the garage door remote may work without relay. You can simulate the button click with simple GPIO toggle with an 3.3V MCU. Toggling this way is very fast. I set mine to toggle in the order of 10 milliseconds which is probably unwise to do with a relay.
One issue I am continuing to fight with is when I want the GDO to open based on presence and if wife and I arrive together HE sends the "on" command then sends it again - which stops the door from opening. I've put some logic around it based on 2 contact sensors plus my multisensor for vibration, but still don't have it quite right. Works great if one of us leave or arrive, but not for multiple.
Here's a dumb question that might help me fix my GDO issue.
If a virtual switch is "on" and a rule turns the switch "on" does that trigger a 'turns on' trigger? or does 'turns on' have to mean changed from another state to "on"?
I have not gotten to setup my rule yet. My plan in this scenario is to check the contact sensors status. I am thinking to ignore any momentary button push command when your door is "opening" or "closing". A door is in these states when the fully opened and fully closed contact sensor is not engaged. Would that work to solve this particular issue?
