"Soft Bypass function
The Soft Bypass function allows me to over-ride the status of the hardware bypass switch when needed. When Soft Bypass is engaged, the bypass button's LED flashes to remind me Soft Bypass mode is enabled."
These sort of things usually end in tragedy. I worked at a chemical plant that had an explosion where the safety systems (SIS) were bypassed. The engineers were trying to light off a boiler and were having trouble. The safety system had a 30 minute time delay between unsuccessful light off attempts to allow the fuel to disperse. The engineers bypassed the time delay and attempted multiple unsuccessful light offs in a row which flooded the boiler with fuel. Finally, it lit and went boom. Those near the equipment in the field were all killed.
Hope you don't have any pets, children, or slow moving adults around.
The "bypass" system seems to only suppress notifications about the door being open. This project doesn't appear to be coupled to any garage door control systems; it can't open or close the door, it can only report on the current state by ringing the doorbell or sending email. It seems that even if the system grossly malfunctions, you're left with a perfectly normal garage door opener. So the lecture about safety is probably not applicable here -- more likely is that the thing catches on fire, or gets the Google account banned after ignoring some rate limit.
I added automation to my garage door, but not before calibrating (and carefully testing) the force limits on the motor, and installing (and carefully testing) an infrared trip sensor which integrates into the motor.
When a garage door doesn't close when requested (or does open when not requested) it will spray notifications to everyone's phone, but it won't try to close the door again.
yes, naming is hard. the gadget is not a garage door control but purely a mgarage door monitor. so the so called bypass is purely related to the monitor. the garage door control, if any, with its safety features remains completely stand alone.
that to me looks like an ok architecture from the safety pov, where the gadget behaves as an eye and maybe a finger pushing the control remote, so the safety is delegated to the garage control.
to state the obvious, I'm fully with you that there should be no override bypass inactive whatever on any kind of safety sensor, and the sensor should always be fail safe, i.e. if a sensor fails the motor stops, if an electronics compinent fails, the motor stops, if not locked shut it should be possible to manually completely reopen the door, etc.
You have this exact same thing with airplane engine starts. Add fuel at the wrong time and the whole thing either exceeds temperature limits and has to be completely taken apart or you cause a fire.
The most common cause of destroyed engines is exactly like you describe for the boiler, it didn't start the first time and someone ignores the cooldown limits and/or cause of the failed start (e.g. not enough power in the batteries to get the turbine up to minimum speed before adding fuel, try again now with even less power because you used some on the first attempt)
> These sort of things usually end in tragedy.
> Hope you don't have any pets, children, or slow moving adults around.
What aspect of a garage door status monitor do you consider a safety item? This has nothing to do with the various safety systems built in to garage doors (IR sensors, pressure sensors, etc) - it just alerts the owner if they forgot to close the garage door. The bypass silences those reminders.
If you're interested in this sort of thing, but lack the time for a hobby project, Chamberlain makes a somewhat-smart device that will do this for you (AND allow you to remotely open and close the door).
We had a terrible problem leaving ours open, and I got one of these to help solve the problem. It's been pretty great, though it does require a phone app, network connectivity, and an account with Chamberlain. OTOH, it's HELLA cheap -- like $25.
I don't know if Amazon links are okay here, but that's where I got mine. The item name there is "Chamberlain MyQ Smart Garage Hub - Wi-Fi enabled Garage Hub with Smartphone Control, Model MYQ-G0301".
There's definitely that possibility, which is exactly why I won't have similar "smarts" in my actual door locks.
There's not much easily steal-able in my garage, and the door from my house to my garage is a heavy fire door with a deadbolt. The Chamberlain device adds convenience and nagging when we DO leave it open, but because of these other measures isn't a major part of household security like a hackable doorlock would be.
If someone could come up with a "Local Only" networking stamp, like how you see "Organic" on foods, I'm sure we could create a market for smart devices that don't reach out to a third party server.
Another option for those that don't want to do hardware hacking is to use a Z-wave garage door sensor and a smart hub that allows programmable alerts (like the Hubitat). Then you can set up a rule that alerts you if the door is open too long.
Yes, although their MyQ cloud platform unfortunately has a very transient reverse-engineered API due to the lack of an official option, AFAIK. This seems to be a regular issue for users that set up an integration between MyQ and a unified platform such as Home Assistant.
Seconded. I tried some of the consumer devices that ran over Z-Wave before OpenGarage. They worked fine but would invariably require a reset after a couple months. This meant I couldn't really rely on this and leave my keys at home. My OpenGarage unit has been super reliable for 2+ years.
Have one, works great and doesn't depend on the cloud. Nice that it can tell the difference between door open/closed, and when closed it can tell if there's a car/truck present.
Throwing my vote behind this. It's extremely slick! I haven't had any issues with it working or reporting correctly and it ties into Home Assistant quite nicely!
Raspberry Pis are great for this kind of DIY IOT hacking. I have a few that monitor my plants for soil moisture and alert me via IRC and Pushbullet when they get too dry.
Very cool homebrew solution. If you need the feature but don’t have time to homebrew it, Chamberlain has a piece of hardware (MyQ) for $30 (occasionally goes on sale for $17-$25) that enables you to poll garage door status and control it remotely using their app and API. Uses a sensor you attach to the garage door, talks using RF to the opener. Some lifter units even have this built in now.
I highly dis-recommend this product, particularly if you want to use it with something like Home Assistant.
First of all, their app usually takes about 30-60 seconds to load up which is pretty frustrating when you're waiting outside your garage and trying to get in, just waiting on a spinner as it has to go talk to some cloud server somewhere.
Second, I use Home Assistant for pretty much everything, and Chamberlain has a habit of breaking their API every few months. From what I've gleaned they seem to change how they allowlist user-agent headers to try to intentionally break things like Home Assistant.
Eventually I got tired of it constantly being slow and/or not working, and I threw it away. I replaced it with a programmable Z-Wave contact relay (Zooz ZEN16). It works much better now.
Lastly, the newer Chamberlain motor units actually don't use simple push buttons anymore, so what I've done isn't even an option. I believe it's some kind of PWM over the wire to the wall-mounted buttons. I would recommend not buying any Chamberlain product because of their anti-consumer behavior.
I just timed it, and it took 4 seconds from opening the MyQ app to being able to open my garage door. The only time I experience the load times you’re talking about is when I’m connected to my house WiFi far enough from the house that I don’t really have any internet connection at all anyway.
MyQ is probably the best $30 I’ve ever spent.
I’m under no illusion that that APIs are appropriately handled, or that it’s great for all workflows. But for me, it delivered the expected functionality, no RPi hacking required.
From what if seen, most modern openers have lost the ability for a simple button to open/close the door.
If you solder, even just a little, you can attach a pair of wires to the board in the wall mounted button so you can control it from whatever project you are doing. I using pins on a pi to flip a relay and complete the circuit as if I pressed the button.
Linus of Linus Tech Tips had a video [1] where he tried just that on a new Chamberlain opener with MyQ built in, and couldn't get it to work, the wires would act as antennas and trigger the door randomly, just from stray RF.
The button panel for my ancient opener runs off of two wires, and each button is connected through a different value resistor, so I'm assuming the way it works is it measures the voltage drop to tell if you pressed the door or the light button. But, just connecting the two leads from the opener is enough to open or close the door, and I never use the built in bulbs on the opener anyways.
I did this. Bought a LiftMaster 883LM button for mine, and soldered an extra wire to the actual pushbutton contact inside. This was cheap so I wasn't really worried about breaking it, plus it is much simpler (easier to solder) than the "big" button that has an LCD and motion sensor. Bonus is I have an extra button to control the door mounted at the front now.
I did this about 8 years ago, and hooked it to an Insteon network, which is ultimately connected to my HomeAssistant setup. If I was doing it today I'd probably do the same soldering but wire it to something I could run tasmota on (and control via mqtt).
For myq if you have iPhone there are “short cuts” available to quickly open the door or close it, it is extremely fast and efficient. Their app is useless and only thing I use it for is for notifications if the door has been open for 5min+
> lets you check garage door status and control it remotely using their app and API.
Presumably by bouncing everything through some anonymous Alibaba cloud ip addresses, effectively granting permission to the Chinese government to open your garage door whenever they like... :sigh:
I’m sure if China wants in my garage they can figure out a way with or without this device
It’s quite nice though. You can even have Amazon drop their packages inside the garage through some single use token or something. I for one have way too much value of tools and whatnot in my garage to grant that trust but I wish I could.
MyQ works well with their stuff but is horrible for integration.
Chamberlain deliberately changes the API every so often just to break folks. Sometimes it comes with a ridiculous excuse, sometimes not.
They are up to v6 or v7 now. It is really because they want the companies (Samsung, Crestron, Control4) to pay them as well.
Compare this with AladdinConnect that comes with the Genie lifts - they have broken people approximately zero times.
They also have a unit that I installed that you can tell it close when no one has been passed the sensor in x number of minutes. It has been great when we've forgotten to shut the door at night. No need to monitor anything yourself.
My fiance worked on the myQ team at Chamberlain. The app integrates with other home IoT devices as well, providing a single app to view / control a lot of the popular ones.
We had 5 garage door openers in our apartment at the beginning of the work from home period for her testing.
A bit tangential but it seems that architectural trends over the last couple of decades has been to do away with the standard human size door that used to be common in garages. Probably 90% of the times I open my vehicle size door, I could have used a human size door instead. I was planning on building a shed in the backyard, which would eliminate most of these door opening events, but then lumber prices went insane so maybe by autumn the time for that will be right.
I'm curious as to why human size doors in garages have become so unpopular in residential architecture. Security concerns? Energy efficiency? Streamlined appearance? Cost cutting?
I went with an uber-simple solution of a wire running through the roof cavity from the magnetic switch to a big flashing LED in the ceiling by the front door.
For non-native smart devices, this type of automation/monitoring could be implemented with a combination of "Bond Home" and Home Bridge (running on a raspberry pi). I have had success with connecting several dumb fans and lights and accessing them via the native HomeKit interface. Even the native iOS home kit automations work with it as well.
I haven't tried it with a garage door, but I do see the home bridge community has built some plugins with some garage doors.
If you want to control non-smart wireless home appliances, you can actually make the Raspberry Pi GPIO do it without the Bond Home.[1] It’s a huge hack, to put it lightly.
I don’t think the OP is actually doing any sort of control though. I personally would recommend a Zigbee door sensor as part of a Zigbee + Home Assistant (open source) automation system which gives you the flexibility of not having to run wires and not having to write code to configure the automations.
Not sure a battery-operated device is even necessary here given powering a garage door motor requires significant power and therefore you're likely to have mains nearby.
I built one of these once. I even figure out a way to remotely trigger the garage door to close. I programmed it so that if I left it open for too long it'll message me telling me, and if I still stays open too long after that it'll be passive aggressive and say something like "I hate having to do all the work for you, but I closed your garage door because you forgot to."
i tried another contact sensor (sonoff) before trying out the aqara/xiaomi sensors - holy crap what a world of difference in build quality and response time.
I have one in our postbox flap to detect when we get mail - the sonoff needed 2-3 seconds to detect anything being open (and thus would miss the event) - and the flap being closed by then made it harder to send out the signal. The aqara's are almost instant and send out while the flap is lifted.
I've been working on a similar project of my own. I've assembled the hardware and enough guides that I'm just about ready to put it together. My solution uses an ESP32 for control/wifi and MPU-6050 tilt sensors to measure how far open the doors are.
Funny I started building my own project for this as well! One wall I hit was trying to have it send the radio signal to open or close the door, could never figure that out (using a 315mhz RF transmitter and an Arduino). Will take a look at this!
Many garage door motors only accept a dumb trigger and don't have any mechanism for sending explicit "close" or "open" signals.
I would avoid basic RF wherever possible; stick with a wired connection between the motor and your microcontroller, and use 802.11 wifi for communications. This is where the ESP32 (or ESP8266) makes perfect sense.
Why do you say avoid RF? I was trying to create essentially a new “opener” button like the battery-powered ones that came with the motor. It’s pretty old so I don’t think it’s doing anything too complicated security-wise. I do see wired being easier but I was driven to figure out RF. And yea I wouldn’t be able to do a specific open/close event but the reed switch would tell me which state the door was in before sending the toggle signal
Mine only has the single trigger, but also has two limit switches at either side (it's a screw drive). I figured out that by flipping the limit switches I can control the direction of the door using just the single trigger button. E.g., if I short the open limit switch and send the door move signal, it will always close.
'ohd' monitors specified GPIO pins on a Raspberry Pi and responds based on changes to the status of those pins. It sends email (or SMS) messages to numbers in its config file.
These sort of things usually end in tragedy. I worked at a chemical plant that had an explosion where the safety systems (SIS) were bypassed. The engineers were trying to light off a boiler and were having trouble. The safety system had a 30 minute time delay between unsuccessful light off attempts to allow the fuel to disperse. The engineers bypassed the time delay and attempted multiple unsuccessful light offs in a row which flooded the boiler with fuel. Finally, it lit and went boom. Those near the equipment in the field were all killed.
Hope you don't have any pets, children, or slow moving adults around.