Elk M1G output relays

[dBeau]

I have two garage door openers. The newer one is the one I expected problems with. Turns out replacing the lighted button with an Elk relay is no problem. A project for another day is use a couple more relays along with the capacitors from the original button to control the lock and light features.

But this is a story about the old opener. It uses a standard non-lighted door bell button. What could be simplier. After testing by touching the wires together all checked out fine. So, I connected it to a relay and tested it. No luck! ...but the relay was clicking and the led was blinking. Turns out, that a "long press" will activate the light without opening the door, while a short press will activate opener. The nice bit is that I finally understand why my wife has had such trouble getting the old button to open the door. The bad is that I cant figure out how to get the elk to close the relay for less than one second.... and one second is considered a "long press".

I ended up hacking my remote and used the elk relay to "push" the button. This works but I'd really rather keep the remote in the car! Any suggestions?

 

[fcwilt]

This is probably overkill for most folks but its what I did:

As you see the typical garage door opener is expecting a contact closure to do various things. Moreover pushing the button doesn't tell the door to open or close, just to change state (open if closed, close if open).

For my home automation I wanted a "smart" garage door. I wanted the home controller to be able to command the door to open or to close, not just change state. So I used an ELK Magic Module and a pair of magnetic switches. By using two switches I had a positive indication of full open or full closed. Each ELK MM has four inputs and four outputs. Two inputs were used for the magnetic switches and two for command input from the home controller (open/close and activate). One output from the MM went to the input on the garage door opener and the other three outputs went to the home controller for status information (open, closed, busy).

The programming capabilities of the MM allowed me to "convert" from the input/output requirements of the home controller to the simple contact closure required by the garage door opener. The timing capabilities of the MM allowed for such things as sending long and short contact closures to the garage door opener and determining if the door has completed its motion in the expected amount of time. For instance most late model garage door openers will reverse direction if something obstructs the motion. The two magnetic switches allow the MM to "see" that the door started to move and then returned to the starting position. The MM then can signal that home controller that something seems to be wrong.

Overkill perhaps but it works just fine and allowed me to have more control over the garage door then I would have using the home controller alone (due to the limited programming capabilities of the home controller).

 

[dBeau]

For my home automation I wanted a "smart" garage door. I wanted the home controller to be able to command the door to open or to close, not just change state. So I used an ELK Magic Module and a pair of magnetic switches. By using two switches I had a positive indication of full open or full closed. Each ELK MM has four inputs and four outputs. Two inputs were used for the magnetic switches and two for command input from the home controller (open/close and activate). One output from the MM went to the input on the garage door opener and the other three outputs went to the home controller for status information (open, closed, busy).

I took a similar approach when programming my elk. I have both a full open and full close indication. The rules for open and close are similar but reversed. They are both derived from a simple statement: If the door is not where it is desired, "press the button".

Most of the time, the door starts from either the full open or full closed position. So this simple rule works perfectly. If a request comes in while the door is either already in motion or is stopped between either full position, its quite difficult to predict what will happen next if the button is pressed. However, since it's not where desired, the button will be pressed and the motion will either be stopped or reversed (depending on which of my doors we are talking about and its current direction of travel). So, this rule also works well when a user presses the button then senses impending doom. At that point it doesnt matter which state they request, the door will stop or reverse.

I should point out too that instead of a doorbell style switch as input, I use a rocker switch. Move the switch to the bottom and the door goes down, move the switch to the top and the door goes up. The only "problem" is that the current state of the switch doesnt always indicate the position of the door. That is, the switch might be down but the door is up because it was opened from the car. In that case it is necessary to move the switch up (causes nothing to happen) then down to get the door to move.

I've toyed with the idea of adding a timer to verify that the door did make it to the desired position but at this point I dont see much in it but trouble. Let's say that after 15 seconds the door is not closed, does the system push the button again? Not if while the original command was in progress, it gets overriden by my wife trying to put the car in the garage! Of course opening the door is less of a problem but also less interesting. I cant think of anytime I'd like to tell the door to open when I am not standing there to make sure it happened.

The external commands from the car remote add potential for trouble. Replacing them with a receiver that talks directly to the elk's inputs would help to eliminate one reason why the door did not do what it was told. There are more of course. What's really needed is an automation friendly garage door opener... separate up and down inputs and fault indicators would be just a start... but I digress.

Sadly, I am still left wondering if there is anyway to get the elk to trigger a relay for less than a second.

 

[amirm]

Sadly, I am still left wondering if there is anyway to get the elk to trigger a relay for less than a second.

How handy are you with simple electronics? Because if you are, you can build a "one-shot timer" using parts from your local Radio Shack. Here is a quick link to what that looks like: http://home.cogeco.ca/~rpaisley4/LM555.html#2. By using the appropriate resistor/capacitor value, you can have any pulse length you want.

Now, I am lazy . So I would first try a simple RC circuit (resistor + capacitor). Depending on debounce logic on the input of the openner, this may work and take all of 5 minutes to build. The wiki page unfortunatley is full of math but does have the simple schematic half way down the page: http://en.wikipedia.org/wiki/RC_circuit. Again, the value of R and C determine the time constant or how long it takes for the pulse to rise and fall. By making this time shorter, then the effective pulse width is also made shorter. But since the waveform in this scenario is no longer nice square wave as the above timer one is, it may or may not work. But it is so easy to try that I would test it first before building anything complicated.

If none of this makes sense, then the magic module may be the right solution. I have not used it but reading the spec on it, it seems quite capable of any task of this sort even if you don't want to design the fancy control system suggested.