Turning an ELK relay on for a split second?

[signal15]

I have a garage door opener button that has a display on it and gets it's power from the opener itself. When I use a relay to bridge the wires together on the back of it to open and close the garage door, it works fine, but the display loses power and it resets the clock and other crap on it.

The relay needs to trip for only a split second to avoid having the display shut off and reset. But the automation rules only allow me to set it down to 1 second minimum. How to make it trip for 0.1 or 0.05 seconds? Any good ideas?

 

[Lagerhead]

Look around for one-shot relays? I have seen them though I have never used them and have no guidance on source or price.

For a simple homebrew, have you considered charging a capacitor during the "inactive" period, and then upon becoming "active" dumping the charge into a second relay which does the actual actuation.

 

[signal15]

I have a garage door opener button that has a display on it and gets it's power from the opener itself. When I use a relay to bridge the wires together on the back of it to open and close the garage door, it works fine, but the display loses power and it resets the clock and other crap on it.

The relay needs to trip for only a split second to avoid having the display shut off and reset. But the automation rules only allow me to set it down to 1 second minimum. How to make it trip for 0.1 or 0.05 seconds? Any good ideas?

I am looking into a capacitor and second relay now. I also have some 555 timers sitting here and some prototype boards, was thinking I could make something with that. I don't think I have a spare relay laying around though. I think I should be able to do it with just the 555 timer though.

 

[BraveSirRobbin]

You might want to consider hacking a remote and interfacing that with your Elk relay methodology as shown HERE.

This way you will not interfere with the normal power reset AND not have to mess with interfacing a one shot (which I think you will have trouble with in the end) to the system.

 

[Old11C4]

How does the wall switch work--try some tests and make observation

1. If you push and hold the switch for 5 seconds does it clear the display.

2. Take a digital multimeter (volt-ohm) and measure the AC and DC voltage across the screw terminals. (If you get only one, AC or DC, then that is all you would need to check from here on.)

3. While measuring the voltage, have someone press the button and observe any voltage change.

To run the backlight and other smarts in the switch, it must have some resistance in the N.O. condition and a lower resistance in the closed position. The GDO must look for a voltage/resistance drop somewhere between the open and semi-closed position while at the same time allowing enough voltage to keep the display powered up.

Have you tested your split second theory yet? Can you short the contacts for 1/2 or 1/4 seconds and trigger the door without clearing the memory. Even if you can, I am concerned that you may still get intermittent memory loses or non-responsive door open signals. No reason to install an unstable system.

4. Disconnect the switch and measure the open resistance and closed resistance across the terminals. Does it go to zero ohms when closed or some over value when closed?

If it goes to say 10mOhms, try putting that same resistance (10 mOhms in this example) in series to the Elk output.

Can you get any access to the relay on the wall switch so you just short across the same switch that it shorts out?

I suspect connecting the output to the GDO where the wall switch connects would cause the same loss of power and loss of display settings. However, check and see if the GDO has a aux connection for a second switch. It never hurts to download a manual and see what they offer.

I have a Chamberlin (Sears) GDO with the fancy wall switch also -- so I may have your same problems when I hook up mine.


Old11C4 - from somewhere in Iraq
OUT

 

[signal15]

I have measured all of that. I originally thought that maybe the button just bridged the two wires with some resistance so it wouldn't drop the voltage completely. But it does not do that. I hooked up an old analog multimeter, and found that it just shorts the two wires for a split second. I can simulate this with some test leads.

I do like your idea of just soldering some wires on the switch portion of it. I think a lot of the stuff is SMT though, I will have to check if this is even feasible.

 

[signal15]

I took your suggestion of just soldering a couple of wires to the switch, works great. Thanks. Don't know why I didn't think of that, considering I've done it on other things.

 

[Old11C4]

I took your suggestion of just soldering a couple of wires to the switch, works great. Thanks. Don't know why I didn't think of that, considering I've done it on other things.

Great, glad it helped.

BTW: I just popped back in to add another thought. Too late for you but may help others.

One of the contacts on the switch may be a common ground to the terminal screws. If that is the case, we would only have to solder one wire to the N.O. contact on the switch and connect the other wire to the screw terminal (switches ground). Sometimes it worth a lot to avoid one more solder joint on a PCB.