Another garage door question

[Ira]

My shop has two regular sectional garage doors. Neither has a GDO on it, mainly because there aren't any cars parked in there. Many times, I will manually close the garage door without locking it when I leave it for a short period of time (usually to keep critters out of it). For example, if I am mowing, I will back the mower out, close the door but not lock it, then go about mowing the yard. Or, I may close the door but leave it unlocked thinking I will be back in a few minutes, but get sidetracked and not go back for a couple of days.

The door has the typical lock mechanism -- two horizontal bars that when locked go thru holes in the tracks. When locked, the bars extend thru the hole an inch or so.

I would like to have my Elk notify me when the garage door is unlocked. The actual notification method (voice message, indicator light, etc.) is not important right now. What I need to find is some way to detect when the door is not locked. I think the best solution would be some kind of mechanical dry contact device mounted over the hole in the track. When the lock bar extends thru the hole in locked position, it would mechanically open the dry contacts. I may couple such a device with a regular garage door open/close sensor so I can differentiate between unlocked/open vs uplocked/closed.

Has anyone seen such a device, or something similar that could be modified to work? For the most part, I would rather buy (and modify if necessary) than build.

Thanks,
Ira

 

[ano]

My shop has two regular sectional garage doors. Neither has a GDO on it, mainly because there aren't any cars parked in there. Many times, I will manually close the garage door without locking it when I leave it for a short period of time (usually to keep critters out of it). For example, if I am mowing, I will back the mower out, close the door but not lock it, then go about mowing the yard. Or, I may close the door but leave it unlocked thinking I will be back in a few minutes, but get sidetracked and not go back for a couple of days.

The door has the typical lock mechanism -- two horizontal bars that when locked go thru holes in the tracks. When locked, the bars extend thru the hole an inch or so.

I would like to have my Elk notify me when the garage door is unlocked. The actual notification method (voice message, indicator light, etc.) is not important right now. What I need to find is some way to detect when the door is not locked. I think the best solution would be some kind of mechanical dry contact device mounted over the hole in the track. When the lock bar extends thru the hole in locked position, it would mechanically open the dry contacts. I may couple such a device with a regular garage door open/close sensor so I can differentiate between unlocked/open vs uplocked/closed.

Has anyone seen such a device, or something similar that could be modified to work? For the most part, I would rather buy (and modify if necessary) than build.

Thanks,
Ira

Why can't you mount a normal door sensor magnet anywhere along the locking bar, and mount a door sensor above or below it? When you lock or unlock it, doesn't these bars move horizontal an inch or more? You may to experiment, but I'm sure you can find a position to mount the sensor so it detects if its locked or not.

 

[sic0048]

Why can't you mount a normal door sensor magnet anywhere along the locking bar, and mount a door sensor above or below it? When you lock or unlock it, doesn't these bars move horizontal an inch or more? You may to experiment, but I'm sure you can find a position to mount the sensor so it detects if its locked or not.

That was going to be my suggestion as well. I'm sure with a little trial and error, you can find a way to mount the magnet on the door and sensor on the door frame in such a way that when the door is locked the magnet closes the loop on the sensor. When he door is unlocked, the magnet does not close the loop and the zone would read open.

This method would not have any moving parts and therefore should be more reliable than a dry contact method. Plus, since you can physically attach the sensor to the door's frame, there should be little chance of the sensor moving or being knocked off. The frame itself will help protect the sensor from accidents.

 

[Ira]

I thought about that, but came up with a few obstacles. Mounting both pieces on the door (i.e., one on the bar and one on the door itself) would be tough to wire back to the panel given the movement of the door when it opens. I don't plan on having wireless out there. To make wiring easier, I could mount the magnet on the bar and the sensor on the track, but then I'm not sure the typical surface mount sensor would have the correct "orientation" (magnet would be perpendicular to the sensor), or that the magnet would be strong enough to work thru the track (since the sensor would be on the outside of the track and the magnet would be on the inside). Maybe a larger/stronger magnet would work.

 

[sic0048]

Well, I would place the magnet at the end of the locking bar. So that when the door is unlocked and the bar is not extended, the magnet is inside the track. When the door is locked and the bar is extended, the bar with the magnet is sticking through the frame (thereby locking the door) and puts the magnet right next to the sensor.

Actually, I think a "normal" rare earth magnet would work well for this. It is small (perhaps 1/2" in diameter or less) and tends to work with a slightly larger gap - I think you'll end up with about a 1/2" gap.

So the whole thing would look like this:

Edit - added picture

 

[Ira]

I took a few measurements. The horizontal bar has about two inches of "throw". When locked, it has about 1/2" on the outside of the track. The magnet would have to be fairly thin (no more than 1/4" thick) so as not to get hit the track during motion. The vertical surface of the bar is not flat (it's a little concave), so two sided tape probably wouldn't work. A hole in the bar takes up about half of the mounting surface area outside the track when locked. Given the violent "snap-back" that happens when the key is turned to unlock the door, it may be difficult to find a way to mount the magnet that would last. May be a job for duct tape.

I ordered a few 1/2" wide bar magnets in various lengths and thicknesses to experiment with. Should know something by this time next week.

Thanks,
Ira

 

[ano]

How about this. Mount the magnet on the bar a few inches inside the rails, then mount the sensor on a piece of aluminum bar that is mounted on the wall outside of the track, but it bends around to be suspended an inch from the magnet. The door can open fine, since the sensor is mounted on the wall not the door. You may need to experiment, but it doesn't sound too difficult.

 

[Ira]

I'm hoping one of the bar magnets I ordered will be strong enough so that I can mount the sensor on the outside of the track, and the magnet will affect it when the lock bar goes thru the hole in the track, similar to what sic0048 drew. I also need to experiment with some sensors and different orientations relative to the magnet. The only spare sensor I have is pretty big, so it would pretty much need to be mounted vertically. I'm sure there are smaller ones that I can mount horizontally if that works better.

ano...the way the door's roller mounting brackets are, such a sensor mounting bracket as you describe would have to extend several inches beyond the inside of the track (and inside of the garage door opening) to clear the roller brackets. That means I would hit it with my tractor or something within 24 hours of putting it in.

 

[cornutt]

If that doesn't work, some sort of leaf switch with a roller-bar end might do the job. You could just epoxy it to the track.

 

[BraveSirRobbin]

I'm a little worried that you may not have enough throw in the locking bar to get reliable open/closed signals from the detector with a magnet solution.

Consider instead a small hinge mounted on top of the opening with the moving plate covering the slot when the bar is in the open position, and then at a 90 degree angle when in the locked position. Place a tilt sensor on the hinge to easily detect its position as shown below.

Or perhaps have a magnet in place of the tilt sensor, and have the sensor itself vertical, at the bottom of the hinge when in the closed position.

Of course you will need none of these methods if your magnet mounted on the locking bar directly works out for you.

 

[Steve]

With only 1/2" coming thru the track, will that be enough to tilt up a hinged sensor? Would be close.

 

[BraveSirRobbin]

With only 1/2" coming thru the track, will that be enough to tilt up a hinged sensor? Would be close.

Good point. Of course a small hinge is cheap enough to just try it.

Darn, now I'm curious (BSR gets shoes on to search garage parts bins for a small hinge).

 

[JayH]

I'm a little worried that you may not have enough throw in the locking bar to get reliable open/closed signals from the detector with a magnet solution.

Consider instead a small hinge mounted on top of the opening with the moving plate covering the slot when the bar is in the open position, and then at a 90 degree angle when in the locked position. Place a tilt sensor on the hinge to easily detect its position as shown below.

BSR - I like this one. I'll have to put it in the bag-of-tricks and hope I remember it. Depending on the tilt switch, dead-level on the hinge might be touchy, you may need a slight offset in the "open" position to make sure a bump or wind gust doesn't trigger it.

I had 2 ideas to add:

1) If you have wireless available, put the transmitter and magnet/switch pair _on the door_. Set it up so when it is "locked" the magnet is pushed up to the contact, when unlocked it is away. Aside from the obvious open/close pattern matching the spring loaded "pop" on the unlock may ram the magnet into the contact if it is in the direction of movement.

2) "U" shaped photoelectric proximity sensor, not the cheapest solution, but no moving parts! Simply mount it on the frame so the bar interrupts the beam. You may need to cover the lock hole.
Example: http://www.fargocontrols.com/pdf/sensors/Fargo_Value_PU.pdf (NPN model for Elk, you can even power off Elk +12V)
Or hit up your local electronics surplus store.

Jay

 

[Ira]

BSR's suggestion is close to what I was talking about in my OP, except that rather than using a hinge with a tilt sensor, I was thinking a spring-loaded hinge with a wire attached to it mounted like BSR's picture shows. Then on the other side of the hole, a metal contact plate with another wire attached to it. Both the hinge device and the contact plate would be insulated from the track. When the door is locked, the bar would push the hinge away from the contact plate, opening the "zone". When the door is unlocked, the spring loaded hinge would snap back against the contact surface, closing the zone. The hinge would not be negatively affected by the fast motion of the bar due to unlocking the door since it would return to the closed position at its own rate. I guess this could actually be done with just a strip of spring steel as long as it would not develop a "memory" when open (door locked) 99% of the time causing it not to spring back.

The bar has a two inch throw. I'm thinking I can vary the position of the magnet on the bar (mount it on the bar so it is barely close enough to close the sensor when the door is locked) so the two inch throw will be enough.

On a somewhat lighter note...I was doing some measuring/experimenting on this earlier today using the garage door on my attached garage, which does have a garage door opener on it. When I finished playing around, I unfortunately left the door in the locked position. When my wife came home this evening and tried to open the door, the GDO had enough force to bow/bend the lock bar enough to pull it back thru the lock hole. My wife said the door was making a terrible noise while it was opening, so she stopped it half way up. After kinda fixing the bar, etc., I checked to see how much force it took to stop the door from opening. I'm guessing it took over 100 lbs. Time to adjust the GDO. Fortunately, the force necessary to stop it from closing was okay.

Ira

 

[Ira]

2) "U" shaped photoelectric proximity sensor, not the cheapest solution, but no moving parts! Simply mount it on the frame so the bar interrupts the beam. You may need to cover the lock hole.
Example: http://www.fargocontrols.com/pdf/sensors/Fargo_Value_PU.pdf (NPN model for Elk, you can even power off Elk +12V)
Or hit up your local electronics surplus store.

Jay

The photoelectric prox sensor looks interesting. I wonder what the light immunity stats mean in the real world. Too much light from another source would trigger a false "unlocked" condition. That probably doesn't matter since the only way that would happen when the door is really locked is if I left the lights on in my shop. I would want to know about that anyway. Am I correct in thinking that too much ambient light cannot cause a false "locked" condition, i.e., the ambient light can't somehow result in the sensor not seeing its own beam?

Certainly seems like a good off-the-shelf solution, and is as cheap as wireless (cheaper since I don't have a wireless receiver in my shop and it may be too far away to ge picked up by the one in my home.

Ira