hooking up my SECU16 to my Ocelot

[Guy Lavoie]

In principle, I'd stick to supervised inputs for an alarm application, but if this is not that critical (like for an outbuilding without valuables or for a unmonitored alarm system) then it might be ok. Just be aware of the potential for missed detection of very short triggers in analog mode. This could be more pertinent for a door that could theoretically be opened and closed very fast!

 

[smee]

Just be aware of the potential for missed detection of very short triggers in analog mode. This could be more pertinent for a door that could theoretically be opened and closed very fast!

Won't this be an issue as well if you were to use this method to detect intermediate garage door positions? How long is very short?

By the way, there is something else to consider with most magnetic switches when they are used to detect position. This affects primarily the long rectangular switches (as opposed to the smaller round ones). As the magnet moves past the switch along the long axes of both, the switch may open and close a couple times. This is not a problem if the two halves move away from each other perpendicularly to the long axis. If you have a switch mounted to the garage door frame and the magnet on the door, they will most likely be moving past each other along that long axis.

This problem may be exacerbated by metal doors/windows and frames.

 

[Guy Lavoie]

My understanding of the garage door position sensing application is that you want to know if the door is presently fully opened or closed. This would be for an application where you want to have them close automatically at a certain evening hour, or want to trigger some kind of alarm or message. In a case like that, the response time doesn't really matter, since you are looking for a static position. It would matter a lot more if you were counting on detecting the door's position to actually stop the motor! The application here seems to just need to detect if the door if fully opened, fully closed, or not one of these two...

 

[smee]

Guy:

I'm looking at BSRs post about using multiple magnetic switches to determine intermediate position (a few pages back).

 

[Guy Lavoie]

I see. Well in that case, I would not use such a circuit to try and detect the door going by a certain physical point in time. Unless the door is moving very slowly (ie: a switch would change state and stick to it for at least 3 seconds) then it can not be reliable. Using individual, latched inputs would work however.

 

[deranged]

At one point, I did play with a 10 turn pot connected to the garage door drive.
It worked okay, I didn't have problems. While i was trying to write down the analog value and work out so close to 5 volts would be open and close to zero closed. And get the right gearing to not rip the pot apart but give me the best analog values. It occured to me I didn't really want to know where the door was. Just if it was open, closed or lost somewhere in the middle. So 2 magnetic switches worked fine.

StevenE

 

[electron]

damn, I just bought 500 different kinds of 1/4W resistors as RS didn't carry the other ones, I guess I will stick with using j ust 2 magnetic switches for now, and use them later when I am running out of inputs.