Interface to a gate controller

[berkinet]

I have a new gate controller(GiBiDi BA230) and would like to interface it to my home automation system. The challenge of the moment is to detect the gate state. There is a controller output that provides 24vdc according to the following pattern

• Gate State - Voltage
  Closed - 0vdc
  Opening - 24vdc pulsed at 1Hz
  Open - 24vdc (steady)
  Closing - 24vdc pulsed at 2Hz

The gate takes around 20 seconds to open or close and I would like to detect the state as quickly as possible each time it changes. I can hook a relay up to the output so I can have dry contacts, or any control voltage I want, like 5vdc from an ifKit.

I have tried using a 1054 but so far without any success. At first it kept measuring 50Hz (I am in France) so I figured that was AC line leakage and I ran the output signal through a diode bridge. That got rid of the 50Hz signal, but did not produce anything useful. My next try will be to use the relay between the controller and the 1054 and switch 5v. Even then, I am concerned about how quickly the 1054 will actually detect changes, and if it will differentiate between steady on and off states.

But, I was wondering if anyone has any other ideas on how to approach this project.

Ideas?

[berkinet]

If anyone cares... I solved the problem in an interesting manner. The 1054 was the way to go, but apparently the relay that was switching the 1054's 5v output back to its input was not clean. I saw a continually changing range of frequencies. So, instead, I used the relay to switch ground to a Digital input on an interfaceKit. Then, in software, I used the high on the DI to trigger a 200ms high on a digital output, and fed that to the 1054. Rock solid 62.5Hz slow flashing, 1.25Hz fast flashing. As a bonus, I can now easily detect a stay on by looking at the DI. A little logic to process all of this and I have my 4 states.

[rootboy]

It sounds to me like you had the sensing wires for the 1054 in the same conduit as the rest of the wires.

The input impedance for the 1054 must be pretty high, which means that it won't require much of anything to trigger a pulse.

I've seen this in industrial equipment, notably a gasoline fill station at the old Saturn plant that if the operator fell behind a ringer would go off. Since the electricians put the leads to the pulser in the same conduit as the ringer, when the ringer would go off the display showing how many gallons of gas that had been put into the tank would look like someone had hit the jackpot.

Based on the one diagram that I saw, I would tie the 5 volts to the "+"terminal on the 1054 board, and place a 1 watt, 2.7k resistor between "-" and ground. Now run the pair of wires out to a dry set of contacts on your gate controller. Dry meaning no power provided by the gate controller. Just a simple N.O. or N.C. contacts (N.O. preferred).

Do your own math to figure out the resistor size and power rating, it's not my best subject.

With the contact open, the resistor will keep your "-" at ground while your "+" will be held to 5 volts giving the maximum differential. Once the contact closes, the "-" will be pulled up to 5 volts, and will be the same voltage as your "+" input.

Now if there ever was an application screaming for isolation, it's this one. One good lightning strike and you will not only fry your controller, but your home automation equipment as well. This is a common problem with copper phone wires where they work fine in dry weather, but become scratchy when it's wet. The reason for this is because the insulation is riddled with holes from lightning strikes.

I would wire up the inputs to opto-isolators and place that board in a weatherproof box and mount it outside the house. Be sure to provide a ground to give a strike a better place to go. That would require some new hardware at the gate end of things to drive the input side of the opto-isolator.

At the very least, use a USB isolator.