Timing a case structure

[karthiksudhir80]

I have a push button with mechanical action set to switch until release(so that it gives an impulse boolean ON for 1 sec). I want to know a way so that the case structure remains active for a specified time after the push button goes off.
Thanks in advance.

[Michel_Gauvin]

Try this.

[mikeporter]

Can you explain a bit more what you are trying to accomplish. It sounds like you might be wandering off into the weeds just a bit in terms of your design.

Mike...

[karthiksudhir80]

Okay, I am trying to implement a bidirectional visitor counter in a room using two PIR sensors wired to a DAQ, the two sensors are kept 1m apart along a single line, and depending on what IR sensor is cut first the counter should either increment or decrement, for entry and exit respectively. However whatever logic i so far used results in negation of the two cases so the counter yields zero. So i am forced to use two doors for entry or a partition of the door to maintain a count. So right now, im thinking of some way to disable one IR sensor for some time while the other is being cut.

[jcarmody]

I'd use a shift register to store the state of the inner & outer sensors and an Event structure timeout case to address the 1 second delay.  Like this:

 The snipppet didn't show what I wanted very well, but the VI is attached.

[altenbach]

If you are interfacing to real sensors, you probably need to poll them at regular intervals and an event structure is not the solution.  A simple state machine would suffice.

You have several states, e.g.:

1. no sensor triggered
2. sensor 1 triggered
3. sensor 2 triggered

if you are in state 1, simply wait for any sensor to trigger and proceed to the relevant state

if you are in state 2, and sensor 1 triggers again, don't increment, because the person has left. if sensor 2 triggers, increment the count and go back to state 1.

if you are in state 3, and sensor 2 triggers again, ignore. If sensor 1 triggers, decrement the count and go back to state 1.

... or similar.

[mikeporter]

A couple thoughts, first that your logic is in a way similar to a quadrature encoder where the leading input telss the direction of rotation.

Second, depending on what you want the software might be working fine. Logically unless someone decides to take up residence in the area, at the end of the day (metaphorically or literally) the people going out will always equal the number of people coming in -- hence the 0. Recording either the inbound or outbound count will give you a simple headcount. The zero could be used as a security check.

Tracking the difference between the in and out counts over a limited time scale could even give you insight into how long folks are staying in the area.

Mike...

[karthiksudhir80]

This is perfect! You have implemented the logic so flawlessly! I have so much more to learn... Thank you so much!

[jcarmody]

altenbach's comment is important to note.  My answer relied on the boolean controls while your situation relies on digital inputs that won't trigger an Event structure by themselves.

[karthiksudhir80]

I can wire booleans to the inputs from daq and use local variables now, can't I?