Nuts and Volts - July 2014

like something you would see on the end of a boom mike (called a dead cat in the industry!) when they are being used outside to filter out wind noise. I was worried that the long haired cat might have a genetic way to bypass detection. Turns out the sensors work great in detecting both cats!

I set it up on the table without anything on the servo and waited to see what the reaction would be. With a few treats left on the table, the long haired cat (not the kitten) made a fast retreat when the servo started moving around.

When set up on the table, I was able to walk around the entire perimeter without being detected. There is a variable in the code called tripDistance which might need adjusting if your table needs a detection range greater than or less than the 40 cm that is default in the code.

Future Upgrade Possibilities

Some future changes that would enhance the functionality would be to mount a piezo buzzer or some other type of loud sounder directly in the box. One could be plugged into the alarm output jack, but it would be nice to have this functionality built in. Currently, there is only a single variable in code that sets the detection range of the sensors. This works great if it will just be used on a single table size that’s round like my situation. A simple code change would allow for sensor 0, 1, and 2 to have different detection ranges. This would allow it to work better on oval tables, for example. This would introduce another obvious enhancement requirement, though. You would need to remember exactly where to place it so that all three sensors line up to the programmed sensor distances. None of these problems exist when using a nice round table.

The way to make that work the best would be to not only have three separate detection distance variables for the three sensors, but to also allow them to be programmed in place. This could be accomplished by adding a new switch to the system. When the switch is pressed, the system would make note of the closest distance it measures from all three sensors. You would simply walk around the entire table so that the system could record the safe distance. When put back into run mode, subtracting 5 cm from these new values would give you a great baseline for the system to have optimal detection.