[Lab] encoders for spiders, one or two sensors?

[Darcy Whyte]

I want to make a "winch servo" to move some spiders up and down on a line.
http://inventorartist.com/halloween/

I'm looking at making a simple encoder out of the sensors we used in the
Arduino Challenge. http://inventorartist.com/oac/

This will be a sort of servo based on a dc motor and I'll use the sensor to
track the progression of the rotation.

In theory I know the direction of the motor since I'm driving it. So each
time the sensor is triggered I know I have progression in the direction of
the drive. I'm tempted to keep it this way and stay simple.

But lets say a 3 year old yanks on one of the spiders. That may give
 rotation against the drive direction and the system will have wrong
information.

To fix this I can add a second sensor.

So an AB trigger of the two sensors is one direction and BA is the other
direction.

Any thoughts on if this is worth it and if it solves the problem?

My intuition tells me the project will be good enough with the single
sensor. Especially since I'll probably use a hacked RC servo as the drive.
http://inventorartist.com/continuous-rotation-servo-hack/

Yanking one of those things backwards against the drive is not easy...

There is another idea I had and it's to use the IR as a limit switch and do
the rest of the work with dead reckoning guessing. If each drive has two
spiders (when one goes up the other goes down), then I can home it on both
sides to guess how long to drive the motor for full movement. And if it
ever drifts it'll hit the home trigger...



Any other ideas on how to move the spiders around?




--
Darcy Whyte

Art+ inventorArtist.com <http://inventorartist.com/> | Aviation
rubber-power.com
Contact: darcy at inventorArtist.com | 613-563-3634 by appointment (no text)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://artengine.ca/pipermail/lab/attachments/20141011/792e5cb5/attachment.html>