Strange Camera Problems

[Hutch]

At the last regional we were at , we finally started really using our camera... It was working great and tracking and everything until we noticed a problem: whenever we turned on our ball collection mechanism, the camera would freak out and go all the way to the left.

Since our collector has two fisher price motors, we tried to just run one at a time: with only one running (regardless of which), it would work just fine (maybe be a wee bit jerky but not freaking out). The motors are located on the complete opposite side of the robot from our camera (camera is top left, motors are bottom right).

We thought it might be the Van De Graff generator thing, but we tried grounding things and it didn't change; it also happens immediately so we ruled that out.

We are really at a loss as to what the problem could be, so anyone have any ideas?

[Andrew Blair]

Here's the one thing I can think of:

When you turn on the Fisher Prices, unless you use some ramp code, the motor brushes arc. Now, somewhere, I heard some tech nonsense about collapsing magnetic fields when arcing occurs(or rather, the arcing is caused by the collapse of the field), and back EMF, and lots of junk like that. Perhaps this could be causing it. I'm sure someone brighter than I can elaborate.

Because it only occurs when you run the collection mechanism, which only contains FP's, I wouldn't think it's a current drop issue, though that is possible. The only thing I don't get is why it occurs only when *both* motors are run. Maybe you've got your FP power lines near your camera wires, and are inducting a power or signal current. Interesting...

[Alex G]

It can be because of noise in the PWM outputs. If the camera draws power from the PWM outputs, I would recommend using an LC(coil-capacitor) filter before the camera. He had similar problems when the backup battery was low. Noise can temporarily lower the voltage, so i suspect this to be the cause.

Here's some info on the LC filters:

http://www.tpub.com/content/neets/14179/css/14179_204.htm

[Keith Watson]

First dump your pwm values to the terminal window to ensure that they are what you expect.

Long runs of multiple wire cable (like pwm wire, also known as servo wire in the world of radio control models) are known to act as excellent antennas. DC brushed motors are an excellent source of noise for that style of cable. Ways to reduce the noise that is picked up include twisting the cables, adding chokes (magnetic rings) at each end of the cable, adding a filter (capacitors) where the power wires connect to the brushed motors, etc.

[coldabert]

To isolate the problem, use an oscilloscope to monitor voltages your pwms.

This might seem like too much for a simple problem, but it will definitely show you which systems are not causing the camera to move.

(Plus its cool and you can tell people that the wavy lines are talking to you in computerspeak)

Let me know if you need help setting up the scope.

[Keith Watson]

To isolate the problem, use an oscilloscope to monitor voltages your pwms.To be clear, the signal line is monitored and you watch the width of the pulse. Pretty cool.