Jerky Tracking?

[Matt Krass]

Well we've set up the camera, and using LabVIEW it tracks beautifully, it's wonderful, but thats driving its own servo. Kevin's code does a great job of searching and finding the light...then goes haywire.

Once locked the camera starts violently pitching itself back and forth and up and down until one of the PWM cables pops.

We're using the code verbatim, so....I'm not sure...the mounts assembled right.

What else should we check?

Thanks.

[Kevin Watson]

Originally Posted by Matt Krass

Well we've set up the camera, and using LabVIEW it tracks beautifully, it's wonderful, but thats driving its own servo. Kevin's code does a great job of searching and finding the light...then goes haywire.

Once locked the camera starts violently pitching itself back and forth and up and down until one of the PWM cables pops.

We're using the code verbatim, so....I'm not sure...the mounts assembled right.

What else should we check?

Thanks.

How do you have the camera mounted? Did you look in the header files for a documented parameter that might explain this wacky behavior?

-Kevin

[Matt Krass]

We're using the kit mount, we tried inverting the camera and that seemed to improve it somewhat but it still jitters way too much to be useful, I've fiddled with the settings but I can't find any noticeable effect. I've managed to slow down the scan speed while searching, is there a parameter I might have missed that controls how fast the servos try to error correct/center when it see's something? It looks like a constant overshoot kind of problem..

EDIT: Looking again after a good break I found some gain parameters to adjust, I'll see if those help.

[Kevin Watson]

Originally Posted by Matt Krass

We're using the kit mount, we tried inverting the camera and that seemed to improve it somewhat but it still jitters way too much to be useful, I've fiddled with the settings but I can't find any noticeable effect. I've managed to slow down the scan speed while searching, is there a parameter I might have missed that controls how fast the servos try to error correct/center when it see's something? It looks like a constant overshoot kind of problem..

EDIT: Looking again after a good break I found some gain parameters to adjust, I'll see if those help.

You can also tweak the gain values via the tracking parameters menu. Just remember to save it to EEPROM once you're happy with the value(s).

-Kevin

[Goldeye]

Something I didn't really notice in the docs:
Does the streamlined version check the EEPROM values saved by the bells and whistles version?

[Kevin Watson]

Originally Posted by Goldeye

Something I didn't really notice in the docs:
Does the streamlined version check the EEPROM values saved by the bells and whistles version?

No, it only uses values from the header files.

-Kevin