Wind And Wind AutoKAP

 
Presented here is a wind-powered and wind-up-powered rig for kite photography.  Lot's of people have done RC-controlled rigs but since I do circuits at work all day, I thought it would be fun to make something fully mechanical.  The Kite Aerial Photography people call it "autoKAP".  The camera is a Canon S95 running a Canon Hacker Development Kit intervalometer that commands the camera to take a picture every 5 seconds indefinitely. 
 
An RC airplane propeller powers the "pan" motion, and a clockspring powers the "tilt" motion.  The device pans around over the course of a couple of minutes, then tilts to a new angle, then pans around again at that angle, then tilts again, pans again and then tilts back to the first position.   The tilt angle selections are adjustable with a wrench.  I've made 3 settings but it could be expanded to many settings by using a wheel instead of the 3 individual pieces.
 
 
 
 Here's a video of the mechanism in action.  There's a zoom in at the end to show the escapement a little better.
 

YouTube Video

 
 
 Here's a short impromtu video tour. 

YouTube Video

 
 
This picture shows a sequence, with the camera pointing down at the dark grey road (which rotates generating a bunch of unneeded pictures) then tilts up some for a bunch of pictures, then tilts out for about 26 pictures, and then back down.  The rig spent about 3.5 minutes at each tilt setting in an unusually unsteady wind. 
 
 
 
 Here are a few more pictures to give you an idea.
  
 
This is the propeller shaft "bearing", a piece of white teflon.  Also shown is the worm (some of a bolt with a hole drilled through it, JB welded lightly), and the worm gear (actually it's a plain old gear; a junk store find).  The gear ratio in this set is 60:1.  The gears don't mesh perfectly, but what do you want for $0.95.   The shaft has some play back and forth to keep friction very low. But the pan motion is reduced by 3600:1 so it doesn't show up.  The shaft rattles a bit because it's bent (a source for improvement).    [updated 2/7/2012:  When I lost the teflon piece during a session, I couldn't get it working reliably again due to the poor centering of the worm's holes.  So I swapped out the home-made worm gears for some real ones bought at stock drive products, and I replaced the shaft.  Need to try it still.]
 
 
 
 
This is a view of the whole thing.  There's a second 60:1 ratio and then another 1:1 to transfer the motion to the vertical shaft that holds the camera bracket.  The input and output shafts both needed to be centered (well, almost) for balance and wind pressure. 
 
 
 
 
 This shows the escapement.  The dark long piece is a spring (a street cleaner bristle).
 
 
 
 
Another view of the escapement.   The angle of the paddles with respect to the camera bracket determines the tilt.
 
 
 
The spring was reduced something like 5:1 to increase the number of turns and decrease the force.  It could still use a governer or some way to slow the camera flip down.  If it's wound fully, it's fairly violent when it does the 270 deg flip
 
 
This is the camera-holding wing nut with its threads filed.  Since the bracket is tapped but the final threads are gone,  the wingnut can spin in place to tighten the camera, but won't fall out.
 
 
I haven't taken it anywhere cool yet, but with the previous version (pan motion only) I took this picture at Coyote Point in San Mateo, CA.    The rattle doesn't seem to affect sharpness for sunny afternoons. 
 
 
 
 
 
 
 
ċ
KAPrig.jpg
(139k)
Bill Maney,
Dec 5, 2011, 1:26 PM
Comments