Two Paper Birds and a Follower That Stayed Up

Cardboard Cam Automata 🎮 Play: Cam Motion Studio

Both paper birds rose. Only one came down. The second remained above its cam while the handle completed another turn, maintaining the posture of someone waiting for an apology. Cardboard Cam Automata, hobby #244, had supplied a useful first problem: how to make a follower actually follow.

Contact is a condition

A cam turns rotation into a prescribed movement. Its outline pushes against a follower, which carries that movement to the sculpture. Rob Ives’s explanation of cams makes an important distinction: the follower’s shape also affects the result. A pointed contact traces details that a broad contact may bridge.

For this build, I chose a circular cam mounted off-centre and a broad, flat follower underneath each bird’s upright rod. The cam lifts the follower; gravity brings it back down.

That second action needs its own inspection. Nothing connects the follower to the cam. If the rod sticks in its guide, the cam can continue underneath an entirely stationary bird.

       rod to bird
           |
        [ straw ]       fixed guide
           |
      -----+-----       flat follower
           ^
        contact
        ( cam )         turns beneath follower
           o            horizontal axle, end view

The sketch shows the arrangement, not dimensions. The follower’s underside must remain horizontal, with enough width to cover the contact point as it moves sideways across the rotating circle.

A circle with its centre elsewhere

An eccentric cam needs no elaborate outline. Cut a smooth circle, mark its geometric centre, then put the axle hole beside that mark. The distance between those points is the eccentricity, e.

As the axle turns, the circle’s centre travels around it. For a horizontal flat follower that stays in contact, the circle’s highest point rises and falls with that centre.

The ideal geometry is:

follower height = axle height + radius + e × cos(angle)
total vertical travel = 2 × e

Here, the angle starts with the circle’s centre directly above the axle. At the highest position, the eccentricity adds to the height; at the lowest, it subtracts. An offset of 4 mm therefore gives 8 mm of travel: 4 mm above the middle position plus 4 mm below it.

That describes a rigid circle and a level follower. A rough edge, tilted shoe or bending axle changes the actual movement.

For alternating birds, mount matching cams on the same shaft with their offsets pointing in opposite directions. A half-turn is 180 degrees, so one follower reaches its highest position as the other reaches its lowest. I marked both offset directions in pencil before assembly. Otherwise, “opposite” becomes a judgement made through a cardboard window.

Give the return stroke somewhere to go

Before attaching either bird, this was my contact test:

  1. Lift the follower by hand and release it. It should descend through its working range without help. If it hangs, inspect the rod and guide before changing the cam.
  2. Turn the bare mechanism slowly. Watch the contact from the side. A gap appearing beneath the follower reveals that its descent has stopped matching the cam.
  3. Check the guide without squeezing it. A drinking straw supports the upright rod, but a crushed or crooked straw becomes a brake.
  4. Attach the figure and repeat. The bird changes the load and may pull the rod sideways. Test the assembled load before fixing the cam permanently.

The Exploratorium’s construction guide recommends a washer or nut when a follower needs help descending. It also leaves the cam adjustable until the movement works.

Weight helps overcome modest resistance. It also increases the load the crank must lift, so adding washers indefinitely is an expensive way to avoid straightening a straw.

Speed introduces another limit: the follower needs time to fall. A mechanism that works slowly can lose contact when cranked faster. Slow testing separates that behaviour from a rod that sticks even when released by hand.

My marble-run session stalled while I checked a ramp with a protractor, measurements and video. Here, the useful observation was smaller: watch the gap. Gravity can only return a follower that has a clear path.

By the end, both birds travelled up and down, their folded bodies tilting on loose paper hinges. The second still hesitated when I hurried the handle. I left its guide accessible and the decorative cover unattached. One bird bowed; the other considered the request.