The cannon-lighting
clown bug, and the
servo underneath that
moves him.
cannon mechanism
rescue diagram
Rescue device
The Rescue Device

The rescue device uses the same type of rail and slide, only this time the slide holds still, and
the rail moves (Figure 10). A slower speed allows a smaller winch diameter, and the stepper
motor is strong enough to operate it all without a counterweight. Limit switches detect when
the device is at the bottom, and when it is loaded with a marble. I didn’t feel a switch was
needed at the top of travel, since the number of steps that the motor makes can govern this
quite well. It just needs to start from the lowered position. It is then possible to finely tune the
height when it is raised by changing the program on the stamp.

The Cannon

The cannon is shot by a 120VAC solenoid, where a cam arrangement transfers the travel of
the solenoid into pushing a rod into the bottom of the cannon (Figure 11). I tried 12VDC and
24VDC solenoids to do this, but none of them were strong enough at the currents that I used
to do the job. The AC solenoid gives 7 lbs. pull when it shoots the marble 6-15” high, and
more would be better. It is triggered by a Triac driver (optical coupler) and a Triac. I thought
that the Triac would give a predictable velocity to the shot since it always switches at the same
place in the changing AC voltage, but this was not the case. I think the shooting power is
variable because of less-than-perfect sphericity of the marbles, and less-than-perfect walls of
the PVC pipe that I made the cannon barrel from. If one in 10 shots miss the basket, then we
get to watch the ball rescue device in operation.

The cannon is “lit” by a circus clown bug. He holds an LED “match” in his hand. The match
lights, and he leans over to light the LED “fuse”. As he leans over, he extends his match arm
to light the fuse, kicks a back leg into the air, and turns his head 90° toward the cannon. A
servomotor moves it all. The bug’s body movements are managed by the front of the bug
rotating on a different axis than the back half of the bug. When he leans over, the relative
motions of the two halves can make the arm, leg and head move. This part of the project
made me feel like a watchmaker. Do you think Mr. Rolex ever played with servos, LED’s and
bugs? Maybe not, but then he probably had all of his marbles.
Servos need to be on a separate power supply than the Basic Stamps. The servos draw a bit
of power, and cause the voltage to drop enough to “brown out” the supply, which would
cause a stamp on the same supply to reset.
Inside of the
clown bug
The LED Cascade

Ten circular rows of 9 (5mm) LED’s surround a tube through which two marbles travel. The lights follow the path of the
marbles as the go down the inside of the tube. Actually, the first marble hits a microswitch before it gets to the tube, and the
lights follow a predetermined sequence that appears to follow the balls. The mechanical marble switch that precedes this lets
go of two marbles at once (Figure 13). A marble separating mechanism is needed so that the two marbles are not right next to
each other, to get the effect that I wanted. A tight spiral inside the tube makes the marbles go down the tube more slowly as
they spin their way down.
Circuit diagram. Only one
input switch and one stepper
motor are shown, since they
all are wired the same. All pin
connections are listed on a
spreadsheet.
Point-to-point
wiring, 30 ga. wire.
More Electronics

The whole thing (Figure 12) was wired with point-to-point wiring on a perforated circuit board with
pads at each hole. A printed circuit would be nice, but since I made a lot of changes, additions and
deletions during the development I think point-to-point was actually easier. Connectors were essential
to be able to plug and unplug the board from everything else. Resistor networks would have been nice,
if I had only realized from the beginning how many resistors of the same values would be used.

I used 3 Basic Stamps, using each one to control several functions. This saved cost, but added
complexity to have the microcontrollers alternating tasks. Using more stamps would have raised the
cost, but would also have saved a lot of time in getting all of the activities to coordinate with each
other, and would simplify the programming. I used a BS2SX, a BS2p24, and a BS2p40. It all could
have been done with 4 or more BS2’s, but I used some stamps that I already had on hand.

My power supplies were wall warts and table top supplies. The stamps need an electrolytic capacitor
across the output of their supply, to help avoid brown-out and resetting. The spot lights are 12VAC
halogen reflector lights (a 35W and two 20W), using a power supply made for these.