Cornell HGR Team Status Report

1
Cornell HGR TeamStatus Report

• 4/8/05

2
Schedule for remaining weeks
Fletcher
Week of 3/28 Decide on format of final
rover Preliminary brainstorming and
re-design Order some parts to start
construction Week of 4/4 Complete rover
design including parts list CAD rover
Order all parts Week of 4/11 Begin
construction Flywheel (No special parts
needed) Negative Friction Foot (NFF)
Hoptuator integration Test
electronics Determine power consumption
by experiment Week of 4/18 Continue
construction
Week of 4/28 Construction complete
Assembly Debug and run rover Week of
5/2 Test rover Hop in paths,
running time, etc. Begin writing final
paper Week of 5/9 More testing
Finishing paper sections Week of 5/16
Proofread paper Hand in paper, test
results
3
PrototypeSystem

• Power
• Separate sources
• Hoptuator (3x9V NiCad)
• Everything else (9V)
• Sensors/Inputs (4 total)
• Primary voltage
• Hop power voltage
• Flywheel tachometer
• Foot angle
• Outputs (6 total)
• NFF motor control
• Kickstand power position
• Flywheel speed direction
• Hoptuator

4
Primary System Power Requirements (No Hoptuator)
Fletcher

• The primary system has low power requirements
• Should be easy to meet with the 9V Duracell
  (above)
• Runtime of 30 minutes with minimal voltage drop

5
Hopping Solenoid
Eugene

• Manufactured by Saia-Burgess
• Ledex Low-profile Series

• Size 2EC
• Rated _at_ 28V
• 2.5 Amps
• 4 58 N
• Wt 64g
• Ø1.125" x 0.580" L
• 0.240 stroke

• Size 4EC
• Rated _at_ 26V
• 5.0 Amps
• 27 - 107 N
• Wt 170g
• Ø1.562" x 0.835" L
• 0.250 stroke

• Size 5EC
• Rated _at_ 28V
• 7.5 Amps
• 42 190 N
• Wt 326g
• Ø1.875" x 1.035" L
• 0.400 stroke

6
Power Source
Eugene

• Hoptuator Requirements
• 26 volts or higher
• Short (0.5s) bursts of high current (6A)
• Total ON time less than 60s
• Energy demand 90 mAh
• Top-side Requirements
• 9V system
• Energy demand 150 mAh
• Non-issue with Alk. 9V

• 9V DC Batteries
• Alkaline (Energizer, Duracell)
• IR 1.70 O
• Energy Cap 625 mAh
• Wt 46g
• Sanyo Industrial Ni-Cad (N6PT)
• IR 0.21 O
• Energy Cap 110 mAh
• Wt 42g

7
Hoptuator Power Constraints
Eugene

• Current-limited
• Sanyo has low energy reserve (lt20)
• Need to fulfill power requirements
• COTS components vs. Specialized Battery Designs

8
Angle Sensor
David

• Potentiometer shaft is attached to leg hinge
• Basic Stamp measures time to charge capacitor
  through potentiometer to calculate the leg angle
• Potentiometer
• Low Friction
• Small
• Cheap

Leg -gt
Potentiometer -gt
Foot -gt
9
Battery Sensor
David

• Similar to angle sensor circuit
• Basic stamp measures how long it takes battery to
  charge capacitor through a resistor
• Battery level is calculated from capacitor charge
  time

10
Aya
Negative Friction Foot Animation
11
Negative Friction Foot
Aya

• Changes from prior nff
• Incased
• Avoid sand
• Gear drive
• Friction drive problems
• Smaller motor
• No need for large motor
• Manufacture hinge
• Wiggle in pre-manufactured hinge

• Units in inches -gt

12
Negative Friction Foot
Aya

• Parts
• Motor
• 4mm Namiki pager motor
• Bearings
• 3x .25 bore, .375 OD, .125 WD
• Gears
• 2x Aluminum spur gears
• 1.5 inch
• .25 inch

13
Optimized Flywheel
Ben

• Reduces weight
• Material is only taken off of one face
• Drive shaft is press fit while the flywheel is
  still in the chuck
• Flywheel is cut off on the lathe
• POC Optimized flywheel is well balanced, proof
  that this fabrication method works

14
Frame
Ben

• Structural frame covered in lightweight skin
• Frame supports the weight when the rover is on
  the ground
• Skin seals the flywheel and electronics from dust
• Easily made out of sheet metal (frame) and carbon
  fiber/fiberglass/plastic (skin)

15
Omni-directional Kickstand
Ben

• Design for a kickstand that works on all sides of
  the rover
• Uses one rotary actuator, with potential for
  using the driveshaft from the flywheel instead

16
Kickstander
Matt

• Our solution to getting the rover to stand is
  installing a kickstand.
• How does it work?
• Pros
• Simple Solution
• Lightweight, easy to build
• Got it to work well on the first try
• Cons
• Rover cant stand up from any fallen position
  without a more intricate mechanical kickstand

17
Next steps
Matt

• There are a couple different leg shapes
  considering
• U-shaped ski?
• Wheel?
• Final placement dependant upon packaging
  constraints.
• Preferred position is just under the rotor.
• If there is no space there, then we can mount the
  servo further down the leg and attach a linkage
  to the kickstand.

18
Gimbal Controller Testing
Julia
Important Features
bearings

• The idea test control law on more balanced,
  stationary rover model
• Little redesign
• Use old POC model
• Changes
• Precision
• Bearings
• New Hinge
• Large Inertia Flywheel
• Increased symmetry
• Adjustable counterweight
• Batteries below pivot
• Fixed, rotating base

Sliding counterweight
Gimbal motor
bearings
batteries
19
Controller Testing
Julia

• Sensor inputs for feedback
• Tilt angle / rate
• Precession angle / rate
• Cant measure with accelerometer
• Tachometer/speed sensor on base?
• Compass (just need period info, really)
• Gimbal angle / rate
• controlled

• I have a control law that works (!)
• I am working on gain-scheduling (initial
  condition dependent gains)
• Comment on figure wall effect independent
  of initial condition