Subsystem Build & Test
Table of Contents
Team Vision for Subsystem Level Build & Test Phase
- Fully build tail prototype and do initial testing outside of water
- Complete walking mechanism to be able to walk on its own
- Test electrical components individually and as a system
- Finalize full CAD design
- Fully built tail prototype and started initial testing outside of water
- Built complete walking mechanism able to move on a stand
- Tested electrical components individually and as a system
- Almost finalized full CAD design
Test Results Summary
Tail Prototype and Testing
Leg System Prototype and TestingAll four legs were re-designed to fit with the new servos that were ordered to meet our torque specification. They were then printed and mounted onto a mock-up board in order to test movement. K'nex were used at the joints because the 3D printed pins were causing to much friction. We are looking into roller blade axles for the final product. The legs are able to move but the structure cannot hold up its own weight because the connections to the servo are not secure. This will be fixed with better hardware which will be fixed in the coming weeks. This will allow us to work on the gait.
Rib and Spine Prototypes
Mini Maestro Testing
The Mini Maestro motor controller has been tested and implemented to produce all the necessary PWM signals at various duty cycles and transition speeds. The maestro has been programmed through Arduino, and the manufacturer Pololu supplies a Library for easy integration.
The maestro is setup as shown in the image below. Testing Videos of the maestro in action with several servo setups. As shown in the single servo test with a scope, the PWM signal controlling the servo is visible in the background. It is also
RC Controller Testing
The FlySky Transmitter and Receiver used for the Robotter is capable of sending 6 signals which correspond to the controller's joysticks and various knobs and switches. The first four of the channels are reserved for the joysticks, and the other two can be programmed as needed.
Testing each of the channels, it is evident that the system sends out PWm signals ranging between 1-2ms, depending on the physical mechanism producing the signal. The figure below outlines each of the channels, what they correspond to, and the ranges for each one.
The pinout established for the robotter is displayed below. When connecting power to the BAT channel, the voltages actually translate throughout each of the other channels. As a result, only the PWM signal data needs to be connected to the Arduino. 3.3 V from the arduino is used as VDD in this case. The other pins connect directly to the PWM inputs. The code used reads each pin under a 30ms period, to determine the pulse widths. This allows for minimal delay between user and the robot response.
The RC Controller testing showed that the controller can communicate with the arduino real time.
Servo Water Proofing Tests
A set of cheap servo motors were purchased for the purposes of perfecting our motor waterproofing procedure before using it on our permanent leg motors.
The test procedure used is:
- Test initial motor function and performance using this simple Arduino program.
- Remove motor casing and cover electronic components with a silicone based adhesive and use a waterproof WD-40 variant on the motor's gearbox
- Reseal the motor case and leave for the sealant to set (at least 8 hours)
- Test motor function and performance with a dry motor using the same program
- Test underwater motor function and performance after the motor has been submerged in water for at least a minute
- Leave the motor submerged in water overnight (at least 12 hours)
- Test underwater motor function and performance one last time.
Test is considered successful if the motor is waterproof (still operates after extended time underwater) and sees no noticeable degradation of performance over the 4 tests.
So far 4 motors have been tested. The first two were successfully waterproof but saw performance issues starting with the second test after the sealant had set. The third and fourth tests were successful on both accounts.
Some more pictures and videos of the process can be found here.
We still have 4 untested 'practice' motors, so we plan to make use of those before waterproofing the actual leg motors.
Preliminary Buoyancy TestA preliminary buoyancy test was done using our electronics box. A series of 0.5 kg masses were added to the inside of the box in order to determine the load at which it can no longer float.
The limit between floating and sinking is somewhere between 2.5 kg and 3 kg. This is heavier than we expect the finished Robotter to be, so we now believe that lack of buoyancy should not be much of a threat.
Risk and Problem TrackingChanges to our Risk Assessment for this phase include:
- Downgrading the Likelihood of a few items
- Adding a risk of legs not supporting weight
- Adding a risk of our waterproof connector failing
- Adding a risk of Robotter not being able to walk straight
A working copy of our risk assessment can be found here.
The problems that we ran into during this phase and our solutions can be found in the above sections.
Functional Demo MaterialsOur Week 6 Functional Demo presentation can be found here.
Notes and actions items from the demo can be found here.
Bill of Materials
Our Bill of Materials can be found here: Bill of Materials
Plans for next phase
- Finalize shoulder and hip designs
- Start system integration testing
- Waterproof leg servos and begin in-water testing
A working copy of our project plan can be found here.