Customer Handoff & Final Project Documentation
Table of Contents
Team Vision for Final Demo and HandoffThroughout the project, the team was aware that this project could be completed entirely, and the goal of this phase was to deliver a fully functional product as ordered by the customer. The prototype delivered should need only improvements should the project be picked up by future teams.
Test Results SummaryTest Results Summary
- Currently only the requirements for weatherproofing and operational temperature have not been sufficiently tested
- All other requirements were met
Future Design Improvements
The prototype was successfully built and found to work as intended. That said, there were a number of findings that might lead to improvements in the future:
Short term improvements to complete the project as is:
- There is a recurring problem with ESD. It could possibly be charge accumulating due to the rubber belt rubbing against metal pulleys. The shafts should be properly grounded.
- Due to line drop across the Ethernet cable, improvements need to be made to the power delivery system. James Stefano has some suggestions involving switching the motors to always run off the battery, and adding a trickle-charging regulator on top of the battery.
- Verify constitution of weatherproofing
- Create calibration sequences
- Verify functionality of Pi/Arduino interface
- Record live satellite pass
- Complete weatherproofing and test
- Mount ground box to pole/railing
- Install on rooftop
- Create a better counterweight system from a single block that is threaded to the rod
Long term design improvements to make if the project were to continue to a new MSD team
- During the design phase in MSD 1, it was thought that a microcontroller would be needed to manage high-speed active feedback for motor control, but since the project ultimately ended up using stepper motors, this high-speed feedback isn’t needed. So, the Raspberry Pi could run the whole system.
- During the design phase in MSD 1, it was thought that should the battery need to power the motors, it would need to be in close physical proximity, to avoid the inductivity of the ethernet cable. So much capacitance needed to be added though, that the battery could most likely be moved inside. If the RF chain was also moved inside the gimbal (which is acceptable, since it was shown through testing that the electronics don’t interfere with the signal), the external “pole box” could be eliminated entirely.
- The feedback path for the lab supply could sense the actual voltage at the gimbal, to correct for dynamic drop on the line. This would provide a more consistent supply voltage to all loads.
- Possibly add a web interface to set paths rather than text-based SSH session.
- Pull satellite paths from online database rather than manually entering pass parameters.
- Improve weatherproofing
- Create PCBs for all power components
- Improve heatsinking on motor drivers. Don't use eval board, include them in the custom PCBs
- Add a camera to the installation site so that the system can be actively observed
- Add lightning protection
Risk and Problem TrackingFinal Risk Summary
Retrospective Analysis of the Project and it's execution has lead to the following conclusions:
The following details the things that went wrong with the project:
- A lot of debugging of the electrical systems was
performed in 2 weeks before Imagine. EE needed motors
with mechanical loads to truly test the power system.
- We didn’t think we had the budget initially to afford a 3rd and 4th motor.
- The mechanical engineers were too busy to build a mechanical load.
- The project should have had another mechanical engineer.
- EE didn’t have the system to mechanically test
the DOF with.
- We could have gotten a lot farther with simulation. EE should have done more testing with the DOF earlier on.
- Same goes for limit switches.
- Nick specifically shouldn’t have tried to do all the electrical on his own. There were 3 EEs and this work could have been split up.
- Code review was never done.
- Electrical hardware review was barely done, and only after most of the system was assembled.
The following details the things that went right with the project:
- A functional prototype that serves as proof of all concepts was created from scratch in 2 semesters.
- Many roadblocks were encountered during the design and fabrication, most were solved, some were avoided.
- All team members remained dedicated to project for 2 semesters, and were ready to actively contribute.
- Several team members were ready to assist with roles not initially associated with their major (ex. EE helping with software design).
Final Project DocumentationGimbal Schematic
Final Presentation MaterialsTechnical Paper
Plans for Wrap-upThe project is nearly complete, but the remaining tasks unfortunately prevent it from being used reliably (see 80/20 rule).
The pole is installed on the rooftop, and is ready to be used.