Rev. 1 works!

A single Rev. 1 board pair (MCU and transceiver) was assembled and tested; they communicate successfully with Rev. 0. This marks the first time we have successfully bridged two CAN buses!

Pictured above is our test setup. The Rev. 1 pair is on the far left and attached to a PC using a CAN-USB adapter. The Rev. 0 board (far right) is attached to a separate laptop PC using another CAN-USB adapter.

Risks

• The ribbon connectors on the Rev. 1 boards are unfortunately facing the wrong direction. The cable can still be attached, but it must enter the connector from the opposite side, causing interference on the transceiver board.
  • Since the cable only interferes with one transceiver “slot,” the other one is temporarily used instead. This is not a significant concern, as we only use one radio at a time anyway.
  • A future revision will re-orient the connectors.
• The ribbon connector is intended to be used with the cable’s contacts facing the board (“downwards”), but the pin layout requires the cable’s contacts to face away from the board (“upwards”).
  • The connector still appears to work with the contacts facing away from the board; we will continue to use it this way for now in order to proceed with debugging.
  • A future revision will either use a reverse-direction ribbon cable, or a “top-facing” connector on one side.
• Timeline is getting somewhat stressed.
  • Core features appear to remain on track, but we will continue to monitor our progress in the coming weeks, especially as testing reveals potential flaws.
  • Depending on how much slack is remaining, the Ground Control UI’s functionality may ultimately be limited to TOM configuration, as it is the most critical feature.
• “Unknown unknown”: Either the inaction of XB1 attention line or the “full-duplex” operation of our SPI drivers is preventing two-way communication back to Rev. 1.
  • Debugging will continue this weekend and into next week.
• Range testing still to be done.
  • Testing will occur soon in the FMS breezeway and parking lot.
  • Additional testing will likely occur at PittRace; the team plans on heading there for track testing in a few weeks.

Design changes

• No design changes have been made.

Updated schedule

• Rev. 1 confirmed to be functional, so we’re fast tracking the next revision for ordering.
• Other schedule tasks have been adjusted to reflect actual task completion dates; dependencies have been shifted accordingly.

Risks

• We tested the Wi-Fi XBee range at a 1 km line-of-sight stretch of the Great Allegheny Passage in McKeesport. The Wi-Fi range was significantly shorter than predicted using the 9dBi antennas, at approximately 200m. Additionally, the directional 19 dBi antenna tests were inconclusive.
  • We plan to recreate these tests in the FMS breezeway and parking lot.

Design changes

• No design changes have been made.

Updated schedule

• No schedule changes have occurred.

Individual updates

• Wi-Fi proof-of-concept has been implemented and tested.
• Bidirectional uncompressed CAN streaming works well (plug-and-play).
• More robust integration and preparation for rev. 1 is necessary.

Schedule status

My deliverables remain on schedule.

Deliverables for 2019 Mar. 16

• Nothing scheduled, as it is spring break and I am out of town.

Risks

• Layout for the 2-PCB design has been completed, but has not yet been reviewed.
  • See Monday’s presentation for review.
  • 2-revision timeline remains on track to ensure any mistakes from rev. 1 can be remedied in time for final demo.
• Firmware implementation may fall behind slightly due to 18-500 design report and homework overhead.
  • More time has been allocated in the schedule by pushing back UART interface implementation, as it is closely coupled with firewall and configuration work.

Design changes

• No design changes have been made.

Updated schedule

• Pushed revision one ordering back by a few days as it has not yet been reviewed (See our Gantt chart for more details.)
• UART interface work has been moved to after spring break, to coincide better with firewall and configuration implementation tasks.

Individual updates

• Merged SPI radio API proof-of-concept
• Continued various driver cleanup efforts
• Made more concrete plans for firewall and configuration implementation
• Spent most of the week preparing the presentation and writing the design report

Schedule status

As discussed in the team status update, UART bringup has been moved to after spring break to better coincide with the firewall and configuration implementation. Other items are in danger of falling behind because of other 18-500 work (design report, homework).

Deliverables for 2019 Mar. 9

• More Wi-Fi XBee integration and testing
• Other integration tasks: codec implementation is becoming more concrete, so it will likely be added into the stack and tested soon.

Individual updates

• Further work on the radio/SPI API to improve its robustness
  • The refactored interface has been tested for both receive and transmit; it is much simpler and more reliable
• Implemented interrupt-based CAN transmit mailbox synchronization to allow better coordination of the limited number of hardware CAN mailboxes (see below)
• Bridging proof-of-concept underway; initial tests will packetize, but not compress

Schedule status

UART driver bringup is falling behind, but the radio API implementation and testing is higher priority. The bringup will not take very long, so it will probably be delayed in the schedule until the configuration API becomes clearer.

Deliverables for 2019 Mar. 2

• Bridging proof-of-concept completed and tested (preferably at >1 m range)
• Integration plans for codec and configuration APIs

Risks

• Layout for the 2-PCB design has been completed, but has not yet been reviewed.
  • See Monday’s presentation for review.
  • 2-revision timeline remains on track to ensure any mistakes can be remedied.
• Preliminary range testing with the 2.4 GHz “9” dBi dipole antennae and Wi-Fi XBees proved successful at ~250 m, but it was likely not representative of our target environment. Furthermore, longer-range testing was inconclusive.
  • Currently looking into whether or not CMU has a suitable RF chamber for our testing purposes.
  • More range testing will occur when the system has been further realized.
  • A 2.4 GHz “19” dBi directional dish antenna has been acquired (c/o Carnegie Tech Radio Club), and will be tested.
• Risks that have been resolved as of this week’s entry:
  • Partially resolved the above issues from last week, but as indicated, further mitigation is required.

Design changes

• No design changes have been made.

Updated schedule

• No updates: We’re still on schedule in terms of PCB schematic and layout, as well as firmware bringup and implementation. (See our Gantt chart for more details.)

Individual updates

• Continued firmware bringup efforts that started in late January:

  • Worked with Deepak Pallerla and other CMR members to refine CAN, ADC, and GPIO drivers and their respective interfaces

  • SPI via DMA implemented and tested (below, left)

• Designed high-level TOM system architecture by outlining required tasks and synchronization objects, including the radio API

• Implemented and tested synchronized full-duplex SPI interface proof-of-concept

Schedule status

I am currently on schedule for my assigned tasks. Radio API design was at risk of falling behind due to missing XBee hardware earlier in the week, but they have since been sourced; I will have enough spare time this weekend to test out various configuration details.

Deliverables for 2019 Feb. 23

• UART drivers brought up and tested

• Radio API tested with XBee

  • Various Wi-Fi configuration/connection details must be ironed out

  • Full-duplex must be tested for robustness

• XBee range tested