| | | Notes, decisions and action items |
|---|
June 4, 2020 | @Adrian Au @Aashmika Mali @Kristen Emiko Shiozaki (Deactivated) For reference: @Jess Douglas Muir (Deactivated) | Steering Stalk | Using the power stalk from MSXII (steering stalk was off-the-shelf) Removing the cruise control from the power stalk - relatively easy to separate Each lever arm is attached to its own set of switches that clicks into their own assemblies so the plan is to remove the cruise control and keep the turn signal one keeping headlight switch which is built into the turn signal, which we need
Wheel has not been touched - rough proof of concept (button layout, rough dimensions for PCB fitting) What was the stalk in MSXII powered off of?
(Makes sense to be off of the board but not entirely sure) Interface board is inserted into the stalk and we think it provides a 3V3 reference Firmware wants to know the power values on the buttons but if this is referenced in the steering wheel, that should be fine? Turn signal can provide a 3-way output if it is reference high to reference high-z to reference ground The turn signal is (prob) a 3-state input (gives a different indication when you signal right, don’t signal and signal left)
Current wheel board 2 boards - one lives in the wheel for the buttons, second (connector board) lives in the steering stalk (unsure of whether it will correctly fit in) We have the wheel board as a button board with 4 buttons at the top and a separate connector board Steering stalk has ‘lane assist’ and ‘high beam’ listed in the drawings Doesn’t know which exposed coppers are for which steering stalk function (control board has 5 exposed metal strips for input from steering stalk) All reference to 3V3 with pull-up (turn signal stalk, max 3v3) Note: stalk goes up and down, in and out headlights should exist (99% sure from Interiors) - did not exist in MSXII but should for MSXIV Someone needs to take the steering stalk from hardware Clockspring is broken but Adrian knows where it’s broken and can direct resoldering (won’t affect turn signals since the clocksprng is about routing info from the steering wheel) 5 Connectors - Connector for 12V in, connecter for push-to-talk, connector for CAN communication, Connector for steering stalk
Connecting to the clockspring (only has 6 pins) Clockspring closer to the wheel has pins, clockspring behind steering stalk has connectors Is there a way to have a nice cable extension from Need 4 twisted connections for the CAN communication and an additional 2 (so 6 in total) Would it be possible to neatly wire the connector board to the button board without going through the clockspring? Control stalk output is the real issue Don’t want an additional controller board because it’s another node in the CAN network, which can break apart the firmware
The steering stalk is a ‘normal’ thing to have in a car and so MSXIV was designed with one Options - We’ll have to modify options soon… SEE IMAGES Keep control stalk architecture with steering wheel and button, have the control stalk output board as another input in the CAN network Keep steering wheel and stalk architecture except have the button board as the only CAN node and find a way for the control stalk output to be routed to the button board Find a clockspring with more wires / pins * Not use the control stalk and move DRLs (headlights) + turn signals to the steering wheel Simplest execution = not using stalk Simplest Mechanical = adding a controller board to the control stalk
Number of pins we need (clockspring only has 6 ) CAN = 4 pins (CAN has 2 lines for both CAN-IN / OUT) 12V to controller board = 2 pins Push-to-talk = 2 pins Horn = 2 pins Steering Stalk Connector = 5 pins → 3 pins Buttons (if routed through clockspring) = 5 pins (2 for power / ground reference - included in the 12V controller pin)
Headphone jack on the centre console? Clockspring Wiring 1 - 4 = CAN 5 = 12 V 6 = Ground
Have to keep the button and controller board, but need to figure out what is happening with the stalk to move forward stuck with the implementation of the turn signal Adrian is not deadset on having the stalk, alright with shifting to the steering wheel May be cheaper in the long-run to have the stalk?
Is there a clockspring with more pins? That would solve a lot of problems… → more pins! Research another clockspring with ideally 9 - 10 pins (option 1) Pin 7 = PTT Pin 8 = Turn Signal Pin 9 = Headlights
|
May 27, 2020 | @Adrian Au @Kristen Emiko Shiozaki (Deactivated) @Catherine Cai @Jess Douglas Muir (Deactivated) @James Carr-Pries @Micah Black
@Filip Tomic @SJ (Unlicensed) @Tom | Takeaways | |
Chassis | Extra tubes running across the bottom of the bottom, on the edges of the occupant space Floor of the pontoons will have a thicker core while the channel running down the center of the car will not (using thinner core for catamaran)
|
Doors | DRI does not exist yet Hinges and latches are currently out of the Structures scope Tabs/Clevices for any sort of hinges - to add it when the chassis design has been finalized Doors can become more structural
|
Seatbelt Mounting | Already designed and built Biggest question - are we mounting to composite or metal of the chassis? Ideal to mount to metal Bulkheads will be structural but mounting it to the chassis tubes makes more sense
|
Brake lines | no changes - still same plan to run down the catamaran
(everything is still in the same place in space) CAD in the actual lines in case cutouts need to be made in the firewall and bulkheads Trying 2 parking brake calipers (heavier than an ATV but not as heavy as a 4-wheeler), decided after thinking from a safety perspective Have a caliper, know what space it will be occupying, how the wire is pulled is up in the air (which is fine - does not matter as much) Interiors will more forward with a mechanical lever, Tom will be an advisor for it
|
Battery Box | |
Parking Brake | Taiping recommended to go with a mechanical parking brake due to concerns for reliability Dual - one on each side, but have to design in the pedal/lever (how to activate it from the interior fo the car)
|
Driver Display & Infotainment | James will be completing the UI this S2020 term Goal is to make a flutter web app, not a react app - currently making sure it works with the Raspberry Pi 0 Jess will be holding onto the screen to test with the Raspberry Pi 0 Infotainment to be discussed later, no DRI atm Open Auto could be run on the infotainment, Jess is concerned we will hit a hard limit on the Pi (pi is too slow and performance will not be smooth). If anything needs to be displayed related to telemetry, it will be through a web app Earlier idea was to have 2 separate Pis (one to control the display) with a display switch
|
Headlights | The headlight itself is massive (17” from one corner to the other) Plan is to 3D-print multiple sections Need to consider materials that will not melt Nita is working on the charger-interface board and will be working on lights next
|