Intro
Who
EngineerComponent Library, Layout: Meghan Dang
ManagerArchitect, Component Selection, Schematic, Reviewer, Schematic: Daniel Puratich
Firmware: Derek Tang Nathan Green
Advisors: Aidan Bowers, Kevin Li
Why
we want to space ESCs out, have a higher datarate, and less harness complexity
CAN supported ESCs exist but more expensive and not what we have alr, we like what we have right now
DSHOT is already a digital protcol that’s busable
single ended
prone to ground loops and common mode SI
100mil hdrs into pixhawk and from ESC
What
board solders onto the ESCs both DSHOT data pins that has a JST CAN connector to pixhawk
the firmware and hardware will not implement an ESC, we will continue to rely on COTS for this assuming the COTS solution features a standard DSHOT 100mil pitch connection (as the APD ESCs we have right now do)
firmware wise this is a CAN to bidirectional DSHOT adapter.
How:
small STM and a CAN transceiver, may require a little LDO for power, see below for design procedure
When
Starting Early June 2024
Where
Table of Contents
Table of Contents | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
|
...
This is a starting place for a simple plane example, but shows how a layout could look with a full CAN architecture. Eventually battery bay could support more CAN and reduce harnessing there, but the battery bay and avionics will be very close to eachother so they’re lower priority than the farther away stuff.
Drawio | ||||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
...
100mil hdr mount onto it
Micro Selection
some STM
https://discord.com/channels/776618956638388305/1241377889035882646/1242282003018154025
LFSPXO072801 ← possible 8mhz HSE, used on the ark one, seems fine to me
CAN Transceiver Selection
https://discord.com/channels/776618956638388305/1241377889035882646/1242282003018154025
copy pixhawk? kinda yea but chosen in discord and no real contest
no need for CANFD imo
common mode choke
need to select
SRF2012A-121YA
https://www.digikey.ca/en/products/detail/bourns-inc/SRF2012A-121YA/5429616
ACM2012-201-2P-T002
...
Power Decision
ACT45B-101-2P-TL003
ACM2012-361-2P-T002
B82786L0104N002
https://product.tdk.com/en/products/selectionguide/cmf.html
doesnt even include CAN lmao
https://product.tdk.com/en/techlibrary/productoverview/automotive-cmf_can-fd.html
ACT1210D
ACT1210D-510-2P-TL00
switched termination
done on ARK board with a low side FET
seems like a bad idea if it gets GND offset, other designs online dont show this, see below
https://www.analog.com/en/resources/reference-designs/circuits-from-the-lab/cn0401.html
https://toshiba.semicon-storage.com/info/TLP175A_datasheet_en_20230525.pdf
https://www.digikey.ca/en/products/detail/toshiba-semiconductor-and-storage/TLP175A-TPL-E/4562829
commit to this one, cheaper
high speed TVS
https://assets.nexperia.com/documents/data-sheet/PESD2CANFD24U-QC.pdf
exmpla pre nice
PESD2CANFD24U-QCZ
NUP2105LT1G
Power Decision
LM66100 for RPP
LDO to power the transciever?
how much capacitance, decent amount, same damping would be nice too lol
https://www.digikey.ca/en/products/detail/texas-instruments/TLV75533PDBVR/9356541 for LDO
Schematic High Level
Drawio | ||||||||||||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Board Outline Sketch
Roughly how will system look? nah no point, see the schematic!!