-
Notifications
You must be signed in to change notification settings - Fork 0
X12 Backplane
IMAGE OF PCB, potentially including its place on the relevant ROV
| PCB Summary | |
|---|---|
| Vehicle | X12, ROV Triton |
| Contributors | Melchor De La Garza, Grant Geyer |
| Predecessors | X11 Backplane, X10 ????? |
| Success? | Yes, after several revisions |
The backplane houses several boards and routes power and communication signnals to their appropriate destinations.
The backplane is in the power box. Inside the powerbox, it connects to the conversion board and 3 Quad ESC boards. It has connectors for CAN and 5V power that go to the pi and pi shield in the logic tube, CAN and 12V connectors for the solenoid board in the solenoid enclosure, and CAN debug and programming connections that go to capped binder panel jacks.
What priorities did you have in your design? What design considerations did you have? What methodologies did you follow? (routing a differential pair, keeping something separate for isolation, etc)
- Big Power for ESCs
- Route CAN with good methodologies for low noise
- Keep the CAN traces close, with ~8mil between them. The differential pair routing too in Eagle is very useful for this.
- Easy to use connectors
- Fit all boards
- Program micros on ESC boards
Can to connectors. Boards connections switch too (see SID). V3 included muxes for programming the ESCs, though they went unused.
Length of the box and covering of ethernet ports on the switch.
See the mux datasheet.
- 17 A on the 12V / ground line (V3).
- 5ish A on the 5V line.
- CAN at 1 Mbit/s
Why did you pick certain components for your board? (If you don’t know the answer/were told, now is a great time to ask)
- Samtec ET60 (of which a male vertical connector does not exist and shouldn't have been picked. But Scott made one.)
- Samtec IPS1 and IPT1 for connection to the ESC boards. They were easy to get and don't bend when unlike like regular headers sometimes do.
- TE 3.5mm pitch buchanan blocks. Screw terminals were easy connectors to make and the connectors came in horizontal and vertical versions.
- SN74CBTLV3257PWR as a mux - because it was used on the ESC boards and we had them laying around.
- JLC doesn't leave much plating around through holes, so some connectors were tough to solder.
- Fixed errors in V1.
- Pinout - The conversion to backplane ET60 connection was switch from right angle connectors to vertical connectors, and the pinout was not un-mirrored.
- Pinout 2 - Several signals on the Quad ESC boards were left unrouted.
- ESC spacing - In V1, the Quad ESC boards were too close together and the middle one wouldn't fit if either of the end ones was plugged in.
- GND current - In V2, routing of the CAN bus, 5V, and 3.3V rails cut the ground plane up. So all the return current from the ESCs went through a tiny sliver (enough for maybe a few amps).
- Connector heights/facing - V1 used vertical connectors didn't fit with the ethernet switch right below.
- Connector sizes - V2 used connectors and the initial eagle footprints had through holes too small to fit the connector pins, so the pins needed to be filed down.
- Connector pinouts - Signals to connectors were not grouped adequately. Ex: there was a connector for 5V, not a connector with 5V, CAN, and GND for the logic tube.
(For this question and the previous one, go back to the board/schematic in eagle and either change them if it can be done quickly enough, or make a note for the future) Everything was fixed in V3, though some signals could be re-routed to get more capacity on the 12V line.
Any addition links if relevant
- Any fun side details
Search keywords.
Summarize the first three questions above into a paragraph for our technical report The backplane routes power and logic signals to many different boards and enclosures. The conversion board and three Quad ESC boards connect to it and it connects to the logic and solenoid enclosures. 12V, 5V, and 3.3V power in addition to the CAN bus are routed through the backplane.
Tools
Enclosures
Frames
Other
ROS