commaai
Rivian: "motor fault"
We haven't seen this in our Rivian, but apparently some users are experiencing a fault on startup.
A couple users hitting this:
- https://discord.com/channels/469524606043160576/1196513491025133638/1411490296600203326
- https://x.com/itskyleconner/status/1961926362698223857
- Motor fault from troy: 0c832cd010dc203a/00000039--36cfb0e2a3/1
- Motor fault from rawsignal: https://connect.comma.ai/574129f33798790b/0000001f--1299806aaf
https://www.google.com/search?q=rivian+motor+fault+detected returns many forum results of non-comma users having similar issues. Since this can happen before the comma 3X even boots, it's unproven if the rate at which these faults occur increases with comma 3X usage.
- https://www.reddit.com/r/Rivian/comments/191jezf/motor_fault_detected/
- https://www.facebook.com/groups/RivianEV/posts/1881270055963600/
- https://www.rivianownersforum.com/threads/r1s-died-at-stop-light-mutilple-fault-codes.6223/
- https://www.rivianownersforum.com/threads/r1s-gen-2-motor-fault-and-bad-efficiency.8854/
- https://www.edmunds.com/car-news/our-long-term-2022-rivian-r1t-quits-without-warning.html
- https://www.reddit.com/r/Rivian/comments/10xixjt/motor_fault_detected/?tl=ms
- https://trailrivian.com/threads/r1s-gen1-motor-fault-and-system-fault.1037/
- https://www.torquenews.com/18000/800-miles-my-rivian-r1t-and-its-already-lost-its-rear-motor-and-gone-turtle-mode-now-im
Our findings so far:
- We are unable to reproduce this issue on the comma R1S.
- Most users have not experienced this issue.
- There are plenty of instances online of non-comma users experiencing this issue.
- It's not clear what the base rate is, and whether installing a comma increases the rate at which this happens.
We'll leave this issue open while we continue to look into it, but so far this seems like a Rivian bug that a comma 3X may exacerbate.
For any users experiencing this, answering the following would be helpful:
- are you using any third party (e.g. Lukas/xnor products) hardware?
- do you have the comma power installed?
- did it happen while running stock openpilot or a fork?
- did you experience the fault before installing a comma?
- how long did you drive with your Rivian before installing the comma?
Interesting findings from https://www.youtube.com/watch?v=QdBDgDWfrU4. Apparantly the car will wake up from deep sleep to charge the 12V battery when it gets low, also causing phantom drain in the HV battery. If this is the case, unclear why the external load from 3X while it's in deep sleep would potentially cause motor faults on boot.
Rapid Discharge Cycles: The battery, designed for internal combustion engine vehicles, is not suitable for the demands of the Rivian R1. When the vehicle is asleep, it discharges the battery in 8 hours, then recharges it for 3 hours, repeating this cycle multiple times a day (1:25). This constant cycling wears down the battery quickly, despite it being a "deep cycle" battery (1:42).
Looking at a few dongles with confirmed motor faults, I don't see the 12V battery voltage drop below normal levels:
Fault route for second user: https://connect.comma.ai/574129f33798790b/0000001f--1299806aaf
Not sure why but it looks like the battery is not reaching the peak voltage before the fault. 14.6 and 14.2 respectively. If comma comes on at the wrong moment could it be pulling down voltage slightly when BMS is testing?
Dont know if youd have to build it into the giraffe but perhaps there would be a way to ignore boot trigger signal when voltage is above 13.6? (if this is indeed the problem?)
If the device was off, we don't have any voltage logs. So we can't say for sure what voltage it was at at the moment it started up in that case. Plus this is recorded every 10 minutes.
we could work on this from a different perspective. theres cameras in rivians such as ones for amazons EVs what about regular vehicles? did anyone ever think the motor from the rivian is connected to the camera already attached by rivian not just from the 3X? reason being it could be a motor limiter thats protecting the vehical sort of like a security system.
We have approached this issue as a potential conflict on the CANbus as both devices boot up simultaneously. We have had potentially good results using a physical USB-C switch on the OBD-C cable to delay the startup of the Comma until the Rivian's wake processes have had a chance to complete. We are using this product: https://www.riitop.com/collections/usb-c-cables-1/products/riitop-usb-type-c-extension-cable-with-on-off-switch-1ft-usb-3-1-type-c-male-to-female-extension-cable-with-on-off-switch-support-video-data-pd-charging
Just a follow-up, user @geomglot on Discord has been testing the switch mentioned above for several weeks now. He was experiencing several motor faults but has not experienced any since using the switch to delay the Comma's booting up. He'll continue testing, but this does seem to point to some sort of conflict between the Comma and the Rivian during the wake process.