Copter-4.1.3 released!

My guess is that ALT with I2C does not work. Let’s see whether this build works or not to confirm:

arducopter.zip (624.7 KB)

This version seems to be fine too!
On I2C I only have compass right now. I had previously also a lidar on I2C (for altitude control) but it never worked well even on 4.0.5. On this copter I only have opticalflow which works quite well.

07/01/2022 12:15:12 : EKF3 IMU0 MAG0 initial yaw alignment complete
07/01/2022 12:15:12 : EKF3 IMU0 started relative aiding
07/01/2022 12:15:12 : EKF3 IMU0 tilt alignment complete
07/01/2022 12:15:11 : AHRS: EKF3 active
07/01/2022 12:15:11 : EKF3 IMU0 initialised
07/01/2022 12:15:10 : EKF3 IMU0 buffs IMU=19 OBS=7 OF=17 EN:17 dt=0.0120
07/01/2022 12:15:09 : Frame: QUAD/X
07/01/2022 12:15:09 : IMU0: fast sampling enabled 8.0kHz/1.0kHz
07/01/2022 12:15:09 : RCOut: DS150:1-4
07/01/2022 12:15:09 : OmnibusNanoV6 00570041 374D500F 2038394
07/01/2022 12:15:09 : ChibiOS: 45395b6a
07/01/2022 12:15:09 : ArduCopter V4.1.3 (2fb939a1)
07/01/2022 12:15:09 : GPS 1: detected as u-blox at 115200 baud
07/01/2022 12:15:09 : Frame: QUAD/X
07/01/2022 12:15:09 : IMU0: fast sampling enabled 8.0kHz/1.0kHz
07/01/2022 12:15:09 : RCOut: DS150:1-4
07/01/2022 12:15:09 : OmnibusNanoV6 00570041 374D500F 2038394
07/01/2022 12:15:09 : ChibiOS: 45395b6a
07/01/2022 12:15:09 : ArduCopter V4.1.3 (2fb939a1)
07/01/2022 12:15:09 : RCOut: DS150:1-4
07/01/2022 12:15:09 : AHRS: DCM active
07/01/2022 12:15:09 : ArduPilot Ready
07/01/2022 12:15:09 : Initialising ArduPilot
07/01/2022 12:15:08 : Frame: QUAD/X
07/01/2022 12:15:08 : IMU0: fast sampling enabled 8.0kHz/1.0kHz
07/01/2022 12:15:08 : RCOut: Initialising
07/01/2022 12:15:08 : OmnibusNanoV6 00570041 374D500F 2038394
07/01/2022 12:15:08 : ChibiOS: 45395b6a
07/01/2022 12:15:08 : ArduCopter V4.1.3 (2fb939a1)
07/01/2022 12:15:08 : Barometer 1 calibration complete

Ok, sanity check - this still fails right?

arducopter.zip (624.9 KB)

This version doesn’t work. The radiomaster no longer sees the sensors. In mission planner messages it stops here. Can I put the previous version back? Thanks again

07/01/2022 12:36:37 : Frame: QUAD/X
07/01/2022 12:36:37 : IMU0: fast sampling enabled 8.0kHz/1.0kHz
07/01/2022 12:36:37 : RCOut: DS150:1-4
07/01/2022 12:36:37 : OmnibusNanoV6 00570041 374D500F 2038394
07/01/2022 12:36:37 : ChibiOS: 45395b6a
07/01/2022 12:36:37 : ArduCopter V4.1.3 (2fb939a1)
07/01/2022 12:36:37 : Frame: QUAD/X
07/01/2022 12:36:37 : IMU0: fast sampling enabled 8.0kHz/1.0kHz
07/01/2022 12:36:37 : RCOut: DS150:1-4
07/01/2022 12:36:37 : OmnibusNanoV6 00570041 374D500F 2038394
07/01/2022 12:36:37 : ChibiOS: 45395b6a
07/01/2022 12:36:37 : ArduCopter V4.1.3 (2fb939a1)
07/01/2022 12:36:36 : Frame: QUAD/X
07/01/2022 12:36:36 : IMU0: fast sampling enabled 8.0kHz/1.0kHz
07/01/2022 12:36:36 : RCOut: DS150:1-4
07/01/2022 12:36:36 : OmnibusNanoV6 00570041 374D500F 2038394
07/01/2022 12:36:36 : ChibiOS: 45395b6a
07/01/2022 12:36:36 : ArduCopter V4.1.3 (2fb939a1)

Ok interesting, so this is down to the I2C1 definition. Let me try something else, can you try this one?:

arducopter.zip (624.9 KB)

This version seems to work, however it seems that radiomaster sees fewer sensors (I can only see those of the RX and the GPS before was more in the second version). the messages screen stops here as in the attached screen. In normal operation I usually saw that ekf could also use optical flow

07/01/2022 12:55:49 : Frame: QUAD/X
07/01/2022 12:55:49 : IMU0: fast sampling enabled 8.0kHz/1.0kHz
07/01/2022 12:55:49 : RCOut: DS150:1-4
07/01/2022 12:55:49 : OmnibusNanoV6 00570041 374D500F 2038394
07/01/2022 12:55:49 : ChibiOS: 45395b6a
07/01/2022 12:55:49 : ArduCopter V4.1.3 (2fb939a1)

Are you using ALT config 1?

No 0

image

Ok, I have no idea why this is happening but clearly the I2C1 definition on pins that are not in use is somehow interfering with the operation of USART1. It’s possible that this is some kind of memory/CPU/threading thing although there are no DMA conflicts/sharing that would affect this. It would be interesting to see the output of (in mavproxy):

ftp get @SYS/threads.txt -
ftp get @SYS/dma.txt - 
ftp get @SYS/memory.txt -
ftp get @SYS/uarts.txt -

For the working/non-working/semi-working version. Does the latest version I posted work well enough for you or is it still too broken?
Maybe @tridge has some ideas

I can try mavproxy if you explain to me how I can proceed. The version that on the bench looks similar to 4.0.5 is this. I also did a small indoor flight in stabilize, I have not tried other flight modes. Later could I attempt a little flight into the garden to see loiter, althold, flowhold?

Andy, I took a 12 minute flight with your latest version of the FW. I tried stabilize, althold, loiter, flow hold and also in auto with 2 waypoints. It performed perfectly, at least like the 4.0.5 I had tested a lot. quad 3 " in use 21 sat hdop 0.5. If needed I have the log.bin

So, just to be clear, this one?

yes Andy, exactly that :slight_smile:

Ok, so it looks like defining I2C1 as an ALT config gets us out of this so I will do a PR for that

2 Likes

Hi Andy, could you do a PR for the Omnibus nano V6.0 problem? Otherwise I can’t test the new realases and I’m stuck at 4.1.3.
Thank you very much!

There was a PR merged, I think it made it into 4.1 as well @rmackay9 ?

1 Like

Copter 4.1.5 Omnibus Nano F4 V6.0
it still doesn’t work, I can’t test it, I have to go back to 4.1.3 :frowning:

Sorry, does not look like it was backported

1 Like

@Alberto_Ds, @andyp1per,

Yes, unfortunately the PR’s project wasn’t set to “Copter 4.1” so I didn’t see it when I prepared the 4.1.5 release and this wasn’t caught during the 4.1.5-rc1 beta testing.

It will go into 4.1.6 but I can’t immediately promise when that will be released because we’re just about to start the 4.2 beta (it will certainly be included in 4.2 though).

1 Like

Thanks Randy, I’m going back to 4.1.3 and waiting for 4.1.6 / .2! :slight_smile:

1 Like