3.4rc1

Navio / Edge Flight stack
21

Have a look at the “extended tuning” screen.
Autotune should still be available in the CH7 and CH8 dropdown list. I do not know why it appeared in the primary flighmodes selection before, because you need AltHold together with Autotune.

22

Compass offsets in beta for compass 1 and 2 were always in the green for me, EKF never had an issue. With the build 3.4rc1 and the “sudo apt-get install apm-navio2” method both give my unreliable compass. With the built arducopter i noticed that iff i kept roll to 180 and calibrated and then used mission plannet to view the mx,my,mz and mx2,my2,mz2 all lined up except on of the axis (cant remember but when i changed comass 2 to no orientation change all compass axis were right about the same until i restarted ardupilot/raspberry.

1 Like
23

So do the developers have any input as to unreliable compass issues?

24

Good question

1 Like
25

Bump

1 Like
26

Same questuon here. I have been struggling with the compass issue for quite some time now.

1 Like
27

Anything on this George? Thank you

28

I loaded RC3.4rc1 on my Navio+

The good news is that the Compass problems are gone but it now has a Pream Check Board Voltage error. Disabling this in QGroundControl does not seem to work and I can’t find the parameter in APM Planner.

Also when connecting with QGroundControl on and Android device is warns that the firmware has not been compiled with JOYSTICK capability…

This is becoming a bit of a struggle …

RC2??

29

I am trying to figure out the compass issue and trying different configurations, hopefully it will help provide some positive information on a fix. If I use compass one only, all is fine except ekf will give errors when the flight controller is moved very rapidly and completely around the all axis… prob would never see that kind of rotation unless totally out of control.

If I only use compass two, it never seems to get a heading and stay locked, if I rotate it and set it down the compass heading takes about 15 seconds to settle and it does not point north, it is off about 90 degrees once it settles.

I have tried this many times, recalibrating, resetting, and using only a cat5 for network and a 10 USB cord (trying to eliminate any interference). Kinda stumped but it seems like there is still a orientation set for compass 2 or ardupilot needs different ekf threshold parameters (only say this because of the information on the docs page that says compass 2 has better precision?)

30

When I upgraded firmware, my secondary was off by 90 too. I tried every possible rotation and it still would not point correctly, and mission planner would give me a compass variance of course lol.

Thanks for trying, believe the team stone walled us today lol!

1 Like
31

@ADSB2016 I tried all sorts of rotations too, using mission planner to check the mag readings. I haven’t looked at the code, not that I know what most of it means but I can kind get the picture of whats going on, I remember a comment that said it would force user to specify a roll of 180 degrees but it might also be driver related and that is waaaay out of my knowledge.

I do not think they stone walled us today at all, I hope that silence means that George and Igor are busy working towards a solution. I have had some issues with my Navio2, hardware/software and both of them and other people at Emlid (never got their name in the emails) have been very helpful. I found an RC bug and the next day i was given instructions on how to fix it. Ardupilot has taken years to get at its current stage, Emlid is still new on the block and have made great leaps!

To George and Igor, my Navio2 is on the bench for more testing, If there is anything I can do to help let me know! I might learn something at the same time!

1 Like
32

2 posts were merged into an existing topic: Compass variance 3.4-rc1 on Navio 2

33

I must say that I applaud your optimism. I myself have shared in the frustrations expressed in the threads of recent days; the greatest of these frustrations has been that a week or so ago I was a leaf on the wind, but now I’m a brick on the ground :wink: I have no doubt that these recent issues will be sorted in short order given that the software, in it’s current state, is unusable. Like most though, I too would appreciate an acknowledgement from the developers that there is in fact an issue that has grounded our craft, and assurances that the issues are being dealt with.

1 Like
34

From my experience, I don’t believe that there are any hardware problems or emf interference which cause the compass or mag errors. I believe that it is a bug or two in the firmware.

I am happy to assist to resolve this by testing Navio+ if required.

in addition to the compass and mag errors please note the observations about

  1. joystick support.
  2. board voltage

Let me know if I can help.

Gary

36

Please don’t forget support for us Navio+ users :slight_smile:

2 Likes
37

Thanks guys

38

Check out my post in another thread:

39

Interesting. I have reverted back to 3.4 unstable and am still having serious issues. I’ll try this next. I wonder if i need to disable the on-board Reach’s wifi as well…

AM

40

Prior to this most recent update, when everything was flying just fine, I was operating with dual 2.4ghz wifi(RPi3 internal w/usb dongle) without any issues at all. Wifi in the field seems to be a must have so as to be able to remotely shutdown the RPi prior to disconnecting the battery to prevent SD card corruption.

41

I do mine over 4g with a dongle non wireless. I’m only sure of one thing, she didn’t fly right before I killed it and now it does lol