Hi everyone,
I've been flying with a Pixhawk for a little while and probably getting away with the minimum amount of knowledge required to make it work.#
After a recent crash (which involved the Pixhawk) due to a lack of elevator authority (see the DIY drones discussion here: http://diydrones.com/group/pixhawk/forum/topics/could-you-please-help-analyse-this-pixhawk-crash-with-log?xg_source=activity) I decided that I wanted to make sure everything was a'ok with the Pixhawk prior to flying again.
I've recently completed a new plane build and am now installing the Pixhawk. Unfortunately a few error messages have come up and I was wondering if there is a problem with the Pixhawk itself (perhaps due to the crash or just that it's broken) or with my set up.
For background I have gone through the Wizard set up in Mission Planner but still have the below issues unfortunately. If you are able to help then that would be amazing! :)
The issues are:
1. Compass issues - error compass variance and bad compass health - the error compass variance seems to come up randomly and generally when I'm moving the board around. The bad compass health seems to also come up randomly although I'm not sure if it's ever present as when I take the board outside to acquire a GPS signal the bad compass health message seems to stick around.
For background I have the Pixhawk pointing to 270 degrees clockwise to the direction of movement and the compass pointing in the direction of movement and therefore 90 degrees clockwise relative to the Pixhawk.
I've therefore set AHRS_ORIENTATION to 6 (Yaw270) and COMPASS_ORIENT to 2 (Yaw90).
I've gone through the compass calibration and it seems to be accepting this.
2. Bad gyro health - this has only come up once and I read here (https://3drobotics.zendesk.com/hc/en-us/articles/201771859-What-to-do-if-I-m-getting-a-Bad-Gyro-Health-warning-on-my-Ground-Station-) that it could be due to a blown 3.3V regulator on the board. I have tested the outer pins on the I2C connector and am getting 4.90V when it seems that this should be reading 3.3V?
Could this also be affecting the compass as it is plugged into the I2C connector?
3. Velocity variance - this is also only occasional and I notice that when I click on the EKF button in Mission Planner that usually the velocity bar is either 0 or very close to it. However, at certain times this does jump up to above 0.8 when it causes a failsafe failure on this parameter.
Thanks for taking the time to read through this and any thoughts you have on any of the issues would be much appreciated.
Cheers, Tom
Replies
Hi Harry and Evan,
Thank you for the messages and your thoughts on the issues. In relation to the issues listed above:
1. Compass errors - The compass was set up relative (its attitude) to the Pixhawk and not the air frame. After speaking with 3DR support (and contrary to what I have read on another forum) it turns out that this is incorrect and it needs to be set up relative to the air frame. Once this was sorted I haven't seen any further compass errors.
2. Bad Gyro Health - I think that the issue was that the Pixhawk wasn't stable straight after powering up and the gyros couldn't calibrate properly as a result. It is also worth noting at this point that 3DR support informed me that the Pixhawk doesn't have the same 3.3V regulator as an APM 2.x.
3. Velocity variance - I believe that this is due to a poor 3D GPS lock and there is subsequently a discrepancy between the internal sensors (accel/gyro/barometer) and what the GPS is telling the Pixhawk is going on. Although not tested, this should be fixed by having a strong GPS signal outside somewhere open.
Thanks for all of the help!
Tom
Tom,
From what I have seen, the compass errors and velocity variance errors you are receiving could be a result of not having a GPS lock. I also had these issues when first starting, and that was the issue. I had one GPS module giving these errors and thought that it was a bad GPS. The new one I purchased also read these messages, so the issues you are seeing may not be a result of a bad Pixhawk or GPS module. Try some of these things to see if you can get it working.
If you haven't already, take it to an open field and start your module up. Make sure you keep everything still for the first 12 seconds while everything is powering on and the Pixhawk is running the initial gyro calibration. The "Bad Gyro Health" is usually attributed to improper initialization of the flight controller, so this should take care of that issue. Make sure you have a GPS lock and your Pixhawk LED should change from yellow to green. Try arming the Pixhawk, and if everything is working, you could try flying if you want.
If you don't get a GPS lock while in the open field, you may not be getting enough satellites due to your location. I live in a rural area, so my module wasn't getting enough satellites. You can fix this by going into Mission Planner and changing the pre-arm safety checks. To do this, go to the Config/Tuning tab. Under it click on the standard parameters and go to the pre-arm safety checks. Here you can edit which checks the Pixhawk performs. I would first bypass the satellites check, and see if this works. Here you can also check and uncheck different pre-arm checks, helping to troubleshoot other issues you may have.
Here is a link giving you a more in-depth explanation of what I just explained: http://ardupilot.org/copter/docs/prearm_safety_check.html?highlight...
After doing this, I personally went again into the GPS configuration, and recalibrated it. I would uncheck the box that times the calibration, allowing you to get as many points as you want. This gives you a more thorough configuration. Let me know if this works! If not let us know, and perhaps take some pictures of your setup and a video of your issues and upload. This could help others see what is wrong.
Cheers
Evan
Tom,
I use Pixhawk for a multirotor, not a plane, however the messages you quote have appeared for me also.
In my case these types of messages show up PRIOR to gpx fix and arming. Once armed the messages settle down and go away.