|
MPU9250 heading is slowly drifting
|
disabling opengl is only for the non rendered boat issue.
to resize the boat window with blue background, put the cursor in the bottom edge of that blue window, wait for the cursor to change shape (2 arrows) and drag, see attached 
2020-01-22, 07:26 PM
Thank you Sailoog. Sorted now, there seems to be a delay before the resize window appears, wasn’t patient enough!
(2020-01-21, 02:06 AM)Opie91 Wrote: I was able to get the acceleration to calibrate and then set the alignment, but not the compass would not calibrate. The heading is slowly drifting about .5 deg/minute. Not completely convinced that my problem is not a bad batch of MPUs. MPU change made all the difference in the world. Accel & compass calibrations happen easily. Heading rock solid. Thanks!
2020-01-24, 04:40 PM
I have a Pi3 with a 9250/9255 at the end of a 1m piece unshielded wire mounted in a block of wood sittin in the middle of the basement on a wooden stool where roll is slowly drifting. I calibrated the compass by rotating the block on each of the 6 sides. When eventually I set it down so that it read 359 and then flip the cube around 180 degrees to face south it comes up with the wrong direction. I have a old release, I'll try this with 2.0.22-stable and try this again, maybe my IMU is toast too? Maybe 1m is too long for a unshielded wire? What are others using? I need at least 1m of wire to get away from the other electronics when I install it in the boat.
2020-01-24, 07:05 PM
1M is pretty far for i2c but should still work. You probably don't need to get it away from electronics as much as you think.
Rotating the sensors on 6 sides does not calibrate the compass. This calibrates the accelerometers. To calibrate the compass you need to slowly (like speed of a boat turning) rotate in 360, but only after you have leveled the sensors as well. It helps to run the "pypilot_calibration" script to visualize this.
2020-01-24, 07:14 PM
(2020-01-24, 07:05 PM)seandepagnier Wrote: 1M is pretty far for i2c but should still work. You probably don't need to get it away from electronics as much as you think. And you need the absolute last version of pypilot, 2.0.5-beta!
2020-01-24, 09:38 PM
I guess that makes sense the Moitessier HAT sits on top of the PI. I'll shorten the wire. I'll put the PI and the IMU in a wood box and position it on the 6 sides to set the accelerometers, then set it flat on the floor mark boat is level and then slowly rotate it. If north is not 0 and south is not 180 I'll get a new IMU.
2020-01-24, 10:41 PM
(2020-01-24, 07:05 PM)seandepagnier Wrote: 1M is pretty far for i2c but should still work. You probably don't need to get it away from electronics as much as you think. Pls have a look at this post, might help with i2c extension https://sites.google.com/site/olewsaa/ya...e-extender Bart
2020-01-25, 02:48 AM
I've had my IMU going out of wack after re-installs and sitting motionless.
It just needed re-calibration as per https://openplotter.readthedocs.io/en/la...ation.html Thanks, --MG
Download BBN Marine OS for raspberry pi
https://bareboat-necessities.github.io/m...at-os.html Video of actual installation: https://www.youtube.com/watch?v=3zMjUs2X3qU
2020-01-25, 09:11 PM
Fresh 2020-01-13-OpenPlotter-Starting-v2.0.img installed and updated with pypilot 2.0.5-beta, the boat graphic is working!
Running thru the accel calibration process. ( thx @mgrouch I've been looking at the 1.x docs ) 1) after the 6 side calibration and the set of green dots, they never turn yellow 2) adding more point to sphere does not seem to help it 2) the calibration age never updates 3) there are time when pypilot_calibrate process runs at 99-100% and the graphic stops updating Calibration is not locked and the Clear button seems to have no effect IMU wire is about 30cm stopped pypilot processes, removed .pypilot/pypilot.conf and restarted no change in calibration age rebooted and the calibration age remains the same ( maybe I should take this discussion to another thread? ) |
|
« Next Oldest | Next Newest »
|
