ACRO bug (fixed in 2.9.1b): while doing flips in ACRO mode, if you switch to Stabilize while inverted your throttle will go to minimum. To regain throttle control you need to switch back to ACRO then back to Stabilize again (i.e. switch to stabilize twice). You never lose control of roll/pitch/yaw.
Loiter/AltHold/Auto/RTL bug: if you switch into these modes with throttle at zero motors will go to minimum until you raise the throttle.
Auto mode altitude bug (fixed in 2.9.1b): setting a waypoint altitude greater than 320m over home altitude may wrap around and instead be interpreted as a low altitude.
ArduCopter 2.9 is now in the mission planner and the downloads area!
The major improvement is we use inertial navigation to improve altitude hold. This increased reliance on the accelerometers means you must do some additional set-up before flying:
3. If upgrading from 2.8.1, modify the throttle and altitude PID values:
Here is the list of major changes (a more detailed list can be found in the release notes):
As per usual PIDs are optimised for the 3DR/jDrones quad with 850 motors and 10" props. If you're using more powerful motors/props and are seeing bad flight behaviour in stabilize, start by turning down Rate Roll P in 25% steps.
Special thanks to our testing team lead Marco and the dedicated bunch on the 2.8.1 release thread who put their copters at risk while testing the pre-release version. Some of their videos are here: 1 2 3 4 5 6 7 8
Please feel free to report issues you find in the discussion below and/or add them to the issues list.
I gave that a try - the Mikrokopter FC had a similar acc reset command at "runtime". Didn't have any effect I could see in the planner flight data AH.
It would seem so. Unless I'm misunderstanding something it would appear that the calibration is done using the min and max values, assuming the range is linear between them.
I tried looking through the source code, but couldn't find the calibration routines, or where the offset and scale are applied to the raw values.
Should the compass_learn be on or off?
In general we recommend that it be left on. So COMPASS_LEARN should be set to 1. The only downside is the possibility that they'll get too big if you have a lot of motor interference.
When you did the accelerometer calibration you were of course careful to keep it level during the first step? Could you try it again? Also please make sure that you've upgraded your mission planner to 1.2.35 and there's no need to push the "Done" button at the end of the procedure (I know it's very unclear whether you're suppose to push it or not).
Autotrim (wiki page) or SaveTrim is another way to get the HUD level but the way it works is it uses the stick-input to move the HUD around...it does not actual use the attitude of the copter which is why when you tried it it appeared to have no effect.
The last and most brutally direct method is to update the AHRS_TRIM_X (roll) and AHRS_TRIM_Y (pitch) values directly through the Advanced Param List in the mission planner. The values are in radians. Like I said, this shouldn't be necessary though because the accel calibration routine should set those for you automatically based upon the copter's orientation at the first stage of the accel calibration routine.
By the way, the accelerometer calibration code is here. It uses the Gauss-Newton method which is described here but in short we use 6 points input to calculate the best fit to a sphere so it's a bit more sophisticated and produces better results than the min/max method.
You should definitely give the AutoTrim a try in a location where you don't have much wind.
Txs for the report. I've created an issue to capture this. If you have any additional details, could you add them to that issue?
Looking just at the RAW message I would have said that your accels are borderline acceptable. The INAV's message Baro Alt vs INAV Alt are tracking very well though so I guess you're ok!
As you found, the accel offsets are looking much better now. one of them is -170 but that's probably close enough to the -150 ~ 150 limit that you're fine.
If I didn't know it was windy I would have said that your Rate P values are a bit high but a windy day would explain why the copter's actual attitude diverges by up to 5 degrees from the desired attitude. I think this is fine too.
Same thing here Randy.
I use carpenter level at the first step to make sure copter strictly flat and still get same tilt left every time I calibrate. I found tweaking TRIM_X and TRIM_Y is the way to go.
What is it about "Done" button? We should not press it?
EDIT: found it ... do not press! That was a problem.
I don't think I posted this question, but I did have that problem once and I believe it was because I hit the done button again and I think that initiates another setup which of course I didn't complete because I thought I was done. The next time I didn't hit the done button and the hud display was level.
OK, so looking at the calibration code and the Rolfe's description of the algorithm, it's not just the end points, but there is still a problem with assuming linearity with these sensors, when so much hinges on knowing when you're level during flight.
I hope that the trimming system can be used to cancel out my 10 degree pitch offset, otherwise I'm never going to get good flight behaviour with this board.