search

iNavFlight / inav

INAV: Navigation-enabled flight control software
https://inavflight.github.io
GNU General Public License v3.0
3.19k stars 1.49k forks source link

Inav 8 artificial horizon not working #10249

Open P-I-Engineer opened 4 months ago

P-I-Engineer commented 4 months ago

Current Behavior

artificial horizon drift

Steps to Reproduce

fly, turn a bit, fly level, horizon is off.

Expected behavior

horizon not lag or drift

Suggested solution(s)

Additional context

version

INAV/AOCODARCF405AIO 8.0.0 Jul 17 2024 / 07:14:26 (4fa40ad0) dev GCC-13.2.1 20231009

PXL_20240720_154747143 inav_003.zip INAV_8.0.0_cli_AOS_20240719_202428.txt

P-I-Engineer commented 4 months ago

screen cap from my plane as well.

speedybeef405wing

PXL_20240720_154634244 INAV_8.0.0_cli_PENGUIN_20240719_075138.txt

P-I-Engineer commented 3 months ago

new log with all things enabled in bblog

inav_001.zip

breadoven commented 3 months ago

I didn't notice any problems the other day using a version 8 that's perhaps 3 weeks behind the master. Didn't have the AHI on the OSD but did have roll and pitch angles which behaved as expected with no offset to the horizon. I assume you're using the current master version 8.0 ?

P-I-Engineer commented 3 months ago

yes, current as of a week ago

b14ckyy commented 3 months ago

Can you check what ahrs_inertia_comp_method is set to? It must be set to Adaptive in Fixed wings. For some reason it seems to be set wrong after Update with DIFF.

P-I-Engineer commented 3 months ago

yes, set to adaptive

P-I-Engineer commented 3 months ago

flew quad again, with roll angle on osd, roll is lagging behind.

P-I-Engineer commented 3 months ago

vid diff included aos 3.5

https://www.youtube.com/watch?v=eRhMadvbffE

INAV_8.0.0_cli_AOS_20240801_090457.txt

tipoman9 commented 3 months ago

I experience similar problem, here is a video of it. https://www.youtube.com/watch?v=wqXGWZxgy7s INAV/SPEEDYBEEF405WING 8.0.0 Jun 28 2024 / 00:18:11 (60a4b714) AR Wing PRO, 1300g AUW, CG on the wing marks. Perfectly balanced prop. I use magnetometer. Looking at the logs, its values do NOT look consistent. This is the moment the plane was on a landing approach. I'm pulling the elevator at max, while the servos are kept at mid-point.... image

Here are the magnetometer data related to Gyro Pitch. I can't find a magnetometer axis that corresponds to the pitch. image

What else in the blackbox data can be examined?

b14ckyy commented 3 months ago

Tagging @shota3527 as the matter expert here. Maybe you have an idea? ahrs_inertia_comp_method = ADAPTIVE should be fine now for both MR and Wings as it switches based on platform regarding to https://github.com/iNavFlight/inav/pull/9360

b14ckyy commented 3 months ago

Update from my end. I recently discovered that my Altus, I used for mLRS tests a lot the last 2 weeks and flew mostly autonomous most of the time and barely looked at the FPV feed, also had strong AHI drift. The one I thought that's not affected.

I found that accidently the Compass got reactivated during a switch back to 7.1 and then back to 8.0. It was calibrated a long time ago but its literally sitting on top of the battery wire so its very useless.

I disabled the compass for the last flights and the drift was nearly gone. I had one small occasion where there was a bit of a tilt after a hard left bank out of a right turn loiter but that might have been caused by a string wind gust from the side at the same time.

So to me it seems like enabled compass exaggerates the drift. All the settings are unchanged and 8.0 default or the same values as they where in the 7.1 presets.

breadoven commented 3 months ago

But how long does the drift last for ? I'm guessing it corrects itself relatively quickly compared to the past problems with AHI drift.

sensei-hacker commented 3 months ago

I found that accidently the Compass got reactivated during a switch back to 7.1 and then back to 8.0. It was calibrated a long time ago but its literally sitting on top of the battery wire so its very useless

So a very wrong compass that's giving false indications of a turn seems to make it worse?

b14ckyy commented 3 months ago

@breadoven about 5 seconds until it recovered. But in that time during some test WP missions the plane sometimes already lost 10-15m of altitude after a turn before it recovered and went back to correct alt. Never had that issue on 6.x and 7.x it was more stable.

@sensei-hacker It might exaggerate it. Unfortunately the conditions where pretty different. Previous test that was so bad I had 10-25kph constant gusty wind. This time with compass off it was calm with just occasional 20kph wind bursts from thermal currents passing over. But still saw some drift now and then.

I will check the old DVRs from that plane's maiden a few months ago as that was on 7.1.2 at similar gusty wind conditions.