Aircraft crashed after switching from RTH_CLIMB_TO_SAFE_ALT to RTH_HEAD_HOME.

[tonuonu]

Even I been learning and flying actual aircrafts, this is my first RC. Neither me not my aircraft (Volantex Ranger 2000) been flying yet. I apologize if this is wrong place to ask, please guide the beginner to correct place if needed.

FC is Speedybee f405 wing.
My settings
INAV_8.0.1_cli_20250712_114844.txt
My logfile
LOG00143.TXT

Video https://youtu.be/gBAo0BAG4BM?si=iL3rY9d7JwCliKpO&t=44

Mission file
crashed.txt
Mission screenshot

I am using INAV Blackbox Explorer and on timestamp 00:52.273, it has state RTH_CLIMB_TO_SAFE_ALT,
navPos[2] 41.88m and baro 44.04m.
On next timestamp 00:52.373 state is RTH_HEAD_HOME and height is increasing up to 56m until aircraft crashed to ground sometime after 00:56.473.

Probably I been doing something wrong, but as a beginner I am unable to figure out.

[sensei-hacker]

I'm on my phone, so can't look at the log right now, but here's a start.

This screenshot shows the pitch while it's climbing through 25 meters:

The camera may be pointed a bit down, but it's climbing at this point, so it has some pitch up.

Then just before it fell, at 47 meters:

Pitched up quite a bit more, then fell out. So classic stall. WHY it stalled needs further analysis.

[tonuonu]

I agree with your arguments. Because this aircraft has never flown before, can it be something in hardware or settings which are simply wrong? And it is something what does not affect during autolaunch but screws everything up during RTH. What it could be? For me there is no apparent logic. To me looks like if aircraft is fundamentally unstable, it would exhibit already earlier.

[sensei-hacker]

Even I been learning and flying actual aircrafts

If "actual aircrafts" means full-sized aircraft, there's a lesson here that may save your life one day. In the video and in your description, I don't see any attempt to fly the aircraft after the trouble started.

Things will happen in the air. It's crucial to "keep flying the plane until after the crash". In this case, that would mean flip to Manual mode (or even angle) and - fly the plane.

[stronnag]

Expecting the plane to climb c. 40m in the distance to home is probably not reasonable (yellow is RTH phase).

RTH invoked at 47s.

Had the experiment been performed 500m from home, or at 45m elevation, the outcome might have been different.

[tonuonu]

I would be happy if you can extend your answer a bit. My only expectation was that iNav is smart enough to bring my aircraft up and then safely return home without me knowing how to operate sticks. Maybe my expectations are wrong but I have not found very little instructions in internet about this phase of flight for beginner. Tutorials tend to skip this part, it simply seems to work as magic, so I expected same. If there is a mistake in my actions, I would like to know. For example, turning on WP mode too early? From your comments I read out that I first need to satisfy preconditions for such mode?

[Jetrell]

@tonuonu It appears you weren't in RTH, but in WP mode.
There was one invalid WP. So as soon as you activated the WP mode,, it attempted a WRTH.

But the log data and DVR don't seem to line-up. The log cuts off before the plane hits the ground, based on the navPos[2], navState and flightModeFlags, when comparing to the DVR footage. And there was no acc spike indicating a collision before the FC power was interrrupted.

[tonuonu]

I now updated post by adding mission file and screenshot of mission. Unsure how to validate it. By naked eye it seemed OK and xmllint was not complaining either. Whole thing why it decided to RTH is a mystery to me. Hard to believe I hit some bug in such mature software. But if this is a case and my FC is useful to developers, I can ship it instantly for debugging help. My long experience is tech have taught me that most of "bugs" of new users are user bugs.

[Jetrell]

Did you load the mission from the eeprom to the FC before the flight, with a Stick command. ? If it wasn't loaded. WRTH will commence when you switch to WP mode, because no mission is available.
The other way to simply load a mission is by having nav_wp_load_on_boot = ON.
It would appear that you where with-in nav_wp_max_safe_distance = 500.

[sensei-hacker]

If I'm correctly understanding the chart Stronnag posted, it looks like it climbed at an much faster rate than the default nav_fw_auto_climb_rate of 500cm/s.

Unless Stronnag or someone sees something I'm missing, that might be an issue for the developers to resolve.

Until we do, you would want to avoid triggering RTH while at low altitude close to home, because it can't safely climb to 50 meters that fast. You might also reduce the RTH altitude IF the terrain around that area allows it. Perhaps not if there are tall trees or buildings around.

[stronnag]

There was no WP mission loaded, the log shows that:

H waypoints:0,0

The vehicle was in RTH (via WRTH).

[breadoven]

If this really was the first flight of this plane then I think the main problem here was not setting it up so that it flies properly in Manual and possibly Acro modes before trying anything else such as Nav modes. This means it should fly straight and level at "Cruise" throttle with pitch/roll centred. Trying any of the Nav modes before this has been done is just asking for trouble.

As @sensei-hacker mentioned though, the main reason it crashed was because it stalled with pitch angles during the climb > 45 degs. It probably only got as high as it did because the throttle was on max from shortly after launch until the log stopped. This was caused by nav_fw_allow_manual_thr_increase being set ON and the throttle stick being on max throttle throughout. Probably better to turn this setting OFF until the Nav modes are working OK. I assume you checked the CG and control neutral positions before flight ?

It also entered Failsafe for some reason just before impact although this was missed by the log. Any idea why ?

You might want to read https://github.com/iNavFlight/inav/blob/master/docs/INAV_Wing_Tuning_Masterclass.pdf.

[tonuonu]

Well, I think it is time to thank everybody. Part of my confusion came from my mistake. I thought I had "Load Waypoints on Boot" enabled. Now thanks to input from @stronnag, @sensei-hacker, @Jetrell, @breadoven and some code reading I think I understand events which lead to crash and can try again.

1. I assumed I had waypoints loaded but they were not.
2. I enabled WP mode which actually ended up with RTH executed.
3. Because HOME was right behind, it attempted to do something which was not the best idea.

I think control planes were on neutral and CG was ok. To understand every other detail explained here I need to fly more and then think of whole thing again. Only thing which confuses me - AFAIK winged airplanes are most unstable at low speeds next to ground. When aeroplane is not flyable enough, it should crash in first seconds during launch, not having serious problems later. I will read pointed PDF, maybe it answers it. Thank You everybody!

[yellowfin1976]

I have the same plane with the same speedy bee 405, same inav version
video of my ranger: https://www.youtube.com/watch?v=AqpAg1NkcVc

now the first thing is you need to check how you have configured your autolaunch , I have mine in aprox 75% throtle and decent climb angle until it reach 200ft or so ... depending on the wind conditions I used flaps or not, auto launch works very good, do NOT TOUCH IT unless completly necesary, every launch I make is visual, what I mean by that is that my googles are on the floor waiting, after every auto launch I put loiter mode to get everthing in order before the the actual flight. RTH you can make a climb and loiter around home.. the only error i see on your video is you touching a lot of stuff very quickly ... does it survided the crash?