INAV 7 - No dive / stick reaction to descent in cruise mode

[jaroszmm]

---

Current Behavior

I cannot say it happens on all my INAV planes but it does happen more than sometimes. Basically in Cruise mode plane sometimes won't dive if I push the elevator stick forward. It will keep flying straight. Climbs are no problem. The only solution for me is to go to Angle/acro/manual and dive manually.

I also noticed it seems to be somehow connected with climb time. Recently I climbed in circles in cruise mode for a couple of minutes and after that I was unable to dive. BUT if I climb for a short time then after leveling out I need to wait a short while and diving is possible again. Feels like a long climb in cruise "locks" diving or something.

Steps to Reproduce

1. Go to CRUISE mode and start climbing with full stick deflection probably for a longer time than a couple of seconds (a minute or more)
2. Level out.
3. Try to dive.
4. If not possible switch to Angle, then go back to cruise and try diving again.

Expected behavior

Dive instantly with stick command.

Additional context

It would seem that there is something with the altitude/attitude controller as sometimes I also notice that climbing in Cruise mode is not instant. I pull the stick and plane takes a second or two to actually increase motor power and pitch up.

---

version

INAV/SPEEDYBEEF405WING 7.1.1 May 6 2024 / 12:49:15 (dd91a87)

GCC-10.3.1 20210824 (release)

[sensei-hacker]

Do you have a blackbox log?
A diff?
Do you have continuously trim servos on?

[jaroszmm]

I can have a diff...

https://pastebin.com/HLhDrMvX

Also, the board has changed, I looked up my diff on my PC but it was before switching the FC so the current "version" stuff is:

version

INAV/MATEKH743 7.1.2 Jun 1 2024 / 00:04:49 (4e1e59e)

GCC-10.3.1 20210824 (release)

and yes, I have continuous autotrim enabled
as for the log I can put SD card next time and give it a try.

[P-I-Engineer]

try lowering your min throttle. i had this issue on my dolphin.
i run mine at 1025.

set nav_fw_min_thr = 1100

[jaroszmm]

Hopefully the log has recorded. Did a short flight today. Autolaunch, a short while on angle and then cruise. Climb was normal, levelled out, pushed the stick, plane started diving. But later after a longer while and another cruise climb plane would not dive anymore at first. Tried even switching to angle and back to cruise and nothing. Climbing was still okay.
Also tried RTH and linear descent worked. Only later, no idea why, cruise descent started working but with a significant delay. So I push stick, plane needs a second or two or more to actually drop the nose.

[sensei-hacker]

If you have any difficulty with INAV 9 please let us know.
About 100 bugs or "idiosyncrasies" have been fixed since INAV 7.

[b14ckyy]

I checked his log and I can confirm the issue being visible there.

Here at the earlier part of his flight the raw stick back or forward results in the vertical speed react accordingly as we would expect with the INAV 7 altitude control.

But then later in the flight after a 50s long climb, the next push forward did not react and the altitude was locked in place.

After that happened, even after multiple mode changes between angle and cruise mode, the plane was only able to climb in cruise but refused to descent.

I wonder if it can be replicated just by doing long (1min or longer) full stick climbs in cruise. or it depends on the total altitude delta I am not sure.

---

This same issue was reported in the past multiple times, also on Facebook but this is the first time that we have it in a log. Thanks a lot @jaroszmm for that!

[jaroszmm]

I wonder if maybe I should now put INAV 9 RC3 onto my plane and try doing the same. If it doesn't happen then we can assume it got fixed. If it is still there then we have a bug.
Since in RTH mode plane went into linear descent with no problems I wonder if this is limited only to cruise/althold modes and autonomous modes are free of that bug.
However during two RTH tries I noticed something. During first one it first directed the nose towards base and then OSD said "begin linear descent" at 1k out from home and plane started diving properly and instantly. I then disabled RTH but when reenabled it again after a short while plane would not say on OSD about the descent and only after a short while it would dive.

[breadoven]

What is navTgtVel[2] doing in the log when the plane fails to dive when descent is commanded in Cruise ? Stay at 0 or show a -ve vertical rate ?

[b14ckyy]

navTgtVel[2] is always 0 in this log no matter what happens. I think thats normal in 7.x?

[jaroszmm]

I'm fine if you guys share the log among yourselves, will probably be much easier to work with it that way. I just didn't want it to be public available, that's all.

[breadoven]

Ah 7.x ... had issues with altitude RC adjustment related to using a rolling altitude position target. One of the reasons for the changes in 8.0.

[jaroszmm]

Okay, a short flight on 9 RC3 and I had no issues. Plane would dive and climb basically instantly when giving stick input. I just went with diff all from 7 to 9 using Darren's tool to convert some stuff and the rest was as RC3 came.
Also plane felt "smoother" in general flight compared to 7? Or was it more just my imagination?

[b14ckyy]

if you mean with smoother the altitude control then yes thats plausible with the better Sensor fusion and faster altitude calculation. Less overshooting and corrections.

[jaroszmm]

This plus it also felt like roll was softer, not agressive etc. But this could be a placebo effect.

[b14ckyy]

That was probably the I-Term Rate Attenuation that Pawel and I developed for 8.0.
Glad you noticed :) So its definitely worth it :D