VTOL Crash after FBWA transition. Could not save with qhover

[Mihai_Comsa]

Hello,

The quadplane in this topic is equiped with CUAV x7 PRO controller, using CUAV NEO 2 PRO GPS. I managed to make it hover quite stable in Qhover and Q loiter mode. Today i tried making the transition to Fixed wing (Qhover - FBWA). The front motor started spinning and the quad motors kept thrust until min airspeed ws reached. After that i got the message that transition airspeed was reached and the plane tilted up and i lost control. I tried switching back to Qhover but it started to spin out of control and eventualy crashed.

I have no idea where to begin to try and figure out what happened. I hope there is someone that can guide me in the right direction.

I have attached the log files (telemetry and dataflash).
https://drive.google.com/file/d/1__wsaK2eMRidGJSeGdPmiyrpvf8M-v2n/view?usp=sharing

HW Specs:
Frame - MFE Fighter VTOL 2430
12 S 12000 mah battery
Matek digital airspeed sensor

Thank you,

Mihai

EDIT>

attached the link with dataflash log and video of the crash

https://drive.google.com/drive/folders/1T6uC1Uu9Q9jYbsCDHAP7i5n8utM2Al0j?usp=sharing

[kd0aij (Mark Whitehorn)]

Where is the Center of Gravity?

[Mihai_Comsa]

It is set up with a slight forward tilt. Attached are some images

[Mihai_Comsa]

Forgot to mention that all the parts were fixed inside the fuselage and the batteries were placed in a custom 3d printed case that was permanently glued to the frame.

AUW is 10kg.

Frame is rated for 11.5 kg max

[kd0aij (Mark Whitehorn)]

Check out this thread: Fighter VTOL 4+1 fixed wing

[GregCovey]

Mihai,

The QHOVER at the beginning of the flight looked good. When first entering FBWA mode, you could see several issues. Did you use the MFE Fighter 4+1 VTOL PID values for both hover and forward flight modes?

After loss of control in FBWA mode, it then switched to QRTL mode before changing back to QHOVER mode. This may have kept the plane from obtaining a proper hover. Whenever QRTL mode is entered, bad things seem to happen when you try to switch out of this mode fast. Never use QRTL (intentionally) on a maiden test flight.

Sorry for your crash. I hope that you can salvage most of the pieces.

[Rolf]

@Mihai_Comsa

Did you check in FBWA mode , that the elevator is deflecting in the right direction? ( Pitch plane up: Elevator moves down )

[GregCovey]

Rolf,

Thanks for he reminder. I always mean to mention it but seem to forget.

I don’t understand the difference in the PIDs from Mihai’s setup and mine. Although the YAW2SRV PIDs look the same, the RLL and PTCH introduce new parameter names that do not match the original file.

Mihai’s PID Values
RLL_RATE_D,0
RLL_RATE_FF,0.345
RLL_RATE_FLTD,12
RLL_RATE_FLTE,0
RLL_RATE_FLTT,3
RLL_RATE_I,0.15
RLL_RATE_IMAX,0.666
RLL_RATE_P,0.08
RLL_RATE_SMAX,150
RLL2SRV_RMAX,75
RLL2SRV_TCONST,0.45

PTCH_RATE_D,0
PTCH_RATE_FF,0.345
PTCH_RATE_FLTD,12
PTCH_RATE_FLTE,0
PTCH_RATE_FLTT,3
PTCH_RATE_I,0.15
PTCH_RATE_IMAX,0.666
PTCH_RATE_P,0.04
PTCH_RATE_SMAX,150
PTCH2SRV_RLL,1.1
PTCH2SRV_RMAX_DN,75
PTCH2SRV_RMAX_UP,75
PTCH2SRV_TCONST,0.45

Greg’s MFE Values
RLL2SRV_D,0.1
RLL2SRV_FF,0
RLL2SRV_I,0.2311478
RLL2SRV_IMAX,3000
RLL2SRV_P,1
RLL2SRV_RMAX,75
RLL2SRV_TCONST,0.45

PTCH2SRV_D,0.1
PTCH2SRV_FF,0
PTCH2SRV_I,0.13
PTCH2SRV_IMAX,3000
PTCH2SRV_P,1.57
PTCH2SRV_RLL,1.1
PTCH2SRV_RMAX_DN,75
PTCH2SRV_RMAX_UP,75
PTCH2SRV_TCONST,0.45

[Rolf]

The differences are due to the new PID system:

New fixed wing PID system and autotune

[ArduPilotPlanePorter] Last month we finally merged the new fixed wing control code slated for the ArduPilot plane 4.1 release. This provides a major improvement in several areas: much better manual PID tuning, with clearer parameters and much improved filtering new fixed wing AUTOTUNE system which produces a much better tune for a wider range of aircraft The key change in the PIDs is we are switching to use the same PID code that is used for VTOL aircraft in ArduPilot. This leverages the g…

[Mihai_Comsa]

Hello,

Sadly the frame is obsolete. I already ordered a new one (not giving up that easily )
Indeed, i switched to QRTL when I noticed that things got messy but realized the VTOL was not going to recover. Then i switched to QHover hoping the less assisted flight mode would save the day.

Still looking into the log files hoping something would come up to keep me from doing the same mistake with the new frame. So far nothing, but the learning curve is steep and I have a long way to go before I understand most of the graphs.

Any insights would be highly appreciated.

Thank you,

All the best!

Mihai

[Mihai_Comsa]

Hello,

Yes, the elevator was moving accordingly.
However I did not see any rudder corrections when testing servo deflecting on the ground using FBWA.

Thank you,

Mihai

[GregCovey]

Mihai,

My recommendation would be to use a previous release firmware that uses the older PID system. In this manner, you can use the proven MFE PIDs as your starting point. I never needed to change mine.

Alternatively, you can ask or wait for MFE to update their .param file to the new PID system. In the numbers below, your P values are way off which is why the plane did not fly properly.

Mihai’s PID Values
RLL_RATE_D,0 (should be 0.1)
RLL_RATE_FF,0.345 (should be 0.0)
RLL_RATE_FLTD,12
RLL_RATE_FLTE,0
RLL_RATE_FLTT,3
RLL_RATE_I,0.15 (should be 0.23115)
RLL_RATE_IMAX,0.666 (should be 3000)
RLL_RATE_P,0.08 (should be 1.0)
RLL_RATE_SMAX,150
RLL2SRV_RMAX,75
RLL2SRV_TCONST,0.45

PTCH_RATE_D,0 (should be 0.1)
PTCH_RATE_FF,0.345 (should be 0.0)
PTCH_RATE_FLTD,12
PTCH_RATE_FLTE,0
PTCH_RATE_FLTT,3
PTCH_RATE_I,0.15
PTCH_RATE_IMAX,0.666 (should be 3000)
PTCH_RATE_P,0.04 (should be 1.57)
PTCH_RATE_SMAX,150
PTCH2SRV_RLL,1.1
PTCH2SRV_RMAX_DN,75
PTCH2SRV_RMAX_UP,75
PTCH2SRV_TCONST,0.45

[Rolf]

Mihai,

I’m still not sure that you’re not accidentally moving the elevators in the wrong direction. If you have flown successfully with identical parameters before, then forget my objection, it must be something else.

Otherwise, look at the reaction of the aircraft to the elevator deflections:

As airspeed increases, pitch increases almost in parallel with elevator deflection. This increases until the A mark in the graph. After that, the elevators return almost to the neutral position (marker B) and the nose of the aircraft also goes down at the same time.

This speaks very much for a wrong direction of the elevators. However, the fact that the aircraft did not immediately make a loop, as would have been typical, speaks against this. However, a loop did not occur simply because the quad engines were still running and steering against it. The transition speed was reached long ago, but the transition and the running quad motors were not switched off, because Q_TRANSITION_MS is set to the default value of 5000 µs. (Note on the side: I have always set this value back to 0, because the sense is not clear to me until today).

Rolf

[Sunny52266]

I have seen a PID converter by tridge somewhere on this forum around the launch of 4.1. maybe that can be found and consulted.

[Rolf]

https://uav.tridgell.net/convert_pid.html

[Mihai_Comsa]

Rolf,

Thank you for checking the logs.

This was my first attempt to transition from QHOVER to FBWA. Although I have triple checked deflections before maiden flight, maybe I missed something. That being said, the tail miraculously survived the crash so i wired the flight controller with the settings i had prior the flight and i recorded a video with the elevator behaviour. I uploaded it to the link containing the logs, if you have a moment to check it out. (sorry for the vertical framing, i noticed it too late)

It is worth mentioning that I did not use a Y splitter for servo connections. I wired them to separate channels. Here is a print-screen with the servo output.

All the best,

Mihai

[Mihai_Comsa]

Greg,

Thank you for this hint. Could you please tell me what firmware version are MFE parameters compatible with?
My intention is to use the new frame with MFE default PIDs. Are there any other parameters i have to consider changing before maiden flight?

All the best,

Mihai

[GregCovey]

@Mihai_Comsa

Could you please tell me what firmware version are MFE parameters compatible with?

Anything below v4.1.0, so 4.0.x.

I have not kept up with the latest changes. Hopefully, @Rolf can answer this question. You can use firmware version 4.1.0 if you use the PID converter mentioned above by Ali and pointed to by Rolf. I am not sure why the automatic conversion didn’t work for you.

[Rolf]

My interpretation of the logfile is wrong, your video clearly shows that the elevator deflections are correct.
Perhaps one of the developers will take a look at the log file. Otherwise, the cause of the faulty pitch control remains unclear. Worse than the crash itself is often when you do not even know why you crashed.

[koseki (yoshio)]

Was the AIRSPEED sensor normal?

I got into a similar out of control without mounting.