4+1 Quadplane Sudden Crash in QLoiter with nose dive

ArduPlane VTOL Plane
, , ,
1

Hi,

I’ve been working on a 2.9KG 4+1 Quadplanem with 2 elevons and no rudder design, as shown below:

image
image1092×713 50.6 KB

It is a Flightory Super Stingray Design.

The details of my build are as follows:

  • FC: Cube Orange Plus
  • GPS: Here 3 Plus.
  • Wingspan: 1130mm.
  • Wing Area: 31dm2.
  • VTOL Propulsion: 2806.5 1300KV with 7042 propellers.
  • Wing Propulsion: 4008 330KV with 15 Inch propeller.
  • Battery: 6S.
  • The COG for wing mode is about 60mm from the leading edge (Marked in red in image), which is recommended, but for VTOL mode, the drone is front-heavy, and the front motors have to work harder than the rear motors in VTOL mode.

I’ve tuned the VTOL mode, and it flew really well and stable. I tried Qloiter, and it performed amazingly. No vibrations or oscillations whatsoever.

You can check the Qloiter stability here:

The flight log for this stable flight can be found here:

The next day, I went to the field to test the maiden forward flight.

I was flying at around 50m altitude, so that if the wing stalls, it has altitude to recover.

I switched from Qloiter to FBWA, and it started going forward. Since it was drifting with the wind (coming from the left side of the drone), I switched it back to Qloiter, and after it stopped going forward, I pulled back the pitch stick to move it towards home location backwards, and it came closer to home location.

I tried to face it a bit into the wind, so it drifts less.

I switched back from Qloiter to FBWA, and again it started going forward. After around 15 seconds, I switched back to Qloiter. After it stopped going forward, I pulled back the pitch stick again, and it was coming closer to the home location backwards.

Suddenly, while coming towards the home location, it flipped and nose dived straight to the ground, without any control.

No warning, nothing whatsoever!

The drone, which flew really well a day before, just nose dived all of a sudden and got destroyed.

The flight log can be found here:

Can anyone please help me understand what happened?

Thanks,
Divyanshu

1 Like
2

The power supply (red line) to the quad motors was suddenly interrupted, as evidenced by the sharp drop in power consumption. The Arduplane control system itself appears intact at the PWM outputs and is attempting to raise and stabilise the VTOL. However, this is futile without power to the motors. Without ESC telemetry, the cause cannot be directly determined. ESC shutdown due to overheating or excessive current flow would be a possible cause.

Screenshot 2026-03-12 071159
Screenshot 2026-03-12 0711592540×1339 71 KB

Rolf

3

What is the ESC are you using ?

4

And, where are the ESC’s located? Do they have any airflow?

5

Hello @Rolf,

Firstly, thank you for taking the time to view the logs. I really appreciate it!

The current suddenly dropping to zero is a very, very important observation! It really streamlines the analysis.

My wiring was as follows: Battery --→ PDB with current sensor + FC Voltage Regulator –→ FC, 4-in-1 ESC, and Wing ESC (all in parallel).

After seeing your comment, it was clear that there was a sudden power loss.

I checked the wiring of the ESC, and it appears fine. There’s no wiring fault or cold joint at all anywhere.

Since the FC was logging at the time of power loss, it means that the FC was getting regulated power from PDB, and only the ESC stopped.

Initially, I thought that it might be possible that the current shunt resistor of either the 4-in-1 ESC or the PDB might have internally broken due to excessive current, causing power loss. I checked the current shunt resistors of both the PDB and the ESC, and both are shorted, so they are fine.

Next, I checked plugging in the ESC to the AM32 configurator, and all 4 ESCs are getting recognized, so that’s also fine.

@dkemxr @juvinski

The 4-in-1 ESC that I used is a Hobbywing Xrotor FPV G2 65A 4-in-1.
It was mounted inside the fuselage with a small 24V 2507 fan on top of it, blowing air away from the FETs. There was a small opening above the fan to throw out the air from the fuselage.

The temperature protection and low voltage protection features were already turned off before the flight.

Do you think that the ESC turned off due to overheating? What else could cause such behaviour?

Thanks,
Divyanshu

6

Hi @harkhka ,

My experience with 4-in-1 esc for Vtol is not good. My tests was with some t-motors f55 was terrible - I lost 2 3d printed vtols - with the 4-in-1 - and the same project, same hardware but using regular - t motor f45a escs - the same vtols are flying great.

7

Thank you for your insights. Really helpful.

I am also thinking about using single ESCs instead of 4-in-1.

Just want to be sure that it was the culprit, and nothing else.

Do you mount your single ESCs below the props, or inside the body?

Thanks,
Divyanshu

8

Those are AM32 so you should have configured for Bdshot. There may have been ESC temp in the log.

1 Like
9

Hi,

I wanted to use that, but the limitation with Cube Orange Plus FC is that it can’t use BDSHOT, as it has IOMCU for ESC outputs, and the communication between FMU and IOMCU is only one way.

Thanks,
Divyanshu

10

Sure it can.

Move the motors to the Aux outputs. You don’t have them assigned to anything. So, Flash the Bdshot version of firmware to the FC and then move all the motors/ESC’s that support Bdshot (at least 4) to the Aux outputs and configure for Bdshot.

11

Yes, that’s correct.

Will do it in the next build. Do you know of any limitations in using bdshot?

12

No. And the telemetry is at a much higher rate than serial telemetry, which you were not using anyway, and a better reference for the notch filter which you did not configure.

13

Thanks for your response.

Usually, I always configure the notches, but I prefer the throttle notch method. All my bigger Copter builds (around 6 Kg) fly very well with it. It doesn’t require separate telemetry or bdshot, and works well.

For BDSHOT or ESC telemetry, you require shorter wires for signal integrity, which can de done nicely in a 4-in-1 ESC mounted close to FC. But if you use single ESC mounted closer to the motors in a big build, using BSHOT will cause signal integrity issues with loong wires running close to power lines.

Also, using RPM notches can quickly overload the system if not set correctly (that is multiple notches and harmonics), and can also introduce much higher phase lag as compared to a properly set up throttle notch. That’s my thought.

Even for this build, I wanted to set up the throttle notch, and I dumped the raw IMU data and used the Filter Review Tool to see the frequency spectrum and filter effect. But the LPF was really doing a good job, and there wasn’t any frequency peak till 175Hz to filter out, so I left it unconfigured.

You can see the stability in the video link.

Thanks,
Divyanshu

14

Hey,

Where have you mounted the single ESCs? Under the props or inside body?

Thanks,
Divyanshu

15

While that might be true if it’s configured incorrectly it’s not difficult to do it right.

Perhaps. The results depend on how Raw IMU is configured. Many get this wrong and the results look great but are not real.

16

Hi ,

inside the body

17

What type of battery you are using, LiPo or Li-ion, and what capacity? It seems you already have high currents in QLoiter, so adding the FF motor during transition might put additional stress on the battery pack as well