Another crash with nothing in the logs

ArduPlane VTOL Plane
1

My large 2.2m VTOL just nose dived after transition and is in a thousand bits. I’ve flown 100s of times with this plane, and nothing in the settings has changed much. I did a flight 10 minutes before with no issues. Back in the spring, I had about 5 events like this with a 5kg quad but all from a low height while testing. The logs just stop while everything is fine. That was with a Matek F405. This time its with a Matek H743. Not common parts between them. I tried everything from batteries to SD cards.

Obviously I’m thinking total power loss. The FC was still powered when i got to the wreck. It does ‘crack’ when the 6s battery is connected and Ive been meaning to fit an anti spark thing. I have powered up the bits on the bench and wiggled everything and have had no power drops.

Honestly, these mystery crashes so unsettling.

My next build with a pix 6c mini is coming along and will have expensive optics onboard so really need to get this figured out.

Any ideas guys?

2

Looks like Power Loss to the Flight Controller as you suspect.

3

The onboard video file was corrupted but I’ve just paid $9 to get it fixed. It also stops at about the same moment so certainly looks like the battery connection but I can’t recreate the fault on the bench.

Anyway the next version will have xt90s with anti spark.

4

I would try and measure voltages when you are hearing the “crack” Imo dual power input like cube is best.

5

The issue occurred when the current was reducing after transition and never exceeded the XT60 rating. In the absence of a smoking gun I’ll change the main battery connection to XT90 which will allow an anti-spark connection.

Also I’ve had enough of crashes when logs just stop and give no evidence, so my latest build, the Kite4, I intend to try using a super capacitor in the secondary supply to the 6c mini (via usbc). In no way is it power redundancy but at least it should give a second of logs point to either MCU lockup or power disconnect.

6

While I’m in no way assigning blame, this seems a systemic enough problem with your builds to at least ask the question: are you certain you have calibrated the battery monitor(s) correctly?

If not, it’s entirely possible that you are too deeply discharging your batteries without knowing it, which could account for sudden power loss.

7

That’s definitely not the issue here or on previous issues. The battery was at about 85% for the flight. No battery warnings or significant sag.

8

That’s the thing, though. Do you KNOW that, or are you trusting uncalibrated/assumed voltage/current sensors?

If you’re just using stock scale factors, you could be significantly off unless you’ve independently verified their accuracy.

I’m definitely not convinced this is the case, but it’s an easy smoking gun if you are unsure of the calibration state of the hardware in use.

9

Off the battery I have an ‘i2c ultra accurate sensor’ then there is the standard matek H743 inbuilt sensor, then there are 5 CanBus ESCs all reporting (admittedly via a credible FC) all stopping dead . I don’t know what more to do,

10

Probably rules out my spitballing.