I have many successful flight hours logged on my ArduPlane platform. Yesterday I had a problem 15 seconds after take off that resulted in a very hard landing/crash. On the bright side, relatively minor damage and happy my plane didn't end up on top of the big tree...
Looking at the video, it appears that 15 seconds after take off, power from my battery was interrupted for about 1/3 of a second. The ESC beeped it's power on sequence, indicating that it lost power. The APM rebooted. My FPV camera/transmitter apparently lost power briefly.
The only thing between my battery and the ESC is the 3DR Power Module. The power connections where all tight. Even if I aggressively wiggle them, I can't reproduce this. I ran the system at full power for a few minutes after the crash and couldn't reproduce it.
I made two changes since my last successful flight (I know, I know, why not only one?).
1) I added a 3DR Airspeed indicator, enabled it, but did not configure it to be used.
2) I upgraded from 2.78b to 3.03
I don't use a ground station so I don't have a tlog for the flight. I do have the logs from the APM that I download post-flight but I don't see anything interesting.
log file starting after APM rebooted
log file leading up to APM reboot
I have my wiring diagram documented, but it hasn't been updated to include the airspeed indicator.
At this point, I'm thinking that either the 3DR power module interrupted power to everything (is that even possible?), or perhaps my Eflight 30A ESC/BEC did something to the power feed.
Maybe I'm pushing the limit of the BEC in my ESC (I probably am) and adding the airspeed indicator caused more current draw from the BEC which caused the issue? I haven't yet determined if the current sensor plugged into APM A1 port is powered from the 3DR Power Module or from the output bus (which is powered from the BEC in the ESC). Power systems are not my forte but I'm learning fast now...
Any advice or input appreciated!
Bill
Replies
You heard the ESC start boot sequence. (you say it has it's own ESC)
You also logged a APM reboot, (powered by power module)
there is no jumper between these power sources.
This would mean the power failure must have happened before APM power module, in the battery pack og connectors between it and the power module.
The only other explanation could be that you heard ESC "got PWM" sound, not "boot sound" . - depending on ESC , the difference can be impossible to mistake.
Thanks for the input Andre.
The ESC boot sound is unmistakable and audible at 25 seconds into the video - it definitely booted.
Your assessment is totally logical and I've already replaced the cabling (but not the power module), replaced the ESC, and I won't be using this battery in the air in the future.
But I'm just having difficulty accepting there was a cabling problem between the battery and APM power module, because the wiring is very solid, a tight deans connector, and even after aggressively wiggling all the cables and connections and where the wires go into the battery, I couldn't replicate the problem.
One interesting note about the ESC & Motor. For the past many months, the boot sound from the motor has been barely audible. But in the video, the boot sound was clearly audible and pretty loud. After the crash, the boot sound is again very quiet. In fact, so quiet that it is impossible to hear with my in-plane microphone, which is the same microphone that recorded the beeping in the crash video.
Can anyone explain why the volume of the ESC/motor boot sound would dramatically change? When I replaced the ESC, the new one has a clearly audible boot sound.
My #1 theory at this point is a failing ESC caused a serious in-flight brownout, even though it is still working fine after the crash.
ESC (BEC) failure would not make APM reboot - not in your configuration. - you can try it - just boot everything up, disconnect ESC, and short the output servo rail GND and 5v (which at this point does not have any voltage)
APM should still be happy.
There is more, nasty, possibilitys;
Aa cold-soldered power connector, where wire is tinned, and "fits" perfectly inside the "hole" (XT60, EC3) - somtimes burning itself (poorly) to the pin/socket when hugh current melts tin.
And: bad battery interconnect - between cells or wire. - only detectable by measuring internal resistance at high current. (same phenomenon can happen here, high current can fuse things together, or let them apart.)
In case your output side could pull down the APM power supply, it's less likely to be the BEC, and more likely a servo to cause brownout. (the nature of servos are so they current, and current of all servos, is hard to predict, and very few hobbyists fuse each servo and provide a power supply that can feed all servo's stall current simultaniously.)
If so, that's bad.
No, JP1 does not have a jumper on my board. I screen scrapped all the photos so not my exact kit.