I’m having a X4 carbon foldable quad. Just got a Pixhawk, all set up with a zener diode, capacitor & seperate UBEC as backup power, firmware 3.2.1. Did an autotune on day 1 and the quad flies well. Was flying on the first battery pack the other day and there was no problem in Stablize, Alitude Hold mode which I flew in most of the time. Make the quad RTL after ~12mins and landed safely. Put on a 2nd battery pack started flying a little further in Alt Hold to test my vtx range, I flew the quad back after the test and decide to try out Pos Hold mode. However after a few minutes adjusting the position, the quad’s 4 motors just stop, the quad flipped and dropped like a rock from the sky from 80m. Luckily it didnt hit anyone, but since then I have no confident of flying it again until I got the problem sorted out. I tried analysing the logs but still can’t come to a solid conclusion. I would appreciate if anyone could point me to the problem.
Looks like mechanical failure as the motors went to maximum but the current did not follow. I think the ESC’s may have gotten too hot and shutdown or something.
Looks like you hit a ESC low-voltage cutout at 14V, which corresponds to ~3.5V per cell on your 4S battery. This would explain the simultaneous shutdown of all motors.
Also note that your VSERVO voltage disappears (begins to drop away and never returns) at the same time as the cutout. VCC (pixhawk 5V) does not droop. These appear under POWR category in the log.
What arrangement of zener diode do you have, and what it its purpose? You should be able to simply connect a single ESC BEC to the pixhawk, and it will only use it if the internal VCC disappears (it has a LinearTech LTC4417IUF#PBF voltage selector device that will select between VDD_5V_BRICK (external input), VDD_SERVO (servo rail), and VBUS (USB input) in that order.
Really, there is no reason to have low voltage cut-off enabled on your ESCs for something like a Quadcopter, unless you have Schadenfreude. Also, you need to be very cautious of Voltage per Cell on your battery pack.
For some reason APM:Copter firmware doesn’t calculate what the battery cell count is and base alarms on this, instead it defaults to the 3S battery pack and typical 3S Li-Po battery low voltage threshold. Unless you customise the low voltage alarm threshold, it won’t alarm until your 4S is depleted because it is set to 3S low voltage threshold. For this reason I always utilise an independent low-voltage alarm attached to the balance plug as backup should I fail to configure APM:Copter appropriately.
What’s the Zener diode for? Did you place it for reverse polarity protection? A “regular” zener will create a significant voltage drop. Was that taken into consideration?
Hi SuperJim, I looked at my blueseries 30a ESC with Blheli 13.2 and the low voltage, temperature protection are not enabled.
[quote=“SuperJim”]Also note that your VSERVO voltage disappears (begins to drop away and never returns) at the same time as the cutout. VCC (pixhawk 5V) does not droop. These appear under POWR category in the log.
What arrangement of zener diode do you have, and what it its purpose? You should be able to simply connect a single ESC BEC to the pixhawk, and it will only use it if the internal VCC disappears (it has a LinearTech LTC4417IUF#PBF voltage selector device that will select between VDD_5V_BRICK (external input), VDD_SERVO (servo rail), and VBUS (USB input) in that order.[/quote]
I wanted a backup power supply to the pixhawk in case of power module fails. I wired a UBEC to servo rail and understand that a Zener diode & capacitor is required to condition the power supply. The diode I used was 1N5339 and the cap was 10v 1000uf. I had it placed in reverse polarity. I did the wiring just as according to the ardupilot wiki.
[quote=“Felixrising”]Really, there is no reason to have low voltage cut-off enabled on your ESCs for something like a Quadcopter, unless you have Schadenfreude. Also, you need to be very cautious of Voltage per Cell on your battery pack.
For some reason APM:Copter firmware doesn’t calculate what the battery cell count is and base alarms on this, instead it defaults to the 3S battery pack and typical 3S Li-Po battery low voltage threshold. Unless you customise the low voltage alarm threshold, it won’t alarm until your 4S is depleted because it is set to 3S low voltage threshold. For this reason I always utilise an independent low-voltage alarm attached to the balance plug as backup should I fail to configure APM:Copter appropriately.[/quote]
I just had my Battery fail safe to kick in at 14v but with no action as I would fly the quad back myself once the Battery Low flashes on my screen which is usually done in less then a min. Am I missing out something? I was monitoring the battery voltage via my screen with minimosd extra also done with the battery calibration.
I’m not familiar with the zener-diode modification, but it’s not necessary unless there are other things like servos (these could induce a back-emf surge onto the rail) also attached to the VSERVO rail. In any case I doubt it’s the cause of your problem.
The fact that the VSERVO voltage disappears indicates that the ESCs lost power completely, as the Blueseries BEC is a separate linear regulator connected directly across the battery input (it can only disappear if the battery voltage disappears). I would look at the power module output; a bad connection there could be the cause. Initially I suspected the current-sense resistor, but the Pixhawk power-supply is sourced down-stream from that.
You mentioned you were testing a VTX, did it fail at the same time as the motors? Where is it powered from?
I’m sorting out a similar case with my hexa running pixhawk (however only from a metre up testing 4S) I get a complete halt in all log events, thus most probably a power failure.
Now, i have all 6 ESCs connected with their power supply as well as the primary power supply. Could that have caused a brownout, if the main power went out and the paralleled power from the ESCs was considered not valid due to just paralleled?
I wouldn’t connect all the BECs in parallel, as it’s possible that they will output slightly different voltages, and get a bit hot (the ESCs); it shouldn’t cause any major issue though. On mine I use the power module for Pixhawk, with a single BEC connected to the VSERVO rail as a backup (this would only be effective only if the power module switching regulator failed, or the power input cable failed / came loose).
If the logging events stopped, your battery connection was probably broken, given you have redundant power sources, a single failure wouldn’t shut down the Pixhawk.
Yeah, must be a power loss of some kind. Unfortunately, my separately fed BEC was not connected for this “just gonna check” flight, so loss of main power erases all backup and logging, stupid of me
But, the batteries are paralelled and works fine. Connected via an AttoPilot power module (need 180A capability). This all grinds down to the AttoPilot circuit and its connections…
Any thoughts among you guys about this break down:
Given:
-Batteries are ok
-Power failure only to motors would keep my logg still running in the PixHawk, as it would still be fed from the power module via the primary power input.
-failure of primary power to the PixHawk would be solved via the BECs as secondary power input, even though paralelled.
Conclution:
Failure of power connection input to the power module or internal failure of the power module, cutting all power output.
Actions:
Strap the hexa down and run it on grund till it happens again with stand alone backup power secured to PixHawk.
