Motor 8 stopped working, and to compensate that the autopilot commanded it to the maximum. 5, 6 and 7 are fine.
the log shows how much motor thrust ardupilot is asking for not how much its getting. so when the motor output went to maximum, its because the motor had failed ardupilot commanded more to compensate until it hit 100% and because that was ineffective it had to reduce power to the other side to compensate.
if your octo was properly tuned and set up it would have continued to fly but because its so out of tune it couldn’t compensate fast enough and rolled.
1 Like
amilcarlucas, My bad, I just realized, no where so far have I mentioned the orientation of the octo on its way to the ground, it didn’t just fall straight down and land flat on its bottom. The octo instantly flipped, what seemed like about 90 degrees, onto its side and drove itself on its side into the ground. If motor 8 stops on one side and 5, 6 and 7 max out on the other side it seems to me that would explain the flip?
its not tuned, it barely flies so when it lost a motor it couldn’t compensate as it could barely fly as it was.
1 Like
Follow these steps:
- Repair the vehicle.
- Reset all parameters to defaults, install ArduCopter 4.5.5
- Download and use the ArduPilot Methodic Configurator >= 0.9.6
- Do not skip any steps, even the ones that do not seam important to you. If a step is not important it is explicitly marked as optional. The process has already been streamlined to the maximum, do not streamline it more!!!
1 Like
You are not understating how closed loop control works. The RC outputs are command signals, not thrust. So if there is a lack of thrust on Output 8 (motor, esc, prop, power failure) The autopilot will command a higher output to compensate. If there is insufficient thrust being produced on that motor this command signal will reach max .It will lower other outputs (not max them out) to attempt to stabilize pitch, roll and Yaw.
Imagine what happens when motor 8 loses thrust for whatever reason?
The opposite motors will have to drop in thrust to stabilize the craft. Now there is a overall lack of thrust.
If this Octo craft was properly tuned it probably would have recovered enough to land it. Wouldn’t happen with a quad. We see this classic output signature on a weekly basis here with quads crashing due to a thrust loss on one motor.
1 Like
Light bulb! Thanks Dave, got it, duh, just because the command for 8 to do something doesn’t mean 8 actually did it or was capable of doing it and it all came unglued from there. What I didn’t get was that what the RCout plot was telling me was what 8 was commanded to do not what it actually did.
I’m not sure how to move on, I don’t know what happened that caused 8 to fail? The octo is a heap beyond repair, I don’t trust any of the components, the crash was so violent that even one of the esc bullet connectors was bent and so were at least two prop screws.
I mentioned at the beginning of this thread that I have a arducopter quad that I hadn’t flown in over a year. I flew that recently and experienced a similar instantaneous flip in the process of just slowly landing straight down. Just as quickly as is it flipped it corrected itself and I was able to land safely. For some reason there are old logs on the quad but no recent logs and no log for the incident which is killing me, why no recent logs?
I figured the quad would be an opportunity to start from scratch with setup and configuration. amilcarlucucas suggested Methodic Configurator so I give it a look and stopped dead in my tracks…“avoid these components” described my entire build. I’ve lost confidence in what to trust and don’t know how to check individual components to see if they are trustworthy. Like for motor 8, I’m mostly suspicious of the ESC, is there some way to test an ESCs that might confirmed that it’s faulty?
1 Like
Don’t ever use those IMO. Directly solder.
They can really only be tested under load. Buy good quality components, not the cheapest you can find.
1 Like
Dave,
I have questions regarding your response to Dave Wiener in 25/28, and I am hoping you can share more insight.
-
If RC outputs are command signals and not thrust, how can you tell if motor 8 went to max because the flight controller was trying to compensate or if the transmitter was calling for more thrust because the stick was pinned as far as possible?
-
If motor 8 lost thrust, is there a way to determine in the logs if there was a failure with the ESC or motor?
-
So motors 5,6, and 7 decreased thrust to compensate? What should have been done in the tuning process that would have prevented this crash? How should the octocopter have recovered?
-
You see “this classic output signature on a weekly basis with quads crashing due to thrust on one motor.” Is it usually due to inexpensive components and quads that are not tuned properly? Would better components and/or proper tuning be the way to eliminate most of these crashes?
-
What was not done in the tuning process that would have prevented this? The autotune was running.
It didn’t, the command signal went to max. The motor could have stopped.
No.
Octo’s, and even hex’s can fly, get to the ground anyway, w/o a motor if there is sufficient thrust/weight to start with and tuning, which is an involved process considering many parameters, is good and responsive.
Sometime you don’t know. Props break, you see it. Motors are hot, which a poor tune with output oscillation can cause. Shit ESC’s like generic “SimonK” units sold for $4. Failed bullet connectors…
It should not have been. Quite a bit of tuning usually needed before Auto Tune will produce a good tune or work at all. Something else we see on the Forum all the time, “failure to level” messages.
2 Likes