I originally switched my 3DR Hex (y6) from an APM to Pixhawk without issue - flies great. I decided to perform the same upgrade with my Quad.
On my maiden everything flew very well; stab, loiter, alt hold. As part of my testing, I usually perform some hard left/right rolls and front/back pitch movements (sticks full in one direction, then neutral) to see how the copter is reacting. Immediately after trying a hard right roll, the copter would not right itself and just flew in one direction - I was able to counter with some stick input, but cut my loses and killed the throttle.
This has happened twice now, first in STAB and last time in ALTHOLD and I cannot pinpoint the issue. I’ve attached both logs in hopes one of you can give it a look with a fresh perspective. Thank you in advanced!
Pixhawk + 3DR power module
850kv motors
10x47 APC props
20amp ESC
3DR OSD and RFD900 Telemetry
both your logs stop midair, this is due to a known cluster/FAT related bug - until it’s fixed you can only compile a version without that bug, or try another, microSD - some are not affected.
Anyway - what I could see, is that you have a little bit too much Z-axis vibrations, this could cause gyro drift.
If we assume your Pixhawk did not came loose and tilted, I’d bet on vibrations - your vibrations gets bad with ThrOut.
[quote=“Andre-K”]both your logs stop midair, this is due to a known cluster/FAT related bug - until it’s fixed you can only compile a version without that bug, or try another, microSD - some are not affected.
Anyway - what I could see, is that you have a little bit too much Z-axis vibrations, this could cause gyro drift.
If we assume your Pixhawk did not came loose and tilted, I’d bet on vibrations - your vibrations gets bad with ThrOut.[/quote]
Hi Andre,
Thank you for your input!
Regarding the logs & cluster/FAT bug, would that have anything to do with the crash? or is this an innocent bystander? I thought this was fixed in the latest arducopter release :\
I will work on the vibrations. It’s very odd since I used the same foam mounts that previously held the APM - never had any behavior close to this.
no. the FAT problem cannot affect the flight, and is not fixed in 3.1.2.
Yes, I find the vibrations theory a bit sketchy too, because your X,Y vibrations are fine - but I don’t see any other reason to such behaviour.
make sure you don’t have an solid part of the frame touching the top of pixhawk’s top/connectors. - a well mounted device, could sometimes get vibrations from above, if a solid plate is on top of it. - I am guessing now - but you should take a look at the Z-vibrations just in case, I think it’s a probable cause.
i took a closer look again at my AccZ and there are some dips at some points but for the most part it is well within the margins. I’m having a hard time accepting this is the cause.
I will double/triple check and make sure nothing is touching the px4 as you suggested…
If anyone else has any ideas or input it would be greatly appreciated.
I agree, - it’s not that bad (like I mentioned) - but the spike follows thrust, - it’s still not impossible that hard input can generate much more vibrations that we see early in the flight.
And/or resonance/harmonics can cause it to spike when a possibly badly balanced motor+prop is running fast…
So, I do not say this is the cause, it’s just a guess based on the short log we see.
[quote=“Andre-K”]I agree, - it’s not that bad (like I mentioned) - but the spike follows thrust, - it’s still not impossible that hard input can generate much more vibrations that we see early in the flight.
And/or resonance/harmonics can cause it to spike when a possibly badly balanced motor+prop is running fast…
So, I do not say this is the cause, it’s just a guess based on the short log we see.[/quote]
Understood - thank you!
So i took the quad totally apart, checked all connections and ensured no wires were touching the pixhawk. I went through 4 batteries of hovering & rolling/pitching without a hitch. then:
[youtube]https://www.youtube.com/watch?v=Xxv5b_LNJ3w[/youtube]
Basically it starts with copter banking left hard, I counter with some stick input and try to bring it down gently, but towards the end I lose yaw completely and start spinning. I’m thinking my problem is either an ESC or motor…
I tried to download log from the pixhawk but the log is no good - whenever I try to download it, I get an error about not being able to rename the file from Mission Planner :.
I’ve ordered replacement ESC’s, but i’ll take any input I can get.
you can also post the .bin file from your microSD.
Yes, a ESC/motor problem could be possible reason to this. Still, it seems you have some lift on each arm, if your props are mounted on the mtor axle, did you check for possibility of slip between the prop and the axle, or axle and the motor casing (the bell with magnets) .
I also hope/assume that all props or prop-hubs(after crash) were firmly attached ? - a loose prop could also give a similar result.