This has happened before on 3.1 and according to what I’ve read in the changelog for 3.2 this issue was to be resolved. Maybe in my case it depends on something else.
Video below.
machine spec:
Flightcontroller: APM 2.6
Frame: Tricopter
Motors: DT750
ESC’s: Plush with BlHeli firmware
Any ideas as to what happened?
[youtube]https://www.youtube.com/watch?v=YbsPGbXb7Gw[/youtube]
Here is the log from today. I cant really interpreate any of this but see that at some point in autotune it just went straight down to the ground
Just curious; Was it windy? The Alt Hold didn’t appear real smooth.
I’m not a long time expert, but Alt Hold was smooth and Loiter was locked in place and producing good logs (from what I could tell) before I did my 1st Auto Tune.
SC
[quote=“philsson”]This has happened before on 3.1 and according to what I’ve read in the changelog for 3.2 this issue was to be resolved. Maybe in my case it depends on something else.
Video below.
machine spec:
Flightcontroller: APM 2.6
Frame: Tricopter
Motors: DT750
ESC’s: Plush with BlHeli firmware
Any ideas as to what happened?
[youtube]https://www.youtube.com/watch?v=YbsPGbXb7Gw[/youtube][/quote]
I really can’t say anything without the logs.
But from listening to the video, it sure sounds like you have some bad motor bearings. It sounds really bad. You might have had a motor lock up or something.
[quote=“FullScalePilot2”]Just curious; Was it windy? The Alt Hold didn’t appear real smooth.
I’m not a long time expert, but Alt Hold was smooth and Loiter was locked in place and producing good logs (from what I could tell) before I did my 1st Auto Tune.
SC
[/quote]
Yes it was a little windy but not so much as to make the autotune fail but maybe get some bad calibration.
What in my log does not show a stable alt-hold? It seems quite steady to me =)
[quote=“Rob_Lefebvre”]I really can’t say anything without the logs.
But from listening to the video, it sure sounds like you have some bad motor bearings. It sounds really bad. You might have had a motor lock up or something.[/quote]
I know, it sounds bad. I think it is something with the camera because it has always given that sound. The bearings are quite new but of course I’ll check that upp. I’m quite convinced this crash would not have happened if not in autotune as the exact same thing happened earlier this year on 3.1. Maybe it is some synchronisation problem with my ESC’s?
Attaching the log here.
What are your BlHeli esc parameters and the setup? I’ve recently switched from simonk to blheli after a crash that was caused by loss of sync… Since i’ve played a lot with blheli I may be able to help if that’s the problem.
Have you tried ( at the bench ) induce a loss of sync to the motors to see if it will happen?
That log won’t do it. You’ll need to download the data .log from the FC.
You will be looking under IMU data and check accX, accY, AccZ to see what the vibrations look like.
You can find those parameters listed under Tlogs (logs stored in your MP folder) , but I don’t find enough detail there to read vibrations.
However, based on the video it doesn’t appear the craft (jerky motions) and conditions were right to Auto Tune.
My advice: Forget about Auto Tune for the moment.
SC
[quote=“philsson”][quote=“FullScalePilot2”]Just curious; Was it windy? The Alt Hold didn’t appear real smooth.
I’m not a long time expert, but Alt Hold was smooth and Loiter was locked in place and producing good logs (from what I could tell) before I did my 1st Auto Tune.
SC
[/quote]
Yes it was a little windy but not so much as to make the autotune fail but maybe get some bad calibration.
What in my log does not show a stable alt-hold? It seems quite steady to me =)
[quote=“Rob_Lefebvre”]I really can’t say anything without the logs.
But from listening to the video, it sure sounds like you have some bad motor bearings. It sounds really bad. You might have had a motor lock up or something.[/quote]
I know, it sounds bad. I think it is something with the camera because it has always given that sound. The bearings are quite new but of course I’ll check that upp. I’m quite convinced this crash would not have happened if not in autotune as the exact same thing happened earlier this year on 3.1. Maybe it is some synchronisation problem with my ESC’s?
Attaching the log here.[/quote]
[quote=“FullScalePilot2”]That log won’t do it. You’ll need to download the data .log from the FC.
You will be looking under IMU data and check accX, accY, AccZ to see what the vibrations look like.
You can find those parameters listed under Tlogs (logs stored in your MP folder) , but I don’t find enough detail there to read vibrations.
However, based on the video it doesn’t appear the craft (jerky motions) and conditions were right to Auto Tune.
My advice: Forget about Auto Tune for the moment.
SC
[/quote]
I’ve uploaded the .log file now. I’ve flown the craft in all modes with success but I experience it to be a little oscillating from time to time when it needs to do a “bigger” adjustment. The yaw PID is maybe not the best as I am having some side to side behaviour but this should be isolated from the autotune as it only works in roll and pitch?
I’ll have to order a new servo to continue flying but luckily it was the only thing that broke on this crash.
What happened on the video I find strange, even if something is a little of with my PID values it should definitely not just flip over and fall to the ground like that? It feels something else went wrong like some sync problem with ESC or something.
[quote=“tabascoz”]What are your BlHeli esc parameters and the setup? I’ve recently switched from simonk to blheli after a crash that was caused by loss of sync… Since i’ve played a lot with blheli I may be able to help if that’s the problem.
Have you tried ( at the bench ) induce a loss of sync to the motors to see if it will happen?[/quote]
It was a while ago i flashed. I could definitely check this out! would be very helpfull
How do you perform a stress test?
I agree with philsson about the conditions to autotune. Maybe it is all about that.
I’ve tested mine by following this video: youtube.com/watch?v=0lRTN5o4UCM
There’s some people reporting loss of sync in autotune with simonk fw , due to the sudden increase in throttle, like autotune does. I’m not saying that this is the issue, but it could be.
Some people advice to change the firmware to blheli, but the real deal to fix that is to tune the motor timing parameter and further the demag compensation/pwm frequency. However this is only needed if you figure out that is the problem by doing the test.
I would advice you to do that on a bench motor by motor. Try to increase throttle from low to top very fast and see if the motor lost sync ( and stops ).
The test should be done with the propellers on. Without the propellers, there won’t be no load thus the problem unlikely will shown up.
So, take all the safety procedures and be very cautious, because it is a pretty dangerous test and just do it if you feel comfortable and safe.
Just let me know if you can reproduce the issue then I can guide you through the next steps.
[quote=“tabascoz”]What are your BlHeli esc parameters and the setup? I’ve recently switched from simonk to blheli after a crash that was caused by loss of sync… Since i’ve played a lot with blheli I may be able to help if that’s the problem.
Have you tried ( at the bench ) induce a loss of sync to the motors to see if it will happen?[/quote]
It took a while, had to resolder the connectors. Here are some screenshots of the configuration. They are all the same.
[attachment=0]Screen Shot 2015-01-08 at 23.37.58.png[/attachment]
[attachment=1]Screen Shot 2015-01-08 at 23.38.04.png[/attachment]
I now have benched. Does not look pretty! I’m even surprised it has not fallen out of the sky for me more than on autotune as the result is TERRIBLE! You might think that I am not giving any throttle half of the time but trust me I am. Going min to max on the throttle punching it time after time and the motor looses sync about twice the time. Great right? So what do i do next? I’ve seen some people changing the PWM frequency on Blheli, maybe that can do the trick or at least its worth a try. Should I go up to mid-high, or maybe high, and what other settings might be good?
[youtube]https://www.youtube.com/watch?v=j18D87zEpy4[/youtube]
Wow, it is really far worse than mine was… The bench has done the duty well.
I don’t have Blheli screenshots to compare and to get them I’ll need to rewire the usbasp ( a lot of work ) , but I do have taken down notes of the parameters I’ve changed from the default.
1st thing: by your screenshot, your version of Blheli seems to be old… Install the latest one from helifreak.com/blog.php?b=2162 .
If you haven’t done, upgrade the Fw to the latest code ( rev 12.2 ) Use the appropriate version for your esc. see if helps by itself.
If not , try this changes from the default multirotor config:
Play with those two parameters, they are key points to solve the issue. My goal was to maintain the commutation time to the lowest value possible without having issues. High commutation time means more performance thus more energy wasted ( also the motor runs hotter ) . Low comm time saves more energy.
The demag is a pretty complicated topic and it depends on the physical specs of the motor. So, you could try as well to disable that. It may fix the problem.
My setup is a turnigy multistar 2216-800Kv / 30A HK esc (F30) / 4S lipo / 12x4.5 prop.
What I have done to fix the issue and archive the best results was:
3** decreased the damping force from highest to medium-hight
4** changed the pwm frequency to DampedLight
** those two last mods aren’t recommended by autopilot manuals . In addition, the DampedLight mode needs to be supported by your esc Hh. I have changed them to play a little and find out that the copter felt more responible , so i let them as it is.
Everything else is at their BLheli defaults.
Try it and let me know ! good luck !
[quote=“tabascoz”]Wow, it is really far worse than mine was… The bench has done the duty well.
…
Everything else is at their BLheli defaults.
Try it and let me know ! good luck ![/quote]
Thank you very much for your input, I’ll try those settings out! I also read a little about “Throttle rate” which is 255 by default gives no delay on speed changes. Maybe that value can also solve some issues if lowered a little?
Yes, for sure, but it is good to avoid any delay to be sent to the motors so, let this as a last resource. I would go for those parameters first ( in order ) :
Please let me know the progress and what you have tried.
PS. I have read some people saying that some motors aren’t compatible with some esc models. I, personally, think that it is a matter of trying hard to find the exact settings for them to work well. It should be always possible.
Regards,
[quote=“tabascoz”]Yes, for sure, but it is good to avoid any delay to be sent to the motors so, let this as a last resource. I would go for those parameters first ( in order ) :
Please let me know the progress and what you have tried.
PS. I have read some people saying that some motors aren’t compatible with some esc models. I, personally, think that it is a matter of trying hard to find the exact settings for them to work well. It should be always possible.
Regards,[/quote]
Started in the order you recommended. I have four videos here from the results and a comment to each
New firmware v12.2 original settings. Changes from the stock settings compared to before was PWM frequency set to high among other things. Sync issues still the same. I had to reverse the direction as the motors rotated backwards before trying and filming.
[attachment=2]orig 12.2.png[/attachment]
[youtube]https://www.youtube.com/watch?v=HNfCloLE7Y8[/youtube][/quote]
High timing now. No big difference I guess?
[attachment=1]high timing 12.2.png[/attachment]
[youtube]https://www.youtube.com/watch?v=2Auxv-AzIHo[/youtube][/quote]
High Demag. I think the sync issue is gone completely but the throttle response is really slow. Adding a last clip just to show that.
[attachment=0]high demag.png[/attachment]
[youtube]https://www.youtube.com/watch?v=a_n8TkzAidA[/youtube][/quote]
Here is the slow throttle response. I filmed the remote so you can se the delay. It goes a little faster when the motor already has some speed. At least there seem to be no sync issues anymore
[youtube]https://www.youtube.com/watch?v=RXZzaFaGwLA[/youtube][/quote]
Any suggestions as how to continue? I noticed the settings are set on high PWM frequency, should I try to lower that? Maybe I should go out for a fly? Curious to hear your input
I looked at the videos a few times and I realize that the motor bell is lifted every time I increase the throttle. Very well it is “loose”. I tried the same test on another motor, this time the left front on my tricopter which was stiff and with standard settings in blheli there are no tendencies to sync issue. Great I think! I tighten the bell on the original test motor (right front) and go back to stock blheli 12.2 settings (besides reverse) but it does the same thing. Now… clearly two ESC’s and motors of the exact same modell and specification are giving different results. One being really responsive without sync issues and the other giving sync issues almost all the time. What is most likely wrong here, the ESC I guess and not the motor?
Here the almost stock settings (on both motors the same now)
[attachment=0]Standard on new motor.png[/attachment]
Here is a clip of the left motor
[youtube]https://www.youtube.com/watch?v=KquSnyjHZ_M[/youtube]
And the right
[youtube]https://www.youtube.com/watch?v=zQ9qKU4tot4[/youtube]
Last test. The one ESC I did not update yet. Works flawlessly doesn’t it?
[youtube]https://www.youtube.com/watch?v=XexYZVb0GOU[/youtube]
Hello Philsson
Weird… One thing for sure it woudn’t be a great idea to have different settings among the motors… it could produce different thrust responce , making things difficult for apm to control the copter. So my advice is to try to find out if the culpid is the motor or the esc by mixing the set and see which part causes the failure…
IN meantime, when you find out , if you have spares to replace thats fine. If not, you could try to fly with all esc’s setup with that dumb slow response. I think the copter would fly but clearly with a slow response… so try to avoid windy days and the autotune.
One last thing on the failing motor: try this settings and see if it works:
Daming force : very low
PWM Freq : low,
demag: low
Temp prot: off ( could be a failing sensor on the esc )
PS; Sorry for the late answer, here is around 10 am and now it is around 38 degrees Celsius , i’m going to the beach and will be back in a couple of hours. 
[quote=“tabascoz”]Hello Philsson
Weird… One thing for sure it woudn’t be a great idea to have different settings among the motors… it could produce different thrust responce , making things difficult for apm to control the copter. So my advice is to try to find out if the culpid is the motor or the esc by mixing the set and see which part causes the failure…
IN meantime, when you find out , if you have spares to replace thats fine. If not, you could try to fly with all esc’s setup with that dumb slow response. I think the copter would fly but clearly with a slow response… so try to avoid windy days and the autotune.
One last thing on the failing motor: try this settings and see if it works:
Daming force : very low
PWM Freq : low,
demag: low
Temp prot: off ( could be a failing sensor on the esc )
PS; Sorry for the late answer, here is around 10 am and now it is around 38 degrees Celsius , i’m going to the beach and will be back in a couple of hours. [/quote]
No worries about response time, I’m more than happy you take the time to help me out and I am in no rush to get it airborne. =)
Of course it would be silly to fly with different settings on the ESC’s. To clarify I was just trying to show the different behaviour and would of course set all the same before flying.
I am pretty sure I have a spare ESC lying around somewhere. Looked around the apartment for an hour before I started doubting myself. Then checked my order history and TADA! I am supposed to have a spare so I’ll just keep looking for it. Think I’ll replace the bad ESC but the idea of a damaged temp sensor is a good one! Will try that out right away actually and if it works I think I’ll leave it like that. If not I’ll have to find that other ESC first and then patch some cables to it for reflash… :geek:
EDIT: Unfortunately it was no temp sensor error… I’ll continue to look for the replacement 