I have a problem where my motors suddenly accelerates for less than a second and then back to normal. It is completely random, It happens on random motors, at random frequency(generally once every 1 to 10 seconds.) It starts happening right after mounting, regardless my radio is turned on or off.
DJI F550 + E300
Rasparry Pi2 + Navio
Spectrum DX8 + AR8000
First of all, you are running a standard set up with our real-time image and APM deb?
When this happens, is APM software running or you simply connect it to Navio+ and Pi?
Does it happen on disarmed copter?
ESCs are connected with ground wires, not only signal?
I am running the provided read-time image with a APM software compiled from source code.
This only happens when the motors are already spinning. Since I have never attempted to start the motor when APM is not running and mounted, I can’t tell if this could happen otherwise. I could try it when I get home.
And yes ESCs are connected with ground wires.
I just want to add that everything else looks normal. I can control the speed of the motors with my radio and if I tilt the hexacoptor, I can hear the speed change of the motors.
It seems to be a bug in the firmware as it happens if I constantly restart the firmware until the initialization of the compass fails.
Then I notice sudden full throttle spins of the motors.
Can you please explain the problem in more details? If I understood you correctly you are experiencing short accelerations of motors when they are already spinning because you armed them?
There is only one known scenario when motors spin up on start up. On initialization we assume that PWM generator is at unknown state and perform full reinitialization. It includes shutting off the pulse generation, which might result in one pulse being cut off, so shorter than normal, if line happened to be in high state when PWM was shut. This behaviour is fine for most ESCs, but for EMAX and probably some others with outdated BLHeli firmware it might result in ESC thinking that a short pulse is actually a signal at different frequency (1Khz) and that it is not less than minimal throttle, but rather high throttle signaled at a different frequency. This issue was reported to BLheli and they told us that it was fixed in subsequent releases. Now it would not switch frequency after one short pulse. Unfortunately some ESC manufacturers might still flash the old firmware. This issue is not Navio specific, it is easy to start motors with these ESCs by simply taking out the connector as it would as well result in a cut pulse and described consequences.
This could never happen on the first start up, as ESC is not yet initialized and would not spin until throttle is held low for some time and until it plays it’s song.
@dgrat which ESCs are you using? If you are sure that described scenario is not what is happening in your case, do you have a scope or logic analyzer to check the signal? Can you describe the actions to reproduce the error? I am not able to start the motors by constantly restarting the firmware.
I will record a video about the problem soon, so you know exactly what happened.
I am using a favorite sky III 4 in 1 esc for my current copter build.
I think it was not that great to use them, but it is at least compact.
see this: http://mymobilemms.com/OFFTHEGRIDWATER.CA/index.php?main_page=popup_image&pID=423&zenid=i8v6vgfcdvl6ks116mjb9pgo47
It is indeed never happening on the first startup.
Nevertheless, I noticed these problems when I wrote the init.d scripts.
So my ESC already were initialized and I restarted the firmware several times for testing purposes.
Which was problematic.
For my bigger copter I use four 60A Turnigy Plush ESCs.
However, I was not testing them so far with Navio+.
Do you know whether Plush ESCs also have this problem?
So when you start the APM, you get a short spin (about 1 sec) and then everything is back to normal?