Reproducible death roll after snappy roll.

tavis

New member
I'm flying a Martian 215, Matek f405, blheli 20es, MT2204-2300kv, 4S.

After doing a fast roll, left or right, before I raise throttle, the copter goes into a death roll. Yesterday it happened twice, each time on a different rear motor. After rolling right the rear right motor stops, after rolling left the rear left motor stops.

Attached are the blackbox log files for both events and an image of the blackbox graph of one of the events.

In both instances the FC commands the motor to 100% to stop the roll maneuver. The rear motor stops producing thrust and the spiral ensues.

Could this be a too high P value ? Could this be desync or demag ?

I've discovered a way to reproduce the spiral of death. At the end of a full stick roll, if I release the stick and let it come back to center via the gimbal spring, the spiral of death will occur. Doing a roll to the right will cause the right rear motor to stop responding, and a left roll the rear left motor.

I've already increased my min throttle command. In all cases the minimum motor command of the motor that stopped was at or above 144 just before the spiral. In the last case it was at 311 . (these are raw values from blackbox)

I'm leaning towards the cause being the 5050 props on the cheap 2204-2300 motors and the motors not being able to keep up with the rapid changes commands with these aggressive pitched props and causing demag events.

View attachment Archive.zip
 

PsyBorg

Wake up! Time to fly!
5050s are aggressive for most motors. Specially if they are tri blades. The fact you lose one motor to me sounds like a typical failsafe condition. You may be drawing so much current you are browning out the other electronics and it only looks like the snap back to center is the cause because the maneuver is such a fast one stick wise. You could also be back feeding current when you let off which is a whole other set of issues.

I pull over 130 amps on 2306 2400 kv motors with tri blades. When running dual blades that drops to ~115 loaded and high 90s unloaded. I am running 45a esc's though so I am sure your not pushing that much current other wise them smaller motors would have lit up like a Christmas tree the first time you tried to flip.
 

tavis

New member
When you say typical failsafe do you mean Rx/Tx failsafe ? There is no indication in the logs that an Rx failsafe occured.
 

PsyBorg

Wake up! Time to fly!
There wouldn't be as far as I know. Unless something changed since the last time I did black box only the motors, gyros and accelerometers are charted. That would show up more like a motor / esc failure. Every time I have had a fail safe in a quad that was set up to drop when the fail safe occurred I have always noticed one of the rear motors shuts off and the quad tumbles to the rear before everything else shuts down. I think that is intentional to stop forward motion and that time frame from the initial motor stop to complete shut down goes by the gyros and accelerometers as they sense the speed decrease. I could be wrong but I believe that is how it seems to work.

Easy way to confirm is to toss on some 5x4 props and lessen the current draw and see if it still happens.
 
Last edited:

cranialrectosis

Faster than a speeding face plant!
Mentor
Did you hard mount or soft mount your flight controller and did you tune the PIDs or are you running defaults?
 

tavis

New member
I tried 5x4 props and it still happened. I also changed the max_ppm value in the esc config and it still happened. I CANNOT reproduce the problem when running 3s battery so my conclusion is that my cheap ass motors can't handle 4s.

regarding failsafe in blackbox, there is a state flag that tracks failsafe and flight mode so an rx failsafe should show up in the logs.
 

PsyBorg

Wake up! Time to fly!
See ya learn something new every day. I have not used nor needed black box in a long time. Specially with the reliability of the Kiss and Emax gear I use.