chromvis

  • *
  • 117
Quad loses control in autotune (16.09-RC1)
« on: September 17, 2016, 06:43:45 pm »
Quad, new build with revolution (FC was used before on other quad without problems) fly well in attitude with default PID.
When I switch to autotune in the first jerk quad flips over and fall.
I tried to imitate autotune jerking in stabilized mode (just rotate slightly pitch/roll stick) and got the same behavior. With a bit more intense jerk quad flips over an fell.
It looks like gyro problem, but it fly in attitude without any problem, even sharp pitch or roll jerk alone (like speeding forward and break) does not cause this behavior.
Setup: Generic quad 250mm, revolution with OpLink control and telemetry, motors EMAX RS2205 2300kv; 30A DYS BlHeli ESC; 3s 2200mah battery; DAL 50x45x3 props. Weight 570g.
Any ideas or recommendations on what could cause this behavior or just where to look for (specific parameters etc.) will be greatly appreciated.

Mateusz

  • *
  • 808
Re: Quad loses control in autotune (16.09-RC1)
« Reply #1 on: September 17, 2016, 08:32:58 pm »
What yours target bank for autotune ? Check if you're flying different bank, then the one you're tuning maybe. I tried on 4 quads in all cases I could successfully finish autotune, but outer loop is a bit over-tunned.

chromvis

  • *
  • 117
Re: Quad loses control in autotune (16.09-RC1)
« Reply #2 on: September 18, 2016, 12:56:52 am »
Thank you Mateusz.
I actually found the cause. I am using OpLink for control and telemetry and use whip 1/4 wavelength (17.2 cm) antenna with ground wire. Apparently it was somehow broken inside and at strong jerk it was loosing connection.
What bother me is that quad did not go to failsafe but just turn upside-down and fell.
I replace antenna to whip with ferrite (Hawkeye) and was able to autotune.

Mateusz

  • *
  • 808
Re: Quad loses control in autotune (16.09-RC1)
« Reply #3 on: September 18, 2016, 06:44:26 am »
Hi again,

Interesting, can you check if powering off the OPlink on the sending side triggers fail-safe ? I think damaged antenna just lowers RSSI and to trigger fail-safe you need a complete no-pulses for some fixed time.