Please help analyse my log - unexplained flip over in loiter

Thanx a lot , I found this parameter , I tried it out , and a noticed a little improvement , I made a strange RTL , but the Quad landed , but the Loiter nor Pos Hold didn’t works at all …
I’m idiot , It’s happened just after I made the Update for the FW from 2.9 to 3.1 and I didn’t backup my settings , and like you I don’t remember the old values for certain parameters :frowning:
With my Quad sometimes the RTL or other FM don’t want to trigger , I switch from one to another but the Quad still Fly stabilised …
This behaviour came as it … For no reason … I didn’t test the GPS cable yet ,but GSC display the right GPS information , I shielded the module , this improve the Sat reception , for me from 2,45 to 1.1 and 0.9 for HDOP … I will test higher values for the IMU filter … Like you this param was downgraded to zero , and I don’t know why … I fear that other critical params were changed too …

I’ve the same problem guys, I’m using the Zeal gel for reduce vibrations on FC. I attached my logs files. Actually I have the 3.2 custom firmware that enables the UART port for telemetry. Default parameters as well. Any suggestions, please? I compare the parameters with another ones default files and i see differences but i don’t know witch the best option for my CX-20. Attached all files that I think you need. Thanks, guys!

macoces, your logging is too low. It needs to be set to nearly-all for testing so we have more information.

Mike

I just come back from the field , this ended with a crash and a prop broken .

I tested with the value 20 , then 42 , and still having the same issue …
I tested the cable => No problems all the wires are conductiv
I tested the GPS with an Arduino Board => No problems and a good 1.1 HDOP INDOOR !! no glitch !
I tested InFlight without any Vtx or 433MHz Telemetry Modem plugged => same issue !
I did a Total reset of the Board : EEPROM erase , Then Re-flash the Firmware , made the mandatory setup , including Compass Motor calibration => Still the same issue … :frowning:

In Flight :
As soon as GPS based mode is triggered ( Pos Hold / RTL … ) => Many overshoots , the Quad always drift in one direction ( N , NW ) , I look at the logs , but found nothing strange … Other than this Issue , the Quad fly as an eagle in Stablised mode , even better since the total reset , but all GPS based moves are totaly erratic … Is my APM dead ?
I will recieve another GPS module soon for another Multirotor , If I can , I will test this new GPS module with the APM and will see if it changes something , I fear of flying with my Quad now , because I don’t want to destroy it …
If someone read this , and can help us to solve these issue , you are wellcome :slight_smile:

Finaly I fixed the Problem .
I downgraded the Firmware to the 3.1.5 Revision , made mandatory setup like the day I recieve the Board , and all the problems was solved . To be able to downgrade , I use the previsous version of APM Planner : Mision planner 1.3.17 , and choose “pick old firmware” in the intital setup > install firmware , and choose the 3.1.5 Firmware :slight_smile:

Yes, I make a new installation of 3.2 custom firmware and then rest to default parameters and I’ll test it again with all UMI logs activated. Thanks guys. I’ll post it the conclusions…

Update :

When downgraded to the 3.1.5 FW :

  • Made a Parameter backup ( named : param-fresh-3.1.5.param )
  • Made the compass calibration : Ok
    => Flew 2 x 5000mAh LiPo = better GPS accuracy in Pos_Hold / Loiter / RTL
  • Made a Parameter backup ( named : param-3.1.5_COMPASS_OK.param )
  • Made a REAL* Compass Motor Calibration : OK
    => Flew 2 x 5000mAh LiPo = better GPS accuracy in Pos_Hold / Loiter / RTL
  • Made a Parameter backup ( named : param-3.1.5_COMPASS_MOT_OK.param )
  • Update from 3.1.5 to 3.2
  • Made a Parameter backup ( named : param-fresh-3.2.param )
    => Flew 2 x 5000mAh LiPo = Perfect LOITER , Perfect RTL : 3 time within a 50cm circle :slight_smile:

At the end there is one thing wich is different than the first time I upgraded :
I have got a 3 position switch to trigger Flight Modes , and I had

[quote]1/ Stabilized
4/ Loiter
6/ RTL[/quote]
Every time I trigger RTL , obviously the switch pass through LOITER to RTL when I switch to RTL from Stabilized ( IMHO ) , and I had GPS issue .
Now I have :

[quote]1/ Stabilized
4/ Auto
6/ RTL[/quote]
And the RTL is Perfect , I didn’t test LOITER nor Other GPS based , but I think there is an Issue when you change the Flight mode list , because as soon as come back to 1/ stab , 4/ Loiter, 6/ RTL => GPS issue .
For now , My Quad flys better than before the first time I upgraded to the 3.2 , but keep in mind that with some kind of GPS module , we have to wait a while to aquire sufficient number of Satelites to perform 3D Fix , and decrease HDOP value , I tested to fly just few seconds after power on the Quad : bad GPS accuracy , I landed , power cycle the Quad , and wait almost 1 mn AFTER the bleue led stay fix : Perfect positioning :slight_smile:.
I think after few days testing that I fixed the Issue :slight_smile: I’m confident now , I can fly in FPV :slight_smile:
I will test with another flight mode list like that :

[quote]1/ LOITER
4/ STABILIZED
6/ RTL[/quote]

  • When I said a Real COMPAS MOT calibration , I mean doing this WITH propellers :

Attach the Quad on the desk and make sure that it doesn’t move , start the COMPASS MOT calibration with the Real current drawn when you fly . and you will see the HUGE difference between the interferences made when the motors was not loaded , and the magnetic field create by the High current drawn when you FLY , if you do the COMPASS MOT calib without propellers , it like you did nothing :slight_smile: