Page 1 of 1

stability of the LPS system

Posted: Tue Jan 28, 2020 3:31 pm
by yousba
Hello,
we are a group of students working on ordering a CRAZYFLIE using the positioning loco DECK.
when trying to fly the drone with the scrypt python "autonomoussequece.py" with a very simple sequence [(0,0,0,) (0,0,0.8) (0,0,0)], we observe that the drone has a completely unstable behavior (it leaves in all senses).

we have already updated the CRAZYFLIE firmware(using the bootloader method in cfclient), anchors and CFclient.(https://github.com/bitcraze/crazyflie-release/releases)
we have already try to activate the kalman filter in config.mk, but it doesn't change anything.


best regards

Re: stability of the LPS system

Posted: Wed Jan 29, 2020 9:04 am
by kimberly
Hi!

Few questions:
  • Is the Crazyflie facing in the x direction of the LPS system
  • Have you tried holding the crazyflie in your hand, move it around and check with the cfclient if the positioning makes any sense?
The kalman filter should actually already be activated automatically if the LPS deck is selected. Have you checked the console of the CFclient to check if the LPS deck is being detected?

Re: stability of the LPS system

Posted: Sat Feb 01, 2020 8:56 pm
by yousba
Hello,
thank you for your reply

yes we have followed all the steps given on the site : https://www.bitcraze.io/getting-started ... ng-system/ and https://www.bitcraze.io/getting-started ... tion-hold/

and we placed the drone in the center of our benchmark. but the problem is that when we execute a simple script the drone it becomes completely unstable

Re: stability of the LPS system

Posted: Mon Feb 03, 2020 11:27 am
by kristoffer
Hi!
Can you please attach a screen shot of the LPS tab in the client?

Re: stability of the LPS system

Posted: Tue Feb 04, 2020 12:52 pm
by yousba
thank you for your reply

here they are the screenshot of th LPS tab and the console in the cfclient

Re: stability of the LPS system

Posted: Tue Feb 04, 2020 3:32 pm
by kristoffer
It looks like the anchors are OK (assuming you are using 6 anchors) and they all are in TWR mode.

The next thing to check is that the anchors actually are at the physical positions that they are configured for, to make sure you have not accidentally mixed up two anchors.

Finally check that you place the Crazyflie on the floor, pointing in the direction of the X-axis when turning it on. This is very important since the LPS system can not measure the orientation and simply assumes it is pointing in positive X direction when started.