Quad rolls to the right

I have managed to get my quad somewhat up and running today, but I still cannnot get anything even close to stable flight without roll stick input. It's a stock arducopter 3dr kit running APM1 with an atmega2560.

Takeoff requires about 20% left roll input, and to hold a stable(ish) hover I need to keep adding an increasing amount of left roll stick input. If I let go, it will roll sharply to the right. I could get the quad to hover today for around a minute at a time, but it seems when it's in the air I will need to add increasing amounts of roll for it to stay in the same place, until I have my roll maxed out to the left on the stick, and the quad still starts drifting right... that's when it gets put back on the ground. Landing it and taking off again seems to reset the amount of roll input required back to about 20%

The Turnigy 9x I'm using has the trims zeroed, I have verified in the mission planner (radio test, and via raw flight info data) that when the sticks are centered that it is sending the correct pwm for roll being centred.

Mission planner seems to show the horizon pretty much centred when the motors are armed, and also if I plug the quad back in and connect to it via mavlink it will show the horizon pretty much flat. I have a couple of logs from today's flights (attached), but I can't figure out how to find out to find out if the quad thought the horizon wasn't level, or if something else was going on. It would be handy to see how the ch1 input corresponds with roll, but I can't figure out how to do this. I only see the ch1,2 and 4 output in the logs. Can someone who knows what they're looking for please take a quick look?

I saw someone else with this issue solved it by running the erase command in the CLI and setting up again, so I'll try that in the morning.

2012-01-04 05-34 6.log

2012-01-04 05-33 5.log

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –

Replies

  • I've tried erasing the eeprom and setting up from scratch again, but still have the same problem.

    Hand testing will start out approximately flat, and over the course of about 40 seconds, the quad will end up on about a 45 degree angle to the right.

This reply was deleted.

Activity