Level automatically adjusting in flight...

I have an odd problem that just baffles me to no end.... I know this is not in the exact correct place, but I have a slight bit of pressure to get the quadcopter to tune-able conditions. I'm in a group working with autonomous vehicles as a project in college, and we changed frames from the older Jdrones frame to a hobbyking SK450, which is really bringing out some odd problems that the larger frames covered up. We have an arducopter APM 1.4, 850KV jdrones motors, 3S 2650 mha lipos, 3DR radio, magnometer, GPS, and a Turnigy 9x. First, we had some major vibration and balance problems, but with 9 years of rc airplane and heli experience, nothing too hard to take care of. Upon takeoff, the quadcopter would tip hard and fast back and too the right. After balancing, the flight characteristics were significantly better. However, we still had problems with back and to the right, but it was different. It was more of an exponential thing, and the quad would start stable, almost perfectly, then gradually move back and to the right, and requires full forward-left control after a while to keep level.

So, I went through the list. I tried leveling, calibrated the ESC's, re-balanced the props, checked the level of the motors, checked the motors for bent shafts, tested every esc and motor individually, checked all the joints, reloaded the firmware, changed version of the firmware, checked the parameters, checked them again, and nothing seems to make sense. Yesterday, I found something out that makes perfect sense and no sense at the same time. I leveled the quad in the mission planner, and flew, and the same behavior still existed, so I landed. I looked at the HUD, and on the same floor, same exact spot, the HUD read 15 degrees down and 15 degrees left. And, after a few seconds, the HUD returned to looking perfectly level. It happens, flying or not, I can manually tilt the quad, to the same effect. This problem seems to persist across all firmware from 2.6 - 2.8.1, so my question is, why does this happen, and how do I fix it? Any help appreciated, this one has me at a real loss.

Views: 1807

▶ Reply to This

Replies to This Discussion

Permalink Reply by Christopher Cooper on December 5, 2012 at 2:51pm

I won't pretend to be intimately familiar with the hardware or firmware here, but I doubt this is related to any of the PID settings.

As I understand it, the artificial horizon only presents information that it receives from the gyros and accelerometers. This information forms an input to the control system, and the PID terms form the control response.

The fact that the artificial horizon is drifting and generally incorrect leads me to believe this is a hardware problem.

Permalink Reply by avionics on December 7, 2012 at 10:35am

 

I am not an expert but when  iloaded your tlog files i dont see your satelite fixed? do you have satelite fixed? and you see 3D fixed on your Mission planner? have you tried to test it when the sat is fixed?


3D RoboticsPermalink Reply by Guillermo Romero on December 9, 2012 at 12:44pm

mmm... switch your props, check if the problem switches too. 

If so, get harder props, APC or CF.

Permalink Reply by Austin Suhler on December 9, 2012 at 1:29pm

I tried eliminating all vibrations, including mounting the board to a rather large kitchen sponge, to no avail. Any thoughts? This log is with the board mounted on the sponge, and not quite in the air, just on the ground on the edge of taking off. 

Attachments:
Permalink Reply by Will Snodgrass on December 9, 2012 at 3:56pm

Vibrations would be random, this clearly is not. Also since this system depends on immediate sensor input to work in the first place, the relatively low resolution GPS signal are not going to affect it. The next logical step is try a different sensor, since this really looks like a case of  malfunctioning sensor.

That's also where the problems lies. The whole APM projects is only here because of the high availability of low cost sensors that has been caused by the explosion of the smartphone market. But with no easy way to replace or add the sensor in question, we kind of hit a wall. Although of all the APM'S, you have the one that is easiest (and cheapest)  to change out the IMU board. Do you know of anyone else that would let you swap out the IMU for  a bit? Because that's the next logical step.

Permalink Reply by Austin Suhler on December 10, 2012 at 7:28am

Unfortunately, while we do have another IMU board, it is currently down, and most likely dead, due to a USB issue. The current hardware seems to be working, just with some incredibly odd behaviors. What would you recommend as a course of action from here? Just a full replacement/upgrade to 2.5, or any other tricks for me to try? I have swapped props, triple-checked them, replaced a motor already because of a bent shaft, etc. It flies well in acro, so it isn't a hardware with my power/radio system. Do you think it could be a problem with the sensors being dirty or having residue built up on them from soldering within the general vicinity? We have done a significant amount of soldering in the last month or so, from redoing the Jdrones V1.0 kit to upgrading frames. What is the professionally accepted way to clean a board like this so I can rule that out? At this point, I don't quite know exactly what's causing this problem, and it's quite baffling. I had a few engineers in the UAV field look at it yesterday, and they came to relatively the same conclusion that it was tentatively a hardware problem. I don't even have the GPS or the compass enabled, or mounted for that matter, at the moment, figured that taking out any possible source of extra complexity would make it easier to debug. 

Permalink Reply by Will Snodgrass on December 10, 2012 at 4:15pm

I'm just speculating here, but I believe that sensor is a sealed unit. Heat can cause issues for pretty much all electronics, but it's impossible to say whether or not that may have caused the problem. As far as cleaning, I would imagine using electronics parts cleaner would be the way to go, although I would check with someone from 3DR before actually going ahead with it........But again with a sealed sensor, unless you have something causing a short it's probably not going to help.

Of course the best way to go is with a new 2.5+, especially since you will get the new dedicated power supply, and at $159 (without GPS) it's a pretty good deal......not to mention the new case it comes with.

Permalink Reply by Austin Suhler on December 10, 2012 at 4:18pm

I was looking at exactly that same board earlier, and I think that's the path I'll end up taking. The GPS that came with the 1.0 kit build should still work with the new 2.5 board, yes? 

Permalink Reply by Will Snodgrass on December 10, 2012 at 4:25pm

Yes, but if your GPS is that old, then for only $20 (when bought with the APM) the new board has some nice features, including a battery backup on board for quicker GPS locks and a  larger circuit board.

Permalink Reply by Austin Suhler on December 15, 2012 at 7:37am

Thank you very much everyone for your help! It looks as if the school has put enough money into the project that they are willing to pay for the APM 2.5 instead of grounding the project. With any luck, after I come back from break, there will be a new board waiting for me to pick up, and I'll keep you posted! I will most likely be back on here soon, we have a plane that we're going to help coding for a competition using the paparazzi autopilot system. With any luck, we hope to have the quadcopter and plane to communicate with each other to direct each other towards targets they identify by post-processing video to accomplish missions. That should be about a year away, though. ;) Thanks again for the help!

        -Austin

Permalink Reply by Jimmy on December 15, 2012 at 8:16am

Hello,

I've the same problem with my hexa. and I try differents solution, but nothing 's good yet.At this time I think it's the cold temperature which create this phenomene.

Permalink Reply by Austin Suhler on December 15, 2012 at 8:18am

I thought about that at first, however, most of my test-flights, in fact all the ones I uploaded logs for, were tested within the confines of my dorm room. 

RSS

Social Networking

Top Content 

1 ArduPlane 2.73 released

ArduPlane 2.73 released
2 HD video for UAV's

HD video for UAV's
3 Mapping Mission Ride Along

Mapping Mission Ride Along
4 APM Copter 2.9.1.b has been released

APM Copter 2.9.1.b has been released
5 Arducopter 3Dr or Phantom

Arducopter 3Dr or Phantom
6 Mechanical Hands From A MakerBot: The Magic Of Robohand (From Makerbot blog news)

Mechanical Hands From A MakerBot: The Magic Of Robohand (From Makerbot blog news)
7 3 axis Camera Gimble connections.

3 axis Camera Gimble connections.
8 ArduVTOL - Wingcopter - Flight Report Nr.1

ArduVTOL - Wingcopter - Flight Report Nr.1
9 ArduVTOL

ArduVTOL
10 Alt_hold and loiter acting crazy

Alt_hold and loiter acting crazy

Contests

Season Two of the Trust Time Trial (T3) Contest has now begun. The fourth round is an accuracy round for multicopters, which requires contestants to fly a cube. The deadline is April 14th.

A list of all T3 contests is here. 

Groups

Advertisement

© 2013   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service