Hi guys, I have a customer with strange level issue on APM 2.5... RMA or there is a fix for this?
He purchased two Arducopter kits . One flies flawlessly but there is a problem with the other.
I need your help to see what you can find in the logs...
Here is what he experienced: (I think they are on latest firmware 2.7.3 but not sure)
I can fly it and it is pretty stable, but I have trouble calibrating the level. If I calibrate the level using the Mission planner, the artificial horizon becomes straight on the display as expected. However, if I perform the auto-trim function, the level becomes out of calibration againonce I land. If I recalibrate the level, the auto-trim is lost again.
I logged a short flight I did this morning to illustrate the problem. The log is attached to this message.
Arducopter is immobile on the ground at the beginning of the log
15:52:24 – Arducopter level performed, Arducopter still on the ground
15:52:32 – Arducopter armed, I keep the Yaw level to the right to start in-flight calibration Red led blinks
15:52:45 – The Arducopter is in flight, it is kept almost still in the air until the end of the flight. As the flight progresses, although the Arducopter is still in the air the Artificial horizon slowly drifts and shows that the Arducopter is far from level.
15:53:13 – End of the flight, The red led stopped blinking, indicated that the calibration is done before landing
After that, the Arducopter is on straight, motionless the ground but the level is now completely off calibration, the artificial horizon even continues to move.
Again, I purchased two identical Arducopter 3DR kits and only one is doing this, the other works flawlessly and the assembly is identical.
Note 1: I am flying indoors in a large garage, without GPS fix.
Note 2: There was also a GPS communication problem with this board hat was fixed by resoldering the GPS header.
Any help would be appreciated.
Yes, I confirm that I am flying version 2.7.3.
Bump for some help? Chris what do you think? Guess will just RMA it and 3DR could have a look at what's wrong
Hey Dany, no real help but some thoughts...
Vibration (motor bearings, unbalanced props, etc.). Pulse motors and watch slow-down behavior. Hold the frames (to compare) carefully with motors near hover/level speed to get a feel. I read elsewhere that these accelerometer units can be fussy about vibration. Since the owner has deux identical frames, the first thing I'd do to eliminate motor/ESC/frame/etc. issues is to swap the good APM into the "bad" one's frame, no other changes. He's probably already done that. I don't really know how to read/analyze logs properly yet. I'm out...
So i had a look at the logs..I wonder if you're actually engaging the auto trim feature. I think you are because I am seeing the accel_cal_x and accel_cal_y values change in the logs but what's confusing is that the raw pwm value for channel 7 barely changes. When you calibrated your radio did you calibrate channel 7?
Here's a graph of your roll and pitch and I think I can see what you're complaining about..that the pitch keeps changing pretty much from the moment you take-off until a bit after you land.
If it's a hardware problem, it's not a super obvious one because your accel and gyros are producing values and they don't look too crazy to me.
I guess you've gone into the cli menu and done a setup, reset and then rebooted your APM? This will reset any parameters that might be causing the problem. For example a very low AHRS_RP_P value could cause some trouble.
Could you also try cranking up your AHRS_RP_P value to 0.4 (the default is 0.1).
By the way, there's a wiki page describing how I use the auto trim feature here. Is this the way you're doing it?
yeah auto trim ch7 is really "sensitive" and really hard to set properly for beginners.
That is why I ask my customers to instead do the arming +20 sec to get 45 secs of "in air trim"
I think that is what Chuck was doing here.
he did the same thing to the 2 other APM 2.5 he's got and all fine (with same frame as this one)
Ah right, i forgot about that way to kick things off. Way down on my list is an item to rewrite the autotrim so that you do the trim on your radio with the regular little trim tabs..then you land and click the ch7 auto trim switch and it then uses those radio trims to calculate the accel offsets it needs. you'd then remove the trim from your radio and fly again to make sure it's worked.
There is a little issue (not a bug exactly) in 2.7.3 (fixed in trunk) which meant that the ch7 trim would take longer to feed through to the attitude solution than in earlier versions. So that could explain why the attitude keeps changing after he lands.
Hello, the same thing happens to me. I make my trims, then using ch7 for saving them. And now, my level in mission planner has changed... This is normal ? If not, did you find the solution ? Thanks
if you use save-trim it definitely will change what is level in the mission planner. That's what it's suppose to do. Basically when you do auto-trim or save-trim you're telling the APM what angle it needs to lean at to fly level. So when you put it back on the ground again it might not be straight. That just means that your copter is a little mis-balanced in some way or perhaps your APM is not level on the frame. I wouldn't worry about it too much though unless it's really far off (10 degrees or more).
yep, true test is in the air, when you fly does it drift? If not no worries.
Later on you can revisit the way you mounted the APM and trying to make it level to the copter.
But really the ground level in MP is not that important, it's not a rover, it's a copter! :)
Thank you both ! You just answered my question : trim has an effect on the level in mission planner =]
My TBS is not very well balanced cause of my battery weight on the rear side and lack of weight on the front (no camera for the moment) It was flying backward. And I have to trim pitch a lot. After the quad was quite stable but there was offset pitch (5 degrees) on the MP level when on the ground.
Thank you =)