Testing APM - Loiter and RTL Drift South Continuously

Hi, I am testing my APM and for some reason when I switch it into Loiter or RTL it will continuously drift south until it is on the edge of my comfort zone and I take control again (300ft maybe). 

The first circle of Loiter and RTL look good but the second is a little south and after 4 or 5 circles it way off and doesn't seem to be heading back.  It does seem to be able to "fly" the plane well (PID settings seem right) Stabilize and FBW-A work great.  It just seems to have a problem with navigation like the GPS points are moving south or something.

The wind was blowing south about 8 MPH but the plane didn't have any trouble flying up wind when told to manually and was making nice circles when in Loiter or RTL they just didn't keep centered on the right location or even return to the correct center once the got a long way south.

I am flying a basic Bixler with the APM mounted in a Bixler Chassis. I did have a camera on there too for some added weight but Stabilize and FBW-A didn't seem to mind the weight.  

I have noticed that when on the ground and hooked up to an XBee (my xbee has very very limited range) that the compass seems to be off.  It has the plane pointing in the almost opposite direction.  I have checked many times to make sure the APM (new purple v2 board) is pointing in the correct orientation in the plane.  Could the Magnets from the chassis be affecting the compass?  could an incorrect compass cause loiter to continuously drift south?

THANKS A LOT if you help!

Views: 978

Reply to This

Replies to This Discussion

Did you calibrate your compass in the Mission Planner?

I did enter/calibrate the "declination" but that is all I remember, would there have been something else?  THANKS for the help.

Is the calibration something that needs to be done? Or can you put in the declination and it be good?

It's best to do the "Live Calibration", too. The MP will instruct you on what to do.

Chris, THANKS A LOT for helping.  I don't always get responses from the forums so I really appreciate it when I do.

I entered the Declination but I believe I did it manually.  I will try the "Live Calibration". 

Could (would you expect) the compass being off to cause the navigation control to send the plane further and further south on each circle?  Could that be my navigation issue or do you think they are two separate issues?

So I ran the calibrate procedure a couple of time and got the compass to work better a little but never very reliably.  I think the magnets that are attached to the chassis and to the plane are throwing it for a loop.  It doesn't surprise me that magnets can effect a magnetometer, but what does surprise me is that other people aren't having this problem with the magnetometer and the Bixler chassis.  I do have an extra set on there so there are a total of 6 magnets on the chassis and 6 on the plane which lock together holding the chassis on very very well...but I guess by doing that I lose my compass.

Does the Magnetometer affect navigation in the way I described?  What effects will I notice with the compass not working properly?

I think a more likely explanation is that the prevaling wind at that altitude is towards the south and you need to turn up your gains to get more navigation authority. Did you load the Bixler config file? That should be a pretty good start, although it's always a good idea to go through this first-time process to tune the gains for your own aircraft.

You could try flying with the compass disabled. It should still navigate, just not in very straight lines, but it should hit the waypoints reasonably closely.

If the magnets are strong enough the compass will read pretty much the same no matter what the plane's orientation.

Yea, the wind was blowing to the south and a bit stronger than other days I have tested (maybe 10 mph). I did load the bixler config file and read through the manual files but when the stock bixler configuration worked so well for FBW-A and Stabalize I didn't go any further with tuning and maybe I should. I did have a camera on there which would add a little weight but other than that it should be the same as every other Bixler with chassis and AMPv2.

What you suggest sounds very likely and I will look into "turning up my gains". I guess I got confused when the airplane kept circling. I guess I thought it would not circle but head straight towards a loiter point or RTL point if it got to far off due to wind or whatever.
Phil, Thanks for the advice I may try that too. I did notice that my compass doesn't turn much at all.
Ok, so while on the ground for all of this testing...I tried the calibration procedure many times while APM was in the plane and got wildly different settings and no completely funtional compass results.

I took the APM out of the chassis containing the magnets and away from the plane with the magnets, ran the calibration procedure and got excellent results and responsiveness from the compass. I brought the chassis close again and the reading skewed without moving the chip and then was no longer accurate or very responsive.

So... I guess my first step is to remove the magenets from the chassis and hold it in another way as to not interfere with the electronics. My second step is to fly again and determine if my gains need to be increased in order to give more navigational authority.

The only question I still have is if the winds are to strong for my "gain settings" and I don't have enough navigational authority will the plane still circle and start drifting down wind while circling or will it turn up wind (towards the waypoint) but just not reach it?

Magnets or any large metal mass right by the compass will definitely make it unreliable.

When I move my Quad when its hooked up to my computer, I get a 15 to 20 degree skew when its setting next to the computer. I input the declination from my location using the web site and that works fine, but magnets near the "compass" in flight will absolutely not work and those little canopy magnets they use are extra strong high Gauss rare earth magnets.

I'll bet a lot of people are having this problem and don't even know it.

By the way there is one other possible cause of drift like you mentioned. If you have any trim on the roll or pitch sticks it will continuously displace the GPS point in the direction of the trim.

Because when in any auto GPS mode, the sticks move the GPS point not the plane so any trim acts as an offset which does exactly that.

It is imperative to set up your APM so no transmitter trim is required.

Reply to Discussion


© 2020   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service