GPS Loiter Mode - quad rises steadily - ArduCopter User Group

Hello everyone. I'm new to this forum and new to the APM 2.5. I need a little help please.

First off, here is what I'm flying:

Frame: QAV500

FC: APM 2.5 w/software v2.9.1

GPS: UBlox LEA 6

Motors: Turnigy 2217/16

Props: APC 10x4.6 SF

Lipos: Zippy 3S, 5000MaH

I tuned my quad on a stand and it flys fiarly well in stabilized mode and in Altitude hold mode. My throttle position for hover is right at 50%. When I switch from Stabilized mode to Alititude Hold mode, the quad responds as expected in that it pretty much holds it's alititude with only some slight deviation. If I add throttle, it rises, if I reduce throttle, it descends. However, when I click in to LOITER mode, the Quad start to ascend at a moderate rate.

Before I fly, I power up my radio, then plug in the lipo on the quad. I let it sit on the ground until the blue light on the APM turns solid blue. Then I arm the motors and take off.

Any ideas on why the quad ascends in LOITER mode only? I have checked the forums and the troubleshooting regarding the quad ascending in ALT HOLD and/or LOITER mode but it does not seem to be my case here. Again, when I click in to ALT HOLD mode, the quad pretty much does hold its altidude fairly well. As soon as I click in to LOITER though.. it's up up and away..

Thanks in advance.

S.Schmick,

Loiter actually uses the altitude hold controller to maintain altitude so it's weird that they're different. Two possibilities are:

you have a lot of vibration on the x/y axis and the inertial nav is getting confused about the altitude when you lean over. You could prove this by going into alt hold mode and then rolling or pitching over to some reasonable angle like 20 degrees and see if it does the same climb as loiter. The solution would be the same as for solving the regular alt-hold issues people face with 2.9.1 - namely more vibration dampening, fixing bent motor shafts, setting INS_MPU6K_FILTER to 20, etc.
the loiter controller pids have been increased a lot and it's becoming very over active so it's constantly switching the desired roll or pitch output a lot. This could expose an issue with the "stability patch" in which it always prioritises roll/pitch over altitude and means it can actually increase the throttle higher than you'd expect if PIDs are much too high. I'd say this is a very unlikely cause unless you've been changing the loiter PIDs.

Replies

S. Schmick March 14, 2013 at 4:18pm

PROBLEM RESOLVED - Operator Error. I thought I had my third switch position set to LOITER but in fact, it was set to RTL. My mistake. Now I feel stupid. lol

Thanks for the reply Randy.
Developer
Randy March 14, 2013 at 2:06am
S.Schmick,

Loiter actually uses the altitude hold controller to maintain altitude so it's weird that they're different. Two possibilities are:
- you have a lot of vibration on the x/y axis and the inertial nav is getting confused about the altitude when you lean over. You could prove this by going into alt hold mode and then rolling or pitching over to some reasonable angle like 20 degrees and see if it does the same climb as loiter. The solution would be the same as for solving the regular alt-hold issues people face with 2.9.1 - namely more vibration dampening, fixing bent motor shafts, setting INS_MPU6K_FILTER to 20, etc.
- the loiter controller pids have been increased a lot and it's becoming very over active so it's constantly switching the desired roll or pitch output a lot. This could expose an issue with the "stability patch" in which it always prioritises roll/pitch over altitude and means it can actually increase the throttle higher than you'd expect if PIDs are much too high. I'd say this is a very unlikely cause unless you've been changing the loiter PIDs.

This reply was deleted.