Copter started to "wander" after about 5 minutes flying with heavier battery

jstroup1986 · March 8, 2025, 12:47am

I upgraded the 4S 2800 mAh LiPo in my copter to a 4S 5500 mAh LiPo - and did some test hovers in my back yard this afternoon.

The copter with the 200g heavier battery seemed to fly quite well with the existing tune - for about 5 to minutes.

At that point the copter had a tendency to fly off or “wander” from the point where I’d placed it using Loiter. It wasn’t severe - but it did require me to constantly stay on the sticks. I landed to check for problems - and took off again - same thing right away.

The wandering reminded me of a beginning of a “toilet bowl” effect. With the amperage to hover increased from 8A to 10A, I considered my compass calibration. I did a new MAGFIT after landing - and there was very little change except to one of the MOT offsets. (The Y I think.)

The added weight brings MOT_THST_HOVER up to within the default range. (just barely) I’ve adjusted PSC_ACCZ_P and PSC_ACCZ_I with this new value. I’m not sure if this has anything to do with the “wander.”

I’ll do another test flight tomorrow at a larger site - my backyard has too many trees to let the copter wander far.

If anyone has any suggestions on things to check before my next test flight, I’d appreciate knowing them. Thanks!

xfacta · March 8, 2025, 3:07am

We can definitely see the variations in the position, but vibrations are not bad.

There is a physical yaw bias from twisted motor mounts or frame twist. This is not drastic, but when yaw compensation is constantly required it will end up affecting pitch and roll at some point. It’s possible this is at least part of the problem.

You can see on the map how much there is variation between the IMU-calculated position and the GPS position - quite a lot!
Try GPS_GNSS_MODE,5 to make the job a bit easier for the GNSS receiver, but you are going to have to get out from under those trees.
There’s a good chance that previous good flights were just fortunate (some conditions giving you better GNSS signals) and that now the poor flight has nothing to do with the battery swap, but changed conditions that are not obvious.