Hello,
I have a 450 XQuad using an APM 2.6 Flight Controller and Neo6M uBLOX GPS. (Rotation is 180)
My stabilise and AltHold modes work beautifully. However whenever I put it on Loiter the Quad does its own thing. Starts Drifting Badly and even the controlls become sluggish. I need to quickly move back to stabilise to regain controll.
It is asif the quad is more locked in with AltHold then with loiter. Please help....
I have only tried Stabilise, AltHold and Loiter till now... Cannot get the loiter to work acceptable.
Can someone please help me by analysing these logs for me. Logs zipped and attached.
BTW - This is just day 2 with my APM 2.6 Quad. And I Just love it. Cant wait to perform an auto mission.
Thanks in Advance :-)
Arshish
You need to be a member of diydrones to add comments!
Replies
Ok Here we go. After Re-Balancing props & Re positioning the GPS this is what I am getting. Managed to get the vibrations much lower. Compassmot interference Vs Throttle is 10%.
Images Below.....
Have not tried a loiter as yet after the changes. Should I try now or is there something else I should test first.
Vibration:
Zoomed In:
Compassmot:
Compas Offsets
The telemetry Log is from when I ran the Compassmot with USB Connected.
Telemetry Logs.zip
Vibrations are ok, compass offsets and interference ok.
However, the telem logs provided is not exploitable. So the best thing to do now is to flight stabilize around one minute (with a gps 3D lock to ensure that it will be ok when you loitering) and check on the telemetry log that magfield vs throttle graph is good enought. Then you will have rock solid loiter.
I dont think I can get a Telemetry log when in flight because i don't have a wireless Telemetry interface.
"the telem logs provided is not exploitable" - What does that mean?
Any other way I could get the magnified vs throttle graph? Worst case i can hold the quad down for 1 minute withe someone else modulates throttle. Will that work? That is essentially what I did for the compassmot and what generated the telemetry log.
Cant fly today as I have a thunderstorm.... :-(
"the telem logs provided is not exploitable" - What does that mean?
I think it's due to compassmot procedure, there's no magfield data available.
"Any other way I could get the magnified vs throttle graph?"
Yes. You have to flip each props and rotate them (exactly same as compassmot), connect your APM in usb and arm in stabilize mod. Then raise the throttle at 50-75 % and hold it for ~5sec. Do it several times in order to have a reliable graph. Then disconnect telemetry :)
My To-Do list for today evening:
1. Prop & Hub Balancing & Recheck vibration
2. Run compassmot Test - Will need to be Compas Vs Throttle as I don't have a power module
3. Gabriel: Have you checked your mag (good offsets) - How do I do this? I have done the live calibration. Rotation Roll is 180. What else do I need to check?
4. Will Log & Post Telemetry Logs
Another Question About Loiter Mode:
My understanding of Loiter Mode is that it is Stabilize + AltHold + GPS Lock & Compass. As I had mentioned in my first post, my Stabilize & AltHold work pretty well. So in that case, if GPS & Compass is the only failing component in loiter mode then the quad should actually still be controllable but may drift. Correct?
In reality in Loiter mode my quad drifts badly and also becomes sluggish to control. It is as-if a ton of expo has added to all controls. It also appears to become less stable... Why would that be?
Hello,
You can find the mag offsets "COMPASS_OFS_X", "COMPASS_OFS_Y" and "COMPASS_OFS_Z" in mission planner full parameter list. They should be between -150 and 150 and are all set by live calibration. Moreover, "compassmot" interference must be below 30 % to ensure a rock solid loiter.
Please pay attention to do the live calibration far away from large metallic structure, outside. Compass should be also as far away as possible from PDB, motors and battery. You can twist the power supply wires together to reduce magnetic field.
In fact, since AC 2.9, the accelerometer is also used to estimate horizontal position (inertial navigation) in loiter mode. So having heavy vibrations can lead the copter to drift away. between +3;-3 on XY axis and -15;-5 on Z is borderline, +1;-1 is far better.
It's normal that loiter mode is sluggish since you only control the target position of the drone whereas in stabilize you control directly the angle of the drone. Therefore the hybrid mode will be introduced in AC.3.2, which hold position when no RC input and switch automatically to stabilize when position ajustement is needed. Good idea to use it...
Ok..
Save trim - yes.. done once. However I still have some trim dialed in. So this could be an issue. Will save these trims and try again tomorrow.
I dont recall pressing the wizard button to be honest... I strongly feel its whwn I work with logs that things mess up.. could it be linked to the GPS fail safe error I saw in the logs and mentioned in my other post??
Calling it a day now... its 12 AM..will save trims and try loiter tomorrow. Is there anything else I should try before that... loiter has been pretty scarry wheneverrive tried it.
Hello,
GPS Ncount and HDOP are ok, vibration level seems to be ok but X and Y axis are borderline. Have you check your mag (good offsets) and set up compassmot and obtain interference below 30 % ?
Could you provide a telemetry log of a hover flight ?
Vibration does not appear to be enough to cause this. What about the GPS number of sats and HDOP value?
Also, do you by chance have trim set on your controller? You can NOT have any transmitter trim set. Remember, Loiter and other position hold modes interpret centered sticks to mean "Hold position". Transmitter trim will prevent the pitch, roll, and yaw from every being centered at zero.
On that note, have you done the SAVE TRIM in no-wind conditions to properly trim out the aircraft? Use SAVE TRIM, not auto trim. It is much more accurate and faster. http://copter.ardupilot.com/wiki/autotrim/
Save trim won't calibrate the PIDs.
-
1
-
2
-
3
-
4
-
5
of 5 Next