I've recently put together a quad with DJI FlameWheel 440 frame, ESCs and motors and APM 2.0
I have few problems:
1. in Stabilize mode it's extremely difficult to hold altitude: when I slightly increase the throttle to avoid crashing the copter continues to ascend and even small change causes it to descend and it bounces up and down. I tried changing the controls but they seem to have no effect so if anyone you guide me which ones to pay special attention to, that would be great. (note: in loiter mode it's much more stable).
2. the copter also seems to have problems with holding position despite solid blue GPS light. I managed to put it to stable hover in stabilize mode and then switched it to loiter mode then without touching the sticks observed what's it doing - the copter was drifting to random positions within 15-meters radius - I would expect better from GPS (note: it was open space and no wind at all).
Any help would be very much appreciated
Tuning throttle is for all modes. But loiter is an autopilot mode, so throttle will be controlled automatically to lock in an altitude and location.
Increasing P will make that control more sensitive, decreasing will make it less. It's basically the amount of power applied to get to the motor speed that your stick is asking for.
Brilliant! Thanks very much!
We have a similar problem ~ 1 click and the copter goes up, one click down and the copter goes down - we changed the throttle P from 0.250 (default) to 0.0 and it made no difference. Maybe I should go lower ?
Thank you all for the discussion above. For shure many of my kind of pilots, pedestrians all over the world, listen and learn.
Get a GPS lock while your USB is hooked up, if your experience is similar to mine, you can watch your Quad in the mission planner flight data section wander around and make periodic excursions even with 9 satellites showing.
I believe that GPS inaccuracy is at least partially responsible for this problem.
As for the first problem I would like clarification that the Alt Hld Thrust / Throttle PID is actually used during non automatic mode operation: EG Stabilize and if this is true why is it prefixed with Alt_Hld.
I've never had problems with my GPS location jumping around. Even indoors, on the top floor of my house, with big windows, I don't have that issue.
I should clarify about the throttle PID. I should be more specific.
It's used in get_nav_throttle() called update_throttle_mode() for the THROTTLE_AUTO case. This means that it's used in any of the guided modes(loiter, position, etc.). This would not include modes like stabilize, were the throttle is manual and in alt_hold, it's not used, since alt_hold has it's own PID.
Thank you very much for your response Ellison,
Thank you for clarifying throttle, I thought stabilize was just an implementation of the stabilize rate and angle PIDs mixed with throttle stick position for maintaining and adjusting altitude and didn't actually have a separate PID for altitude or thrust control and I believe youve confirmed that.
Specifiaclly the only PIDs that affect stabilize mode are the Stab Rate and Stab ang PIDs (OK the Yaw PIDs too, but only for Yaw). Please confirm.
As for the other issue.
With 9 satellites showing on a solid GPS lock and my quad sitting on my living room table, I watched it go around the neighbors property, come back, hang out over my living room table for a little while and then move about half a block in the other direction and sit in the middle of the road and stay there.
Maybe some kind of indoors interference, but always showed 8 or 9 satellites and this is not dissimilar to my experience in loiter.
I will try this hooked up outdoors and see if it continues.
Thanks again for response.
Same problem here: my gps dances around whole neighbourhood when indoors and I suspect this may be source of my problem with copter doing chaotic dance in loiter mode. Is there any way to fine tune it, or is it because of how the gps is mounted (maybe interference with ESCs etc?)
I'm currently on holiday overseas but have alreay ordered the bluetooth kit and will try to connect it to Android version of the planner and will see if changing any of the PIDs will make visible difference to how it behaves.
Another post showed interference with motor power wiring which improved greatly when APM was relocated away from the wires. (Even more so for magnetometer).
I put an aluminum tape shield on my fiberglass intermediate separator plate which is between the Power Distribution plate and my APM and saw an immediate improvement.
GPS lock had been taking 5 to 8 minutes, now under 2 minutes.
Also, some of the problem appears that it might be from GPS Sat signal attenuation when indoors.
A final note, any radio absorptive dome material can definitely cause problems.
Black plastic usually contains free carbon particles as pigment and free carbon definitely absorbs microwaves eg:GPS Sat signals and will attenuate them.
A few other non pigmented plastics also absorb radio waves and this is not something they talk about much. Plastics are all primarily carbon but how the carbon atoms are arranged determines whether they are radio absorptive or not.
I just built a 540mm frame, ntm 26-28a motors 30amp esc. I get a perfect gps lock and can not fly in stabalized simple mode..