Two issues with position hold.

Big Ben said:
OI Photography said:
This is purely speculation on my part, but it seems perfectly normal to me that you'd see the yaw effect from that type of stick input.

Yes... if it weren't for the fact that the NAZA is supposed to have a bit more 'brains' than a spinning coin and is supposed to actively keep its orientation with the help of the many sensors it has, one of them being a compass which should give it the possibility to 'stick' to a given magnetic course until the pilot tells it otherwise. Otherwise being left stick rudder input. In the absence of such input I expect it to maintain its compass orientation.

The only thing I can figure is that compass orientation not being a precision measurement, if it constantly corrected for heading, it might jitter all the time, yawing left/right constantly. Add to that all the electronic equipment that these things carry that can also interfere or cause "noise" in the compass heading, and I bet they were forced to build in some slack where the system has to take an average over time to determine heading and possibly even update the preferred heading.

So the system may average multiple readings over time and even "recalibrate" the actual direction you are "supposed" to be heading. If it is making those calculations while you are grinding on the right stick like you're making a milkshake, some of the calculations and averaging might get confused and the system might actually update your "target" heading because it thinks it's getting erroneous compass readings: and it's trying to correct those. It does seem like something that could probably be fixed in firmware though: like stop certain heading calculations if quick motions are being made on the flight stick(s). Or just never update preferred heading unless the left stick is moved. Of course, this stuff is quite complex so I'm not sure what else that might affect (potentially adversely).

Curious though, what do you think was the primary problem that got fixed for your initial flight issues... what was the cause do you think?

Mike
 
Difficult to say without extensive knowledge of the inner workings of the NAZA/its firmware. My best guess is that quite possibly post-factory a firmware update was done by the German importer/distributor but without loading the default settings afterwards which left somehow interfering values in memory. The instruction to load those defaults after a firmware update surely is there for a reason. Once I completed the procedure by doing just that things appeared to work as intended.

Because of my problem I had been very sensitive to/aware of any yawing motion and so I noticed the 'directional drift' when doing the circular stick movements which I just did as part of testing the Phantom and seeing how it reacts to all kinds of input. I see no reason why the Phantom wouldn't be able to hold direction unless it receives rudder commands. I hope DJI reads this too and will fix this in a future firmware update.
 
Big Ben said:
OI Photography said:
This is purely speculation on my part, but it seems perfectly normal to me that you'd see the yaw effect from that type of stick input.

Yes... if it weren't for the fact that the NAZA is supposed to have a bit more 'brains' than a spinning coin and is supposed to actively keep its orientation with the help of the many sensors it has, one of them being a compass which should give it the possibility to 'stick' to a given magnetic course until the pilot tells it otherwise. Otherwise being left stick rudder input. In the absence of such input I expect it to maintain its compass orientation.

I see your point, and that probably is a reasonable expectation, I just expected the kind of net force that comes out of that kind of unusual stick input might be outside the range of what the NAZA can completely compensate for.

Another test that could point to that is (while flying) to give forward pitch only for 5 sec (no roll), return the stick to center, then right roll only for 5 sec (no pitch), and then repeat that a few times. The Phantom will/should fly in a zig-zag pattern, but after several repetitions of that the forward path may not be quite be parallel to what it was originally, due to some net yaw to the right....even though the Phantom "should" maintain parallel flight paths all along.

All in all though I get where you're coming from, and I'd also be looking for every little inconsistency in that like you have been.
 
Keep in mind too that even DJI makes better IMU's and GPS units than the NAZA has, so you may have just stumbled across a limitation of that non-top-shelf hardware.

If the skies clear up enough today/tomorrow I'll run your "milkshake test" and see what results I get on a Phantom that doesn't have other yaw issues.
 
I don't think it's torque wind up as such. I believe it's more to do with the fact that when you bank left the craft loses some lift on that side. As you rotate it continues that slipping in the direction you rotate the stick. You can see the effect this has at the end of a straight line run at full speed(full speed then reverse direction. Make sure you have enough height before trying this). The aircraft will dip in height as it stabilizes.
I'm not sure I'm explaining this very clearly I hope it makes sense. (6 am and no coffee yet)
 
Unfortunately my problem(s) do not seem to be completely solved. Saturday I flew on another location again with a pretty stiff breeze so I kept it low. One time when hovering it suddenly rotated 90 degrees left. Another flight it did about a 70 degree turn left. Sunday I again loaded defaults and did all calibrations and generally it behaved but I saw a few small spontaneous rotations of perhaps 10 degrees each, strangely also one to the right this time. If it never would go above that I could live with it but I cannot be certain it never will.

On top of that it has serious troubles with (constant) rotation in CL. When I do a pirouette in CL it goes off course about 25 degrees and it even loses altitude. I did some last tests today close to sunset after another compass calibration in the field and then I discovered it seems to hold course much better doing such a pirouette move in HL which could be of some significance. It suggests to me be that in HL it might fly its course purely on GPS information and not use the compass while in CL it does use the compass (too), and that is when it cannot make a properly coordinated flight because somehow it cannot properly mix the compass input with its other sensor data.

I remembered a video of someone doing this too in CL and it looks like its moving in a straight line much better here. Unfortunately you cannot see what right stick movements he is making or not making.

Have others here have experience with such a pirouette flight in CL? I'd be very interested to hear about it.

The problem is that with only one Phantom at my disposal I cannot adequately compare the flight performance of my bird with others and with a sample size of one I am unable to determine how close or far from the norm my Phantom performs. I have a period of 14 days in which I have the right of return no questions asked and tomorrow the 18th is the last day of that period. After that the web shop can fall back on regular warranty which could mean much more hassle for me. I have to decide whether I keep it or return it and get another one.

If people would be able to try such a pirouette flight in CL and HL within the next 12-16 hours and report back I would be able to use that info before making my decision. I will try to make a few final flights tomorrow morning and compare the CL and HL performance.
 
Have you checked to seee if your phantom has dji propellers on it? i bought what i thought were dji props and they turned out to be XOXO's and my phantom would do all kinds of crazy things... New dji props, and flies like a dream.
 
Ben, have you "upped" the gain setting on just the YAW? I had similar issues awile back and raising just the yaw cured it. Just a note on gains - the upper 4 gain settings are how the aircraft reacts to wind,surroundings, etc... the atti gains (lower 2) are how the aircarft reacts to stick inputs.
 
This sounds strangely familiar to prop balancing, third party prop vendor problems I have read elsewhere.
Did you try another set of props? Sounds exactly similar to other members trying everything under the sun, even returning the units only to find the problem being props.
 
Hi,

I had the same problem of the Phantom rotating when I first got it. I was advised to more aggressively take of to a height of 2 metres then let it sit for a few seconds. Ever since then its been straight as an arrow.

Mine also drifts while yawing with left stick, but as long as I don't touch the right stick it will recover to its original position once yawing is complete.

The Phantoms seem to temporarily loose gps while yawing. If your bird keeps yawing then this is way you could be loosing lock.

Damien.
 

Members online

No members online now.

Forum statistics

Threads
143,066
Messages
1,467,358
Members
104,935
Latest member
Pauos31