OK I'll work on getting some data from my bird up for us all to see here.When you apply forward elevator in P-GPS, the FC attempts to fly a course that matches the aircraft heading - i.e. it should fly in the direction that it is facing. Assuming that the aircraft heading (yaw) actually matches the camera view (i.e. the gimbal is aligned with the aircraft), then it should move in that direction. That's achieved by applying additional negative pitch (forward tilt) relative to the attitude while hovering, and then applying roll corrections if the track diverges from the desired course. The track is computed from the IMU data corrected by the GPS data.
When that doesn't work, as is clearly the case here, there are only a couple of possible causes. Bad GPS data could cause issues, but that's very unlikely to result in a constant course offset. The obvious mechanism is simply that the aircraft has its heading (yaw) incorrect - in other words it is not facing in the direction that it thinks it is. In your examples above, if the IMU yaw value were off by 5° to the left, then that would be the course that the FC attempted, and succeeded, in flying.
The IMU yaw value is initialized at startup from the compass data, which requires a couple of things to go right. Firstly, the aircraft needs to be powered up at a location with no external magnetic interference, so that all the compass sees is the earth's magnetic field (plus the aircraft's magnetic field - but that is subtracted out during calibration). Secondly, it needs to know the local magnetic declination so that it can convert its magnetic heading to a true heading. That's achieved with a global magnetic declination model in the firmware. Interestingly, the P2 shipped with an incorrect model that caused somewhat similar issues, but that doesn't necessarily imply that this is the same problem.
It's impossible to diagnose further without one of the DAT files - the smaller mobile device DAT file is good enough. That contains the raw compass data, the magnetic declination calculation, and the resulting IMU yaw calculation, and should show unambiguously why it is not flying straight.
yup this video that I posted here is not me or my bird but is the same exact thing that's happening.As a flyer of a P3 and P4, I can tell you that in a steady sideways wind both will drift smoothly sideways with a perpendicular wind. The only time you see camera motion at all is with strong gusts.
Wind is never constant, but on the video the movement was. Either way, OP just stated that it does it in 0 wind as well.As a flyer of a P3 and P4, I can tell you that in a steady sideways wind both will drift smoothly sideways with a perpendicular wind. The only time you see camera motion at all is with strong gusts.
Looking at the video only without any telemetry, the drone is actually pointed in the right direction but is drifting left. The bird is drifting left holding the correct direction.
Two possibilities - there is a strong wind from the right side pushing the drone to the left, or the right hand joystick isn’t properly centered in the left-right axis.
Try flying a tap and fly intelligent flight and see what happens. A waypoint flight would provide the same information. Also try identifying the wind direction, and try a flight parallel to it.
Could you please direct me to whomever these guru's are so the may be able to help me read the flight logs.My friend, you have encountered the main issue that has plagued V2.0 users. I had the same issue with mine. You can look through the various threads here by searching “yaw drift” and find where some of the gurus went thru my flight logs and found a YAW/magYaw separation in the data. They believed the flight controller was holding one or the other value fixed and rotating the drone until the values matched.
I finally received a new replacement P4P V2 a couple weeks ago, i had to send it off twice. First time it was sent back to me saying “no issue”. I prodded their engineers with the flight log findings of the gentlemen here and they asked me to send it back. At that time they deemed the GPS module was bad and replaced my drone. I havent flown it to see, i was just going to unload it and try to find a V1. Never had an issue there. I hope youre still in the warranty period, and maybe that ive helped.
Could you please direct me to whomever these guru's are so the may be able to help me read the flight logs.
I would love to know if your replaced GPS modual has fixed the problem.
That is a horible to hear the replaced GPS did not correct the problem.Sar104 (who has posted in this thread) and BudWalker.
See post 6 I already flew the new Phantom and its yaw drift was observed less than 15secs into the first flight. I, like you, regret ever getting rid of my V1 for this machine.
That is a horible to hear the replaced GPS did not correct the problem.
Thanks for the guidance to the flight log guru's.
Maybe some day DJI will address the problem.
It did however take them almost 7 months to correct the 106 firmware release that grounded the p4 standard. So I guess we shall see.