Man I dig that music. LOL
I think most advice you are going to get on this will be IMU calibration as you have already noticed but have you checked your stick calibration?
THE DREADED YAW DRIFT. Seems many have this where...
- Thread starter Blade4
- Start date
Man I dig that music. LOL
I think most advice you are going to get on this will be IMU calibration as you have already noticed but have you checked your stick calibration?
Hi all, I'm fairly new to the Phantom 4 Pro v2.0, but I have experience the same thing and found changing my altitude made all the difference. Before my P4P2.0, I flew ultralights, mine was an American Eagle, no tail, front facing smaller wing, Kinard (sp) design, and I can tell you first hand the wind at ground level may be very different at any altitude, from no wind on the ground to strong winds at 60' to no wind again at 100' and extreme at 500'. Please understand in an ultralight I had to fly above 500'. Wind conditions were dependent on a number of factors, how hot, how humid, whether you were flying over water, or freshly plowed field or asphalt, and so far, my experience with the P4P2.0 is consistent with the ultralight...., except the P4P2.0 is a heck of lot more stable.
I suggest next time you think you have drift, let go of the stick, if it stays put, then you are being affected by the atmosphere, if it drifts, you have a real issue.
Either way, I am interested.
I suggest next time you think you have drift, let go of the stick, if it stays put, then you are being affected by the atmosphere, if it drifts, you have a real issue.
Either way, I am interested.
This is not a wind issue, period. The “stability” of the Phantom is not being brought into play. The P4P V2 can still fight a 20mph wind while stationary (position-wise) as it yaw drifts either left or right autonomically. It will also yaw drift ten feet off the ground in a warehouse with closed walls and no moving air. DJI has already tried the “its the GPS module” excuse, but they sent me a “new one” and it does the same thing. There is a true hardware issue. (Also I find it prudent to note that my “new drone” had a new SN, but only had stickers on the arms and not the sensors or camera, and that the UV filter has a small scratch (probably someone elses used drone). One might find it coincidal that the P4P V2 had the shortest sales and production duration of any DJI drone model. I don’t.
I have gone back and reviewed the video clip several times and still I have to agree with David_Cambridge, the P4P V2 is attempting to hold a line with a slight twist to the right as though it is correcting for a fairly stiff breeze coming off the water to it's right. This is normal behavior. In your warehouse example, you shouldn't have a strong GPS signal, being indoors, a drift would be expected since you're pretty much in atti mode.
It is your assertion that the drift is either left or right that is more concerning and the reason I think it is atmospheric in nature. A sticking control stick will cause an issue, but it will be consistent, in fact, most things I can think that might occur will be constant in behavior. I'm guessing you've checked your props, and pretty much done a complete once over of the aircraft to make sure nothing is damaged or twisted?
I wish you the best and hope you find your source of distress.
It is your assertion that the drift is either left or right that is more concerning and the reason I think it is atmospheric in nature. A sticking control stick will cause an issue, but it will be consistent, in fact, most things I can think that might occur will be constant in behavior. I'm guessing you've checked your props, and pretty much done a complete once over of the aircraft to make sure nothing is damaged or twisted?
I wish you the best and hope you find your source of distress.
You still miss the point entirely, sir. We arent grumbling about the physical/lateral movement of the drone! This behavior CAN BE OBSERVED WITHOUT FORWARD MOVEMENT. Park the P4P V2 somewhere in the air and point it North. Set the remote down. The drone will end up pointing 40-60 degrees either NE or NW BY ITSELF WITH YAW (think turning your head on your shoulders to look left or right, your body stays still) MOVEMENT ONLY in less than 90 seconds. THIS SLOW AUTONOMIC MOVEMENT is what stops the Phantom from being able to fly a straight line because its flight controller is seemingly never in agreement on the YAW and magYAW values and holds one or the other constant, while allowing the drone to drift to “correct” the separation that the FC finds unsuitable. The movement is visible in the flight data with ZERO stick input. Whats causing the separation is unclear.
Ive tried numerous calibrations, (and yes checked the props; God if it were that simple) and received replacement drones from DJI. The results have been consistently disappointing. I hope this helps you understand the issue better.
- Joined
- Jul 29, 2016
- Messages
- 446
- Reaction score
- 235
- Age
- 81
You are describing an entirely different issue than the original poster. Using the video data supplied there, it is clear that the drone’s direction is lined up with the road the entire flights. There is no change with what is apparently the intended direction straight up the street.
The movement in question is a left drift, with the drone still maintaining a CONSTANT heading. The only 2 reasons that I can think of that would cause it would be a wind from the right, or an improperly centered right joystick. SAR104 poses an alternate explanation of a mis-alignment between the inertial guidance system and the compass, which may also be an explanation if I understood it clearly.
The movement in question is a left drift, with the drone still maintaining a CONSTANT heading. The only 2 reasons that I can think of that would cause it would be a wind from the right, or an improperly centered right joystick. SAR104 poses an alternate explanation of a mis-alignment between the inertial guidance system and the compass, which may also be an explanation if I understood it clearly.
Those two behaviors can be directly related however - both resulting from a discrepancy between the IMU yaw and the magnetic yaw values. The incorrect IMU yaw value causes the flight course to be offset from the aircraft heading, and the slow rotation is one method that the FC uses to try to reconcile the error - slowly rotating the aircraft while freezing the IMU yaw value.
- Joined
- Jul 29, 2016
- Messages
- 446
- Reaction score
- 235
- Age
- 81
I don’t see the yaw changing in the video, although there could be some some very slight change that isn’t visible. Would the difference between the imu direction the compass direction show up in the flight logs? That would settle this debate if we could get them.
Yes - it will be in the mobile device and aircraft DAT files, as I mentioned in #8.
This is exactly what I was explaining, maybe not as eloquently.
How can the heading be straight and the center reticle be locked if the drone is strafing left and not yawing?
Its a hardware issue.
I also am a little confused now. It seems at first you say the drone in the video (and yours as well) was crabbing left. Then you described a yaw left when it’s hovering, which as David C. says, are two different things.
But now as I watch the video and put it together with the reticle comment, are you saying that it’s crabbing left by yawing RIGHT this keeoing the reticle fixed on a certain spot?
I must concur with those stating that this is absolutely not a wind issue. The Phantom 4 pro version 2.0 doesn't get pushed sideways by the wind, at least I have not come across any valid reports of this occurring. It does however move in the yaw axis uncommanded. Another simple explanation for this is that the gimbal is looking to the right. So the Drone may actually be flying perfectly straight but the camera is looking to the right. This happens to me all the time and it's frustrating so I have to use extra finesse to fly the Drone diagonally to pull off a straight looking shot.
I think this issue can simply be explained by the fact that the camera is not facing forward. In the case of the video posted by the OP, the camera is pointed slightly to the right, so when the pilot pushes the stick forward it appears to drift to the left.
This happens all the time on my P4P V1--so it's not just a V2 issue. I didn't really notice it much until a year ago when I tried chasing airplanes down a runway, as most of my flying before then didn't really rely on flying straight in sports or ATTI mode as fast as possible. You don't notice it in waypoint flying because you're not using the camera to track your flight and the yaw angle of the camera gimble is so slight it's not entirely obvious. It's also not a wind issue. In my situation it would seemingly "drift" to the left even when flying a course 180 degrees from previous.
If you hold a straight edge (non-ferrous metallic otherwise that will make your gimble twitch) across the front of the drone, you can see that the camera lens is slightly pointed off center. Just like the video above, my camera points slightly to the right. I've tried ever fix I've come across to make sure the camera centers properly; like twisting the gimble (with power off) to its stops in opposite direction with a little extra pressure, tried auto-calibrating the gimble while resting the drone on it's rear--resting on it's rear motors and landing legs (camera pointing up). Nothing has helped.
I did notice a year ago that the Go 4 App now includes a yaw offset (just like a roll offset control), and that is great for aligning the camera with a straight edge--however, it only lasts while (if at all) during that same power up. As soon as you change batteries the gimble no longer points forward. I've thought about trying to set up some kind of easy template so I could align my camera manually and quickly after each power up, but haven't thought of anything practical.
I do still experience actual "yaw drift" in flight. Often not noticeable unless shooting long exposure, and thats when you can see a distant light at night as a single pixel start to slowly creep across a grid line overlay over several seconds. Even this seems to be randomly "fixed" by finding a certain heading hold where it won't drift--but I'll go through a whole battery trying to find that sweet spot and then it's not the same heading the next time I fly; and of course sometimes that heading doesn't frame your subject like you want.
***EDIT*** Didn't see Dallas Drone Guy had already said the same thing...
This happens all the time on my P4P V1--so it's not just a V2 issue. I didn't really notice it much until a year ago when I tried chasing airplanes down a runway, as most of my flying before then didn't really rely on flying straight in sports or ATTI mode as fast as possible. You don't notice it in waypoint flying because you're not using the camera to track your flight and the yaw angle of the camera gimble is so slight it's not entirely obvious. It's also not a wind issue. In my situation it would seemingly "drift" to the left even when flying a course 180 degrees from previous.
If you hold a straight edge (non-ferrous metallic otherwise that will make your gimble twitch) across the front of the drone, you can see that the camera lens is slightly pointed off center. Just like the video above, my camera points slightly to the right. I've tried ever fix I've come across to make sure the camera centers properly; like twisting the gimble (with power off) to its stops in opposite direction with a little extra pressure, tried auto-calibrating the gimble while resting the drone on it's rear--resting on it's rear motors and landing legs (camera pointing up). Nothing has helped.
I did notice a year ago that the Go 4 App now includes a yaw offset (just like a roll offset control), and that is great for aligning the camera with a straight edge--however, it only lasts while (if at all) during that same power up. As soon as you change batteries the gimble no longer points forward. I've thought about trying to set up some kind of easy template so I could align my camera manually and quickly after each power up, but haven't thought of anything practical.
I do still experience actual "yaw drift" in flight. Often not noticeable unless shooting long exposure, and thats when you can see a distant light at night as a single pixel start to slowly creep across a grid line overlay over several seconds. Even this seems to be randomly "fixed" by finding a certain heading hold where it won't drift--but I'll go through a whole battery trying to find that sweet spot and then it's not the same heading the next time I fly; and of course sometimes that heading doesn't frame your subject like you want.
***EDIT*** Didn't see Dallas Drone Guy had already said the same thing...
Last edited:
In post 26 I described stationary autonomic movement to the LEFT OR RIGHT could be observed
I also described that this movement causes trajectory issues when the drone is in motion.
I never said anything about the drone in the video, I just responded to the title (as I do every time I see V2 Yaw drift, its very common and not going away). Everyone else has speculated on that.
My reticle comment was intended to stimulate thought in those who believe wind is the culprit; IF the wind is crabbing the drone left, and you are flying straight and your reticle is holding position, your drone would have to be yawing right, would it not? The possibility of the camera not being aligned center-forward with the drone may help explain this particular phenomenon, but again, if the drone is flying straight with a misaligned camera and appears to be crabbing left with a locked center reticle, it is rotating to the right. Slowly.
Last edited:
No sweat.
Sorry if I seem irritated, I am, but not with anyone here. This issue has plagued me since 9/19/18 when I “upgraded” from the V1 to V2 and I want a solution for myself and others. My V2 is going back for the 4th time on Monday. Ill report back with what I can.
Similar threads
