droneranger said:
Mistaken however, if he is in SE Australia as he indicated. The largest declination down there is around +14˚. You have to head over to New Zealand to get up into the 20s.
Fixing DJI's Compass Problem
- Thread starter ianwood
- Start date
Mistaken however, if he is in SE Australia as he indicated. The largest declination down there is around +14˚. You have to head over to New Zealand to get up into the 20s.
Actually N017RW, no. What you are referring to is magnetic declination; the difference between true North and magnetic North. I was addressing magnetic deviation which is essentially a magnetic error due to local fields (either man-made or natural).
Maybe where he flies he actually has 10 degrees positive local magnetic deviation on top of the declination. Correction for that is going to depend on the automatic compass calibration the Phantom performs. Maybe the next beta will include saving that data in flash memory and help solve the local deviation problem too.sar104 said:
N017RW
Premium Pilot
droneranger said:
You're absolutely right.
But I don't take every post so literally.
Though we may share a common language, we all communicate differently.
I understood what 'Pazz' was trying to say and 'riblit' further clarified the subject IMO.
One more thing...
I would like to know how to measure my local magnetic deviation.
Is 22 degrees of deviation a lot?
Thanks.
N017RW said:
Ummm yes
. "Is two cups of bleach too much with my good suit?"Try this: http://magnetic-declination.com
damoncooper said:
I'm confused. I've got magnetic-declination.com on a tab on my browser. How do you get it to show magnetic deviation for a local area? Does it just automatically do that?
Magnetic deviation is one of the variables you are zeroing out when you calibrate the compass. It is any disturbance in the magnetic field localized to your P2. The P2 by itself will cause some level of deviation. Things you add on like a vTX will too. It will alter the way the compass sees the magnetic field around the P2.
On a plane you will have a compass card that indicates the deviation in quadrants for the mechanical compass. It gets updated whenever you change equipment much like we should recalibrate the compass when we change equipment.
On a plane you will have a compass card that indicates the deviation in quadrants for the mechanical compass. It gets updated whenever you change equipment much like we should recalibrate the compass when we change equipment.
ianwood said:
Are you sure the compass calibration zeros out mag deviation, like local iron deposits? I thought the Phantom calibration dance was just to make sure you hadn't done something to magnetize the compass. Basically just making sure forward on the quad equals forward on the compass. DJI doesn't even have you orient the quad in any particular direction during the calibration. How could it possibly know what was deviation relative to true north from local effects (external to the machine) versus magnetic declination? All it sees is a single magnetic field.
I'm sure you're right that adding things to the Phantom disrupts the magnetic field, requiring calibration. But I don't buy that it has anything to do with correcting for external causes of magnetic deviation. You've done more research than me so maybe you can explain.
Hi, I'm not sure about this concept of the system learning. My view is that the NAZA autopilot system is readjusting to the signal updates from the GPS system. Talking to RAAF pilots here in Australia the GPS system can be up to 250 meters offriblit said:
at any given time. They only use GPS as an aid to navigation not a rock solid given. This is workable for large aircraft travelling at high speed over large areas. However our little drones mainly operate within the field of error of the GPS system. Usually we operate at comparatively low speed & at 400 meters ranges. The result is that GPS corrections cause the drone to hunt to the next update to establish position, hence TBE & flight direction errors/hooking at the end of runs. Therefore its not learning, just readjusting to the GPS corrections. PS re GPS accuracy in USA, you have an extra system in place to improve accuracy, benefit of it being run by the American defence dept. So despite Mag. variation,in the USA the GPS system is better then the rest of the world
I don't know exactly how it works but I would guess by taking a 360 degree sample, it looks for consistent deviations across what would otherwise be a 0-360 sample, i.e. ones that move with the compass and are effectively part of the P2's magnetic fingerprint. That can include magnetization of the compass to a point as long as it isn't so magnetized that it can no longer "see" through it.
I would also guess it could look for local anomalies like the iron deposits you mention that would be represented by a transient anomaly as it passes from 0-360. Think of a deviation on an otherwise linear line.
That is just a guess. But I do know the compass calibration is both for local magnetic fields and for magnetic fields generated by things on the P2 itself.
Paz, GPS jumps happen but mostly when the field of view to the sky is changed dramatically. Unless the RAAF use an array of GPS antennas or gimbal stabilized antennas which I would have thought they do, the GPS can suffer from sudden changes of FOV say in a barrel roll which would result in a GPS jump. We don't have that issue under an open sky. GPS is not the issue here. It is the compass. It has been proven. And the system definitely learns on each flight. And then unfortunately, it forgets. If it wasn't declination related, everyone would have the problem evenly which isn't the case.
I would also guess it could look for local anomalies like the iron deposits you mention that would be represented by a transient anomaly as it passes from 0-360. Think of a deviation on an otherwise linear line.
That is just a guess. But I do know the compass calibration is both for local magnetic fields and for magnetic fields generated by things on the P2 itself.
Paz, GPS jumps happen but mostly when the field of view to the sky is changed dramatically. Unless the RAAF use an array of GPS antennas or gimbal stabilized antennas which I would have thought they do, the GPS can suffer from sudden changes of FOV say in a barrel roll which would result in a GPS jump. We don't have that issue under an open sky. GPS is not the issue here. It is the compass. It has been proven. And the system definitely learns on each flight. And then unfortunately, it forgets. If it wasn't declination related, everyone would have the problem evenly which isn't the case.
Hi Ian, GPS jumps occur for all sorts of reasons. Satellite postions, whether wide apart or close together. The GPS system even guesses postion of where a satellite should be if it can't find one to get triangulation Signals reflected off objects, which can happen under open sky, if the angle to the satellite is low to the horizon. Try out your "find my phone"app & over a number of days, it can vary in accuracy as much as 50 meters or sometimes can't find you. Also car nav. system(which use GPS) can sometimes have you travelling 25 to 50 meters off the road or slow to catch up your postion. I still think is idea of "learning" is not founded. As we take the drones up higher after takeoff the GPS jumps occur less as interference reduces compared to closer to the ground. Thats would explain why it does not remember, as there is no learning, just on the fly adjustments in that flight.ianwood said:
That's why I askedsar104 said:
I'm in SE Aust. and my magnetic declination is +12. Paz may actually be referring to his deviation (as he posted). Correctly explained by droneranger.Pazz
If you read through all of the three major threads related to this magnetic declination issue you will find...
People effected are in areas with generally more than 10 degrees of declination. There is almost nobody effected with less than 10 degrees declination. People have flown their Phantom in areas of both high and low declination and have attested to the difference in flight performance. DJI is in a low declination area and probably never sees this in their testing.
Almost all of the people affected have observed that the TBE, drift, JHook, and Course Lock error are very noticeable in the first few minutes of flight and then virtually disappear. How long it takes to disappear varies with declination. If you land but keep it powered, then takeoff it "remembers" the refinement and doesn't have a problem. If you land, change the battery, start it up and takeoff again it goes through the whole process again from scratch. There is no doubt that it is learning. In a video I posted recently you can watch the problem go away (fairly quickly with the Beta 3.05) as it "learns" when the motion it commands based on the magnetic compass doesn't match the motion it gets from GPS data.
Geometrically GPS requires three satellites to get a position but four is considered a practical minimum. DJI requires a minimum of six to be in GPS mode. Each satellite adds reliability and accuracy. Often I see 8 to 12 in flight which further reduces errors and fluctuations. Certainly there are GPS errors and they have an effect but that is not what is causing TBE, drift, JHook, and Course Lock error. If it was it would happen randomly throughout the flight - but that is not what we are seeing.
There is a subset of us (not me) who rotate their compass, usually twice the declination. They claim that reduces the problem greatly - probably because it reduces the mag/GPS discrepancy. Most feel this doesn't solve all the problem and is not a good long term fix.
That's why I think it is declination, it is learning, and it's not GPS errors.
If you read through all of the three major threads related to this magnetic declination issue you will find...
People effected are in areas with generally more than 10 degrees of declination. There is almost nobody effected with less than 10 degrees declination. People have flown their Phantom in areas of both high and low declination and have attested to the difference in flight performance. DJI is in a low declination area and probably never sees this in their testing.
Almost all of the people affected have observed that the TBE, drift, JHook, and Course Lock error are very noticeable in the first few minutes of flight and then virtually disappear. How long it takes to disappear varies with declination. If you land but keep it powered, then takeoff it "remembers" the refinement and doesn't have a problem. If you land, change the battery, start it up and takeoff again it goes through the whole process again from scratch. There is no doubt that it is learning. In a video I posted recently you can watch the problem go away (fairly quickly with the Beta 3.05) as it "learns" when the motion it commands based on the magnetic compass doesn't match the motion it gets from GPS data.
Geometrically GPS requires three satellites to get a position but four is considered a practical minimum. DJI requires a minimum of six to be in GPS mode. Each satellite adds reliability and accuracy. Often I see 8 to 12 in flight which further reduces errors and fluctuations. Certainly there are GPS errors and they have an effect but that is not what is causing TBE, drift, JHook, and Course Lock error. If it was it would happen randomly throughout the flight - but that is not what we are seeing.
There is a subset of us (not me) who rotate their compass, usually twice the declination. They claim that reduces the problem greatly - probably because it reduces the mag/GPS discrepancy. Most feel this doesn't solve all the problem and is not a good long term fix.
That's why I think it is declination, it is learning, and it's not GPS errors.
I've had to take off with only 5 Sats. but acquired more once airborne
Does the P2 know where it took off from as in "return to home"?
Great information!
Does the P2 know where it took off from as in "return to home"?
Great information!
Navman said:
I don't believe RTH is even valid without 6 sats reported. In fact I've seen my Phantom reset RTH all on it's own (greens flashing) well after manually setting it. If it goes from less than 6 to 6 sats, no matter where it was when it finally acquired them, it resets RTH.
As a practical example, I often fly from my backyard which is hemmed in on 2 sides by other homes, making it difficult to get 6 or more sats while still on the ground. So what I usually do is wait until I get 4 or 5, then take it up 20-30 feet to capture more. At that point, coincident with the 6th (or more) sat being acquired, I'll see the Phantom reset the home point. This is fine, since I'm doing it directly above my origin, but if you fly the Phantom away with less than 6 sats, you'll likely end up with a homepoint in a completely random spot. Just don't fly away with less than 6 unless you you plan to fly ATTI or manual.
- Joined
- Sep 4, 2013
- Messages
- 2,219
- Reaction score
- 7
varmint said:
Yes, in fact, I don't think it even lets you "manually reset" your home position until it's automatically set one. So if you flipped S2 repeatedly before ever getting 6 satellites, it will override it with a "true" automatic home point as soon as 6 are acquired.
To be clear, it doesn't reset your home point if it loses satellites and then regains them, this is just at first acquisition.
John Shaw said:
The following Official DJI NAZA Manual details how to calibrate and solve the drifting and TBE.http://team-blacksheep.freshdesk.com/he ... s/12145075
Page 11, details compass calibration.
Page 29, fixing the TBE.
NAZA-M_Quick_Start_Guide_v1.20_en.pdf
1.2 MB
This problem is across all DJI NAZA auto pilot systems & all air frames. Solution is to rotate gps module, but how do you do that with phantoms. Also as pointed out by many other rc forums, we are only dealing with a hobby level system. Get what you pay for.
Pazz said:
I really think that you need to read up on the background to this issue, as detailed here and in several other threads, before posting further. The points that you are raising have all been beaten to death previously, and it's not very helpful to start all over again from scratch with this discussion.
Sorry you feel that way. I have read the threads & GPS rotation has not really been examined. If you read other threads of different airframe users (here & on other forums) these issues are being looked out differently. The problem with the phantom is that the GPS unit is glued in under the top of the shell, so no one is prepared to put in apart & see whether is works. Easier on other airframes where GPS unit is external & adjustable.Just looking for a software fix may not be the answer.sar104 said:
Similar threads
