I have thought about what could trigger the persistant compass error that happens when you are flying for the first few minutes. Could it be the magnetic field generated by the rotating four motors?

i set up an experiment i put a sensitive compass close to the motor while i started the inspire Voila!! there is a strong magnetic field generated by each motor

watch for the compass needle in the attached video

That's interesting .. how far from the motor does it effect the compass?  The only problem that I see is if the motor is producing the problem wouldn't it be a constant problem while the motor is on?

WOW, wonder how DJI will get around this?

Maybe they need to MOVE the compass sensor?

I have the Phantom 2V+ and I will see how this compare as well.

Good finding, i too see this same error on each flight, each take off.

Best regards,
Steve

The magnetic force generated by the motors are surely compensated for already even the little cooling fans 1 in the cam and 1 under the hood.
They could try and shield them though

I do not understand much about it but then, why that occurs only with the new firmware and not with the previous firmware.
My INSPIRE1 had no problems until I update the firmware.
Since then I have the problem of compass and change to att and a small nod of gimbal that seems lost for a second horizontality.
Sorry for my English

I doubt they would miss that before production. I'm pretty sure the magnetic field generated from the motors will be visible with sensitive devices, but I doubt it would account for the nav problems. If so, everyone would experience compass issues. I don't fly my inspire that often but I've only experienced gps loss shorty after take off a couple of times. If the issue was hardware related it would've occur all the time.

On a sidenote, I do notice that my inspire has trouble calibrating the compass when I'm indoors, where my phantom doesn't have any such trouble, so somehow it's definitely more prone to interference compared to the phantom.

I just recently updated so I haven't been flying with 0.17 yet. I'm actually crossing my fingers that I won't run into trouble!

xzces Posted at 2015-3-14 03:44
I doubt they would miss that before production. I'm pretty sure the magnetic field generated from th ...

0,17 they changed the threshold for interference in the software. So it makes sense now small compass interference has SAME effect as LARGE before

arunmehta Posted at 2015-3-14 04:30
0,17 they changed the threshold for interference in the software. So it makes sense now small comp ...

Well that does explain it. It still doesn't make sense to me that they would narrow down the treshold if now more people are having compass problems. We need both compass &  gps to fly in p mode right?

xzces Posted at 2015-3-14 06:31
Well that does explain it. It still doesn't make sense to me that they would narrow down the tresh ...

the threshold was lowered to make the gps failure less likely (compass error too much makes gps error)but there is a fine balancing art....or they have to change the whole algorithm..which is the route i would take at this point..change the whole algorithm..
As it is the plane is flyable and you are able to finish the mission i just make sure i am hiigh up and away from obstacles when it loses compass and ATTi mode kicks in for a second ...be alert and takeover you are fine...

All electric motors generate an electric feild when current is applied to them. The problem I see with your theory is that the amount of flux should increase with applied current and the field will be different on any motor given its demand. Image an almost stalled motor on one side and a nearly cranked motor on the other. There will be a fairly large change in the fields. Anyway, if what you say is true, what makes the aircraft de sensitize to the fields? Why does it not happen immediately on start up and normally shortly after flight? And why does it not persist?

The compass error develops after the aircraft takes off and gets to around 5 -10m height. It lasts a short while and then goes away. Sometimes, it recurs after a minute but thereafter, no problems.

I've encountered the problem only in high latitudes. Never at near the equator so I speculated it has something to do with magnetic declination. So far, no response from engineers / forum leaders like Blade or Tahoe.

It looks to me that you had the compass right under the motor. Since the effect would fall off exponentially as you move away from the motor the compass should be mounted on top of the center unit as this is where all the electronics are located. It would be interesting to see what your compass does then.

The compass will react like any magnetic compass placed over metal, it will deflect. The question should be why is it correcting? Why does it happen only some times, I have had flights without this error. And finally, do the engineers feel like telling us anything? Or is this top secret? Can't very well apply for an SFOC with an unreliable aircraft now can we?

The DJI compass are different  thing, the famous the "calibration or copter dance" is necessary  almost mandatory  because diiferent place has different " magnetic declination"

https://www.ngdc.noaa.gov/geomag/declination.shtml

The cheaper pocket compass (video)  isn't rated  as  EMC ,(electromagnetic compatibility)  any small magnetic variation caused by external source (electric motor ) will affect it.

There is an compass app on the iPhone I believe it comes stock on most iPhones. This app requires you to calibrate it by rolling the ball around the perimeter of the compass dial. In effect it is exactly the same as the aircraft compass. They are both digital instead of analog

Süleyman Posted at 2015-3-15 12:26
The DJI compass are different  thing, the famous the "calibration or copter dance" is necessary  alm ...

Upon further analysis and speculation:
where i live the magnetic declination is 23degrees. The motors definately create four equaly distant magnetic fields. I have a feeling the software is "poling" the compass every few minutes? looking for the direction of aircraft and  compute the current orientation. The purpose of the compass is to determine the" current" position of the aircraft in space and its orientation relative to the magnetic/geographic north. because of the magnetic input from the motors this computation of "direction" will be at odds with the GPS computation of location/orientation.
But why does this error only happen certain times and more important goes away in seconds ?

i find the error is most likely to happen when i suddenly rev up like going up and THEN slow to idle like hovering. the strong magnetic field during revving the motors will confuse the last polling of software and throw an error!  BUT once the motors are back to "normal" revs and normal magnetic field when there is the NEXT POLLING by the software (i think the frequency of polling is same as the time this error lasts ie seconds)  the error goes away and the gps mode is safe to fly

Only places where the magnetic declination is large will show this up MORE OFTEN  hence i would be interested in knowing the magnetic declination of people who are NOT getting this error
Those motors are very strong and have a robust magnetic imprint and that imprint intensifies and is stronger when revving them up.

the solution:
***Software tweeks (they tried in 0.17 update by playing with threshold )
***magnetic shielding of motors and the compass (wouldnt magnetic shielding of compass make it ineffective?)
***Have magnetic declination entry on first handshake of the software with hardware so the software
       compensates for the magnetic declination of each particular location


Arun

arunmehta Posted at 2015-3-15 15:01
Upon further analysis and speculation:
where i live the magnetic declination is 23degrees. The moto ...

the solution:
***Software tweeks (they tried in 0.17 update by playing with threshold )
fine tuning to the Mathematical algorithm (firmware)

***magnetic shielding of motors and the compass (wouldnt magnetic shielding of compass make it ineffective?)

probably increase the internal IMU shield.

***Have magnetic declination entry on first handshake of the software with hardware so the software
       compensates for the magnetic declination of each particular location
Different manufacturer request  the manual input for optimal accuracy (magnetic declination) or auto mode.

arunmehta Posted at 2015-3-15 15:01
Upon further analysis and speculation:
where i live the magnetic declination is 23degrees. The moto ...

I have no idea why you are postulating on these theories when the problem has been fixed in the next roll out of firmware.

Blade and Ed (and others) have already done exhaustive testing on this very issue with the current beta firmware and this bug and the gimbal twitch (which are intrinsically linked) have been addressed.  

Job done -

markjacobs.talk Posted at 2015-3-15 21:21
I have no idea why you are postulating on these theories when the problem has been fixed in the ne ...

Excellent news Markjacobs...i had no idea that this problem was fixed when is the new firmware coming out?

Will it also fix the tolit bowl affect I get in the first minute of flying?

arunmehta Posted at 2015-3-15 15:01
Upon further analysis and speculation:
where i live the magnetic declination is 23degrees. The moto ...

No problem at equator. Never. >30 flights.

markjacobs.talk Posted at 2015-3-15 21:21
I have no idea why you are postulating on these theories when the problem has been fixed in the ne ...

This is a technical problem. Please go look up what magnetic declination refers to. Solving the issue is not so simple. I will say that it was less of a problem in later versions of the P2 at similar latitudes both north and south of the equator. I've gone as far as 55N and 45S in addition to being 1N so I know what I'm talking about. DJI will get to a solution but it will take a little time

markjacobs.talk Posted at 2015-3-15 21:21
I have no idea why you are postulating on these theories when the problem has been fixed in the ne ...

Credit goes to DJI for the postulations.  DJI should try to keep their customers in the loop when possible.

Maybe Markjacobs can respond...When is this supposed update to firmware coming? anyone know? 0.17 has too many flaws
Markjacobs wrote :

I have no idea why you are postulating on these theories when the problem has been fixed in the next roll out of firmware.

Blade and Ed (and others) have already done exhaustive testing on this very issue with the current beta firmware and this bug and the gimbal twitch (which are intrinsically linked) have been addressed.  

Will this new firmware possibly solver my problem of compass calibration failing and makes my inspire an ultra expensive paper weight.

mattks Posted at 2015-3-22 20:29
Will this new firmware possibly solver my problem of compass calibration failing and makes my inspir ...

there is a way to fix the compass with a magnet but you better leave it to dji...i know it works for phantom but Inspire i am not sure
your problem ...if i recall is 800 number in compass it should be more than 1300  you can increase your number by running a strong magnet about 3 inches from the compass...

i dont think a new firmware will solve your problem...your compass may have been demagnetized by a strong speaker magnet in an adjacent shipping box ...just speculating

I guess what I'm no understanding is with all of this chatter with compass problems that DJI has not reacted with an update in the firmware (if that is the way to fix the problem).

I spoke with DJI at great length last night about the possibility of the positioning of the compass with the motors causing compass errors in flight.  They did not acknowledge this problem at all.  They told me that I would have to re-callibrate my compass on a daily basis or tie up my inspire by shipping it back for service.  When I asked how could this be the case when many people all over the world are seeing this problem.  DJI didn't even acknowledge what was being presented here on this discussion.  Kind of disappointed to hear.

Mike-the-cat Posted at 2015-3-16 00:39
No problem at equator. Never. >30 flights.

so that confirms my proposal...and gives clues to a possible answer..

Entering the current co-ordinates and a few tweaks in software will solve this issuefor most of us it is not a FATAL problem BUt try telling it to the guy whose gps lost and his inspire hit his garage door ( all he had to do was take over and steer it back away from garage in ATTi mode)

Hello from Miami where I too have had the same issue.  The other day, after compass calibration (which I do every time I fly) within the first minute of flight the Inspire 1 displayed "Compass Error," lost control, and flew directly into a building at about 15 feet high.  Did tremendous damage to the drone.  I called DJI and they told me it could be a compass malfunction or interference, but never mentioned that anyone else was having the same problem. I had to ship it back at my expense ($160) and now I will wait for a response.  The scariest thing about this is that with the speed of the blades someone could get seriously injured.  Luckily there was no one around when the Inspire 1 went crazy.  DJI should address this issue immediately before someone gets hurt.

I was told that the compass is placed in the "nose" of the bird.
Do you really think that magnetic variations produced by the motors can actually influence the compass readings?

Bornish, I am not sure theoretically it should...

I wish I had a fix... I have owned the Inspire 1 for 2 months now, acutally two since the camera had some issues on the first. I got compass errors on the first and second unit right out of the box, doing all of my compass recalibration, IMU etc., and still have issues. I get the error usually withing the first couple of minutes of flight, then it usually goes away for the rest of the flight. Freaks me out especially since I fly in windy conditions sometimes. So if you have wind, are near objects, and suddenly you go to ATTI mode the craft can fly into a building or tree. That means you have to be really careful about where you fly, anticipating losing control. I own a Phantom 2 also and never had problems like this. I recently sent back the second Inspire and they are looking at it now but the techs act like this is a new problem. Guess not. Any solutions please post!

I wish I had a fix... I have owned the Inspire 1 for 2 months now, acutally two since the camera had some issues on the first. I got compass errors on the first and second unit right out of the box, doing all of my compass recalibration, IMU etc., and still have issues. I get the error usually withing the first couple of minutes of flight, then it usually goes away for the rest of the flight. Freaks me out especially since I fly in windy conditions sometimes. So if you have wind, are near objects, and suddenly you go to ATTI mode the craft can fly into a building or tree. That means you have to be really careful about where you fly, anticipating losing control. I own a Phantom 2 also and never had problems like this. I recently sent back the second Inspire and they are looking at it now but the techs act like this is a new problem. Guess not. Any solutions please post!

I wish I had a fix... I have owned the Inspire 1 for 2 months now, acutally two since the camera had some issues on the first. I got compass errors on the first and second unit right out of the box, doing all of my compass recalibration, IMU etc., and still have issues. I get the error usually withing the first couple of minutes of flight, then it usually goes away for the rest of the flight. Freaks me out especially since I fly in windy conditions sometimes. So if you have wind, are near objects, and suddenly you go to ATTI mode the craft can fly into a building or tree. That means you have to be really careful about where you fly, anticipating losing control. I own a Phantom 2 also and never had problems like this. I recently sent back the second Inspire and they are looking at it now but the techs act like this is a new problem. Guess not. Any solutions please post!