Fixing DJI's Compass Problem

[nzvideoguys]

What I know is irrelevant. I have never wished ill will on anyone based on a statement taken out of context or without a complete understanding of that persons intentions, Nor will I ever.

Most folks here (including myself) are trying to get answers from others that may have experience in solving similar problems or just keep up on pertinent topics. Some (again including myself) offer clean, friendly, fact based assistance for the sake of perpetuating our hobby...

You sir have been seemingly argumentative and somewhat combative from the beginning of your posts... what is your goal here?[/quote]

The goal, appears to be, to troll.

For a Wikipedia essay on the topic, see meta:What is a troll?
In Internet slang, a troll (/ˈtroʊl/, /ˈtrɒl/) is a person who sows discord on the Internet by starting arguments or upsetting people,[1] by posting inflammatory,[2] extraneous, or off-topic messages in an online community (such as a forum, chat room, or blog) with the deliberate intent of provoking readers into an emotional response[3] or of otherwise disrupting normal on-topic discussion.[4]

 

[spunrdneck]

The goal, appears to be, to troll.

For a Wikipedia essay on the topic, see meta:What is a troll?
In Internet slang, a troll (/ˈtroʊl/, /ˈtrɒl/) is a person who sows discord on the Internet by starting arguments or upsetting people,[1] by posting inflammatory,[2] extraneous, or off-topic messages in an online community (such as a forum, chat room, or blog) with the deliberate intent of provoking readers into an emotional response[3] or of otherwise disrupting normal on-topic discussion.[4]

Troll indeed, I have never seen this written out before... kinda sad it happens enough to have a title!

 

[damoncooper]

Kids, kids....

 

[droneranger]

Any updates from DJI on a potential release date for 3.05? I understand that another beta is going to be released to the testers, but at the very least I think DJI should release 3.05 as it is and then release 3.06 whenever they get the next set of tweaks dialed in. From the tester's reports, 3.05b seems to be a huge improvement and as one who has been grounded for over 2 months, I'm getting very impatient because my quad is DANGEROUS to fly in its current state. Why can't we at least get the current fix while we wait for the "improved" version DJI?

 

[SilentAV8R]

as one who has been grounded for over 2 months, I'm getting very impatient because my quad is DANGEROUS to fly in its current state.

Seriously?? I live just south of you and I fly almost every day. I put in 12 flights since Friday. How is it that you consider flying your Phantom to be DANGEROUS!!?? Mine flies fine. I use the sticks to correct any deviation due to compass or GPS errors.

 

[droneranger]

as one who has been grounded for over 2 months, I'm getting very impatient because my quad is DANGEROUS to fly in its current state.

Seriously?? I live just south of you and I fly almost every day. I put in 12 flights since Friday. How is it that you consider flying your Phantom to be DANGEROUS!!?? Mine flies fine. I use the sticks to correct any deviation due to compass or GPS errors.

Yes, seriously.

I'm glad your Phantom flies fine. That is obviously not the case for many of us. Mine has a severe tracking error to the left and a huge J-hook and TBE when stopping from forward flight. Much of my flying is done in relatively confined areas where the flight control must be precise. I know how to pilot the quad and put in corrective inputs, but that precludes getting the footage I want. It is dangerous because if I clip an obstruction due to the quad's poor handling characteristics and unpredictable movements, it could potentially deviate from the intended flight path and injure someone who would otherwise be well out of the way of my desired course.

Finally, just because your "just south of me" (it is nearly 100 miles to the nearest OC location from here) means very little when it comes to deviation. Local deviations can be very big over fairly short distances. That is why DJI recommends that you re-cal the compass whenever you change locations by any significant amount.

 

[sar104]

as one who has been grounded for over 2 months, I'm getting very impatient because my quad is DANGEROUS to fly in its current state.

Seriously?? I live just south of you and I fly almost every day. I put in 12 flights since Friday. How is it that you consider flying your Phantom to be DANGEROUS!!?? Mine flies fine. I use the sticks to correct any deviation due to compass or GPS errors.

Yes, seriously.

I'm glad your Phantom flies fine. That is obviously not the case for many of us. Mine has a severe tracking error to the left and a huge J-hook and TBE when stopping from forward flight. Much of my flying is done in relatively confined areas where the flight control must be precise. I know how to pilot the quad and put in corrective inputs, but that precludes getting the footage I want. It is dangerous because if I clip an obstruction due to the quad's poor handling characteristics and unpredictable movements, it could potentially deviate from the intended flight path and injure someone who would otherwise be well out of the way of my desired course.

Finally, just because your "just south of me" (it is nearly 100 miles to the nearest OC location from here) means very little when it comes to deviation. Local deviations can be very big over fairly short distances. That is why DJI recommends that you re-cal the compass whenever you change locations by any significant amount.

Could you post the gains that you are using?

 

[xtonex]

as one who has been grounded for over 2 months, I'm getting very impatient because my quad is DANGEROUS to fly in its current state.

Seriously?? I live just south of you and I fly almost every day. I put in 12 flights since Friday. How is it that you consider flying your Phantom to be DANGEROUS!!?? Mine flies fine. I use the sticks to correct any deviation due to compass or GPS errors.

Yes, seriously.

I'm glad your Phantom flies fine. That is obviously not the case for many of us. Mine has a severe tracking error to the left and a huge J-hook and TBE when stopping from forward flight. Much of my flying is done in relatively confined areas where the flight control must be precise. I know how to pilot the quad and put in corrective inputs, but that precludes getting the footage I want. It is dangerous because if I clip an obstruction due to the quad's poor handling characteristics and unpredictable movements, it could potentially deviate from the intended flight path and injure someone who would otherwise be well out of the way of my desired course.

Finally, just because your "just south of me" (it is nearly 100 miles to the nearest OC location from here) means very little when it comes to deviation. Local deviations can be very big over fairly short distances. That is why DJI recommends that you re-cal the compass whenever you change locations by any significant amount.

The fix is for tbe and j hook and in return course lock thats not straight. if i keep moving forward it doesnt hook. and since i know it will hook once i let go of the stick i make sure i dont let go until ive gotten into a space where a hook wont hit anything or i correct it before it happens. and thats not even mentioning after a few mins it cleans up very nicely and i can then fly how ever i want. If your in danger of clipping something or worried about the jhook into a crowd where its possible to hurt people then your flying to close to things and people anyways

Also DEVIATION is NOT the problem even being worked on here. deviation is what is solved by you doing the imu calibration and compass dance in the field. DECLINATION is whats being worked on here in the firm.

Everyone understand that you want the fix, theyre not gonna be releasing the beta in the meantime... they will release it when the problem is solved. Until then asking about it or claiming you had to "shelf" your phantom isnt doing u, us or dji any good

 

[droneranger]

xtonex,

Since you don't know the specific conditions in which I am flying and the proximity of objects and people to my flight path, it is presumptuous of you to assume I am too close to either. If a rotor is damaged or broken by contact with something, the quad can seriously deviate from its planned path at high speed. People that were otherwise well out of the way can be in jeopardy in certain situations.

As far as deviation vs. declination is concerned, 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).

Finally, I am not claiming anything. It is a fact that other than for testing purposes, my quad is effectively grounded because it cannot currently perform as DJI says it should. I have every right to voice my opinion and displeasure regardless of how it appears to you.

 

[droneranger]

Sure sar104......I'll post those gains as soon as I am back at the "hangar".

 

[SilentAV8R]

Finally, I am not claiming anything. It is a fact that other than for testing purposes, my quad is effectively grounded because it cannot currently perform as DJI says it should. I have every right to voice my opinion and displeasure regardless of how it appears to you.

Well, if you are not flying due to deviation errors you will continue to be disappointed since DJI is not working on a fix for that. In addition, I seriously question your judgement if you are indeed flying that close to people, objects, etc. regardless of the platform. So, yes, we all have a right to our opinion and we all have a right to provide our opinion on those opinions

 

[Andy T]

Finally, I am not claiming anything. It is a fact that other than for testing purposes, my quad is effectively grounded because it cannot currently perform as DJI says it should. I have every right to voice my opinion and displeasure regardless of how it appears to you.

Well, if you are not flying due to deviation errors you will continue to be disappointed since DJI is not working on a fix for that. In addition, I seriously question your judgement if you are indeed flying that close to people, objects, etc. regardless of the platform. So, yes, we all have a right to our opinion and we all have a right to provide our opinion on those opinions

I certainly hope the fix addresses deviation + declination! and Droneranger is correct in saying the thing don't fly as it should and is doing the right thing by complaining here about it. I also think this spat has played itself out now.. lol

 

[droneranger]

Finally, I am not claiming anything. It is a fact that other than for testing purposes, my quad is effectively grounded because it cannot currently perform as DJI says it should. I have every right to voice my opinion and displeasure regardless of how it appears to you.

Well, if you are not flying due to deviation errors you will continue to be disappointed since DJI is not working on a fix for that. In addition, I seriously question your judgement if you are indeed flying that close to people, objects, etc. regardless of the platform. So, yes, we all have a right to our opinion and we all have a right to provide our opinion on those opinions

I certainly hope the fix addresses deviation + declination! and Droneranger is correct in saying the thing don't fly as it should and is doing the right thing by complaining here about it. I also think this spat has played itself out now.. lol

Agreed!!

 

[sar104]

Finally, I am not claiming anything. It is a fact that other than for testing purposes, my quad is effectively grounded because it cannot currently perform as DJI says it should. I have every right to voice my opinion and displeasure regardless of how it appears to you.

Well, if you are not flying due to deviation errors you will continue to be disappointed since DJI is not working on a fix for that. In addition, I seriously question your judgement if you are indeed flying that close to people, objects, etc. regardless of the platform. So, yes, we all have a right to our opinion and we all have a right to provide our opinion on those opinions

I certainly hope the fix addresses deviation + declination! and Droneranger is correct in saying the thing don't fly as it should and is doing the right thing by complaining here about it. I also think this spat has played itself out now.. lol

if the deviation issues are due to local magnetic field variations then I'm not sure how that is really fixable, since it would require a map of the variations. How do you see that working?

 

[Andy T]

if the deviation issues are due to local magnetic field variations then I'm not sure how that is really fixable, since it would require a map of the variations. How do you see that working?

Using a declination map wouldn't fix it but a quick learn and save method would. My android seems to be able to point to true north after traveling 10 or 20 meters and is rock solid after that. Of course I'm not going to go to the trouble of proving that but is sure seems that way. Doubt it uses a declination map

 

[sar104]

if the deviation issues are due to local magnetic field variations then I'm not sure how that is really fixable, since it would require a map of the variations. How do you see that working?

Using a declination map wouldn't fix it but a quick learn and save method would. My android seems to be able to point to true north after traveling 10 or 20 meters and is rock solid after that. Of course I'm not going to go to the trouble of proving that but is sure seems that way. Doubt it uses a declination map

Actually I'm quite sure that it does use a declination map of some kind - all the common chips include onboard declination correction. The problem with more local variations is the potentially disparate length scales involved. Fixed variations, due to aircraft equipment, are corrected by compass calibration, but trying to map and store fine-scale local deviations would be most impractical, especially if the variations are altitude, as well as location, dependent.

 

[ianwood]

Either some of us have different definitions for deviation or people might be getting a little confused. The compass calibration that you currently do is for deviation. It measures a full circle in two axes to get the "background" magnetic fingerprint. That includes all your gear that you've stacked onto your Phantom and the environment you happen to be flying in.

Other than local field calibration, there is no other way to compensate for deviation.

Andy, depending on which Android device you have and how old it is, it could be using GPS derived course which is only determined after you establish a clear pattern of motion. Or, it is using an internal compass in your phone.

 

[Andy T]

Actually I'm quite sure that it does use a declination map of some kind - all the common chips include onboard declination correction. The problem with more local variations is the potentially disparate length scales involved. Fixed variations, due to aircraft equipment, are corrected by compass calibration, but trying to map and store fine-scale local deviations would be most impractical, especially if the variations are altitude, as well as location, dependent.

On second thought yes you are probably correct and if they decide to use local declination for correction then it wont correct for the local deviation. I thought I had to move before my android made the correction but I may have been mistaken.

 

[Andy T]

Either some of us have different definitions for deviation or people might be getting a little confused. The compass calibration that you currently do is for deviation. It measures a full circle in two axes to get the "background" magnetic fingerprint. That includes all your gear that you've stacked onto your Phantom and the environment you happen to be flying in.

Other than local field calibration, there is no other way to compensate for deviation.

Andy, depending on which Android device you have and how old it is, it could be using GPS derived course which is only determined after you establish a clear pattern of motion. Or, it is using an internal compass in your phone.

yeah deviation has a couple of definitions.. I guess we are not talking about the on board deviation or deviation caused by metal objects around us, but rather the deviation caused by some geological formation in the area causing a fairly widespread deviation from the expected declination. Not sure if it is a big issue anyway

 

[Pazz]

Either some of us have different definitions for deviation or people might be getting a little confused. The compass calibration that you currently do is for deviation. It measures a full circle in two axes to get the "background" magnetic fingerprint. That includes all your gear that you've stacked onto your Phantom and the environment you happen to be flying in.

Other than local field calibration, there is no other way to compensate for deviation.

Andy, depending on which Android device you have and how old it is, it could be using GPS derived course which is only determined after you establish a clear pattern of motion. Or, it is using an internal compass in your phone.

Nope, What you are referring to is Declination ( local magnetic interferences) . Thats what compass calibration Zeros out. Deviation(or variation) is the difference between True north & magnetic north, which varies depending on location & time (as it drifts over time) Compass calibration does not correct this.
Suspect now the issue may be the fact the GPS is set to reading true north values ( same as a map) & the compass magnetic north values. However can you explain the dynamic relationship, in the auto pilot of the NAZA software, of the GPS, compass & barometer. are they connected?, do they give positive feedback or do they operate independently ?ie. gps for position, compass for heading & barometer for height. Do you have any insight it to the science behind how an auto pilot system works, as there seems to be very little material on this on the net.