Beware of Smart Phone compass readings

Discussion in 'Cinestar 8' started by Dave King, Jan 8, 2013.

  1. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    I went to the Army store and got a real magnetic compass. I compared the compass values to both an Android compass app and a smart phone compass app. See the pictures for the results. Here in Pittsburgh Magnetic north is about negative degrees in relation to geographic north. It appears that the Android app is definetely pointing toward geographic north by 9 degrees compared to the compass. Next I borrowed my friends IPHONE and used the IPHONE app that comes with the phone. You can see that north is way off in relation to magnetic north. I then downloaded 2 more android compass apps which showed closer results compared the magnetic compass but they are off about 4-5 degrees. Only one of these apps asked me to configure the compass doing a figure 8 pattern.

    I also checked each setting by itself with the other devices away from each other to make sure that there was no interference going on and the results did not change. I also made sure that I was away from anything magnetic where I placed the compasses.

    CONCLUSION: Get a good magnetic compass to calibrate your rotorcopter!!! IF you think about it this way, if the smart phone compass can be off as much as 15 degrees, and you add in the variance of actually pointing the booms to true magnetic north can be 10 to 15 degrees, the compass on the copter can be as much as 30 degrees or more off.
     

    Attached Files:

  2. Gary Haynes

    Gary Haynes Administrator
    Staff Member

    Joined:
    Jul 6, 2012
    Messages:
    5,211
    Likes Received:
    460
    Dave not sure where you are in Pittsburgh but the magnetic declination for about the 50 yard line center field at Heinz stadium is -9.25 degrees from true north.
     
  3. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,310
    Likes Received:
    1,153
    Dave:
    I suspect that you're on to something, but I do have to point out that the Smartphones do have magnets in them so you may find that the proximity of the iPhone, for example, will cause the simple magnetic compass needle to swing. I just did a quick test and can swing a compass about 15 degrees just by bringing my iPhone the same distance away as you show in your images.

    That said, I believe the iPhone's compass is calibrated for use in the iPhone -- but I really don't know how accurate it is.

    I think we need to do some more testing, perhaps using geographic landmarks and Google satellite imagery to verify -- but I wouldn't put a simple magnetic compass next to a Smartphone!

    An alternative means might be to look at Google satellite view, establish true North, and then find out the magnetic variation (aka declination) for that location. Here's such a map: http://www.windtesting.com/4.html

    Andy.
     
  4. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    Yeah most of Pittsburgh has a negative 9 degree magnetic declination.
     
  5. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    I understand what you are saying but I did the tests by themselves as well with no magnet next to phones and the results were the same. Maybe I can try to print out a google map with pin point accuracy of geographic north and I can verify the results.
     
  6. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,310
    Likes Received:
    1,153
    Dave: Yeah, if you see that map I linked to you can see you're pretty much smack in between the 8 and 10 lines.

    Andy.
     
  7. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    OK, see attached pics. I picked a good place near by that I can do some testing. I saw the map that Andy posted above and I am graphically sandwiched in between lines 8 and 10. I used this site to get an exact declination angle
    http://magnetic-declination.com/

    From there I created a X and Y axis for geographical north. After I created my X and Y axis I printed the picture out and used a compass to draw the actual magnetic target line 9 degrees from geographical north to give me an exact aiming target. I picked the pitchers mound on the baseball field because it was a easy reference point to accurately find. Now I can do some testing with different apps for droid, and Iphone. I have a droid and my wife has an IPHONE so I can test both. IF anyone has any specific apps that they want me to test please let me know and I will download the apps to the cooresponding phone. I will test each app against the true magnetic compass as well and I will make sure that the magnetic compass is at far away from the phones when I test the phone apps. I should be able to test this Saturday or Sunday. Early weather predictions for Saturday are 63 degrees!!!!
     

    Attached Files:

  8. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    I added Magnetic east to the reference point which will help me when it comes time to recalibrate the compass which I will do after I verify all the data.
     

    Attached Files:

  9. Steve Maller

    Steve Maller UAV Grief Counselor

    Joined:
    Oct 30, 2012
    Messages:
    3,979
    Likes Received:
    807
    Seems there's an app opportunity there to mashup this data with Google Maps' data and produce a location-accurate compass calibrator. But maybe somebody should reach out to Holger (or others) and determine if this is a red herring, or if it's a real issue. But thanks for the detective work, Dave!

    Edit: did a little poking around, and there's a $0.99 iPhone app called "Magnetic Declination" that looks promising.
     
  10. Gary Haynes

    Gary Haynes Administrator
    Staff Member

    Joined:
    Jul 6, 2012
    Messages:
    5,211
    Likes Received:
    460
    Dave you could also get a reference check by measuring the angles of the baseball first and third base lines. Google sat photos don't always match exactly and you have some really natural compass points that you can check.

    I also agree with Steve that this isn't something that needs to be to a zero tolerance. I would guess that withing +\- 10 degrees would work. Other wise we would all be building some type of frame to hold the copter while we spun it around on the axis. It is an awkward exercise at best.

    I use multiple apps on the iPhone that provide compass bearings and declination info

    Theolodite - Hunter Research
    Declination - amphibious technologies Immediately know your mag declination for adjustment
    Free HD Compass -Imaginatr
    Compass - OnDemadWorld This one seems to be the most accurate for me
    Gaia GPS - GAIAGPS.COM I use this for offline topt maps and True North is also very accurate but doesn't have a compass mode
     
  11. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    Thanks for the replies. IF you look at some of the pictures in the first post of this thread some of the results were more than 10 degrees (up to 30 degrees by eyeballing) from actual magnetic north that the real compass was showing. 30 degrees is a lot, now add in the slop factor of the actual calibration of the copters compass which could be 30 degrees???? and now you have a 60 degree variance. It would be nice to hear from someone at microkopter for feedback. What drew me to this experiment is that there seems to be a good number of people that are having issues once they hit PH CH or AH. Now Andy says he calibrates the compass for every location but I do believe by looking at the chart that if you are not going far distances your compass calibration should be really close.
     
  12. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,310
    Likes Received:
    1,153
    Hi Dave:
    The problem is that there are often local magnetic anomalies that will otherwise cause problems. I've seen them in the form of railings, buried metal pipes, power lines, etc. etc.

    Thus you cannot rely on web-sites to tell you specifically what's going on at a particular place (although it's good to know of the http://magnetic-declination.com/).

    The position hold depends on the accuracy of the compass calibration so, to reduce the flying in circles when on PH (aka "toilet bowling"), I simply re-calibrate every time [Edit: every time I go to a new flying location] rather than discover when I'm airborne that I should have calibrated. :)

    Andy.
     
  13. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    Are there any apps you would like me to compare?
     
  14. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,310
    Likes Received:
    1,153
  15. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    I went out yesterday to the field and did some comparisons. I have interesting results. I'll post tomorrow with pictures.
     
  16. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    OK so I did some testing back to the ball field over the weekend and did some measurements with android apps. I would have tested spy glass but I didnt' know my wifes apple ID and didnt want to waste the good weather. First thing I did was use a real magnetic compass and verified magnetic north and compared it to the magnetic maps I used previously last week. It confirms that magnetic north is about negative 9 degrees plus or minus a degree. I couldn't use the pitchers mound as a starting point because was very muddy!! But I lined up the pitchers mound to home plate in the outfield and picked the best alignment point I used for reference. I then marked magnetic north with a yard stick and then compared that to the apps I had. Before I started I put the magnetic compass back in the car and far away from the phone. I had one app off as much as 35 degrees. I attached pictures for reference.
     

    Attached Files:

  17. Andy Johnson-Laird

    Andy Johnson-Laird Administrator
    Staff Member

    Joined:
    Jul 31, 2012
    Messages:
    10,310
    Likes Received:
    1,153
    Interesting information, Dave. I wonder how the iPhone would have compared? The issue is really how the "smart" phones' Hall effect compasses are calibrated. The iPhone allegedly does continuous calibration -- not sure how the Samsung's compass works. But the lack of repeatability is worrying, given that it was the same phone, same place and more or less the same time.

    Andy.
     
  18. Dave King

    Dave King Well-Known Member

    Joined:
    Dec 24, 2012
    Messages:
    2,711
    Likes Received:
    311
    I will try to go back this weekend and try at least the stock Iphone compass app and eye spy as well as the other andriod apps.
     

Share This Page