1. EgonW3

    EgonW3 New Member

    Joined:
    Jun 20, 2011
    Messages:
    37
    Location:
    Poland
    I've just found quite simple method to detemine Z-axis position of COG:

    1. Determine where is COG on X and Y axis. I think nobody has a problem to find it.
    2. Tilt your copter on the edge of landing skids until you find a balance (copter doesn't tend to flip to right or left)
    3. Make a picture. In my copter green line is where I found my COG on X and Y axies.

    COG.jpg

    4. COG is where red line crosses green line

    What do you think about such a method?
     
  2. eyeball

    eyeball Member

    Joined:
    Aug 4, 2011
    Messages:
    118
    Location:
    Cheshire England
    I like the idea EgonW3....but I dont think this will give you your Z axis as your will only find the balance point of the copter in relation to the front edge of the landing skid. For example if you were to shorten the length of the skid and redo the same test, you would find the C/G moving even lower down because you would have to lean the copter further forward to the right (off the edge of the table) inorder for it to balance again.
    I may stand corrected... or have understood you wrongly. Good try though!!:apathy:
     
  3. kauaicuda

    kauaicuda New Member

    Joined:
    Oct 12, 2011
    Messages:
    10
    I just tie a string to the arms and hang it from a doorway, then the same MS paint lines, works great.
     
  4. EgonW3

    EgonW3 New Member

    Joined:
    Jun 20, 2011
    Messages:
    37
    Location:
    Poland
    Your bethod is based on idea that when an obiect is suspended, so that it can swing freely, it will come to rest with its COG vertically below the point of suspension.
    My method is based on idea that obiect become unstable when its COG falls outside of the obiect's base.

    Both methods should give the same results, but my method doesn't require a string and because COG is usually quite close to the surface of the arms arms, you get quite sharp angle so it is more difficult to determine it precisely.
    Your method may give you COG position on two axies (by rotating hanging copter by right angle and making another photo).
     
    Last edited by a moderator: Jun 20, 2012
  5. EgonW3

    EgonW3 New Member

    Joined:
    Jun 20, 2011
    Messages:
    37
    Location:
    Poland
    It will give you the same result despite of edge of base you chose. But in this method it is better to chose the edge that is farthest from predicted COG to achive better accuracy.
     
  6. eyeball

    eyeball Member

    Joined:
    Aug 4, 2011
    Messages:
    118
    Location:
    Cheshire England
    I understand what your saying, I will have a try of this method later.
     
  7. mitsaras

    mitsaras New Member

    Joined:
    Mar 7, 2012
    Messages:
    85
    Location:
    Greece
    Cool idea, between this and the "suspend by a string" method, it should be quite easy to determine COG along the Z-axis. Thanks!
     
  8. janoots2

    janoots2 Member

    Joined:
    Jan 22, 2012
    Messages:
    269
    Location:
    Western, PA
    It's been my experience with the WK that the z axis doesn't matter. When I keep it at the middle of the IMU, my setup is much more solid. I actually tried today to dial in the CG right on the money. My hexa started twitching and swaying. Brought it back down, changed the COG back to the middle of the IMU and it was rock solid again...
     
  9. mitsaras

    mitsaras New Member

    Joined:
    Mar 7, 2012
    Messages:
    85
    Location:
    Greece
    If I understand you correctly, you simply measure the vertical distance from the GPS to the center of the IMU (or Naza, I presume?), and use that as the "Z" value? You ignore the vertical location of the COG?
    Sorry but I am new to this...
     
  10. BorisS

    BorisS Active Member

    Joined:
    May 16, 2011
    Messages:
    3,434
    Location:
    Austria, Vienna
    this might help:



    boris
     
    Last edited by a moderator: Nov 22, 2014 at 3:24 AM
  11. janoots2

    janoots2 Member

    Joined:
    Jan 22, 2012
    Messages:
    269
    Location:
    Western, PA
    Correct, that's what works best on my setup at least. There was a thread a while back with others talking about it too. Yesterday I tried it out and confirmed that yes (at least on my setup), keep the IMU z axis at 0, and then measure appropriately for the GPS z axis (and x and y) off the middle of the IMU - or - the unit is much less stable. Test it out both ways though to find out what works best. My unit still flew and was controllable with the true COG inputted in, but a lot less stable
     
  12. 139 jockey

    139 jockey New Member

    Joined:
    Sep 28, 2011
    Messages:
    40
    Location:
    New Plymouth, New Zealand
    I am with Borris on the c of g, all the lines and strings might work but a simple balance point is exactly what c of g is.
     
  13. mitsaras

    mitsaras New Member

    Joined:
    Mar 7, 2012
    Messages:
    85
    Location:
    Greece
    aha... so DJI gear considers itself to be "the center of the universe"!!!
    :nevreness:

    thx for your help!
     
  14. BorisS

    BorisS Active Member

    Joined:
    May 16, 2011
    Messages:
    3,434
    Location:
    Austria, Vienna
    guys pleas don't just enter 0 for the Z axis value for the IMU there is a reason why you can change that value. Its not rocket science to find a close value.

    Setp 1:


    CXA1.jpg

    1. I choose a imaganery line (in this case blue) which i take as an starting point to consider were theheavier parts are. Blue line is on the booms, because of the prespektiv it dosent work out in this foto.
    • Whats over the blue line ? The motors and the props and some plates and the IMU plus the Herkules ESCs.All probably adds up to 4x KW 139g a piece so 556 g. plus props and plates IMU ESC lets say another200 grams. Total 756 grams.
    • Under the blue line 4 s 6000 mah Nano Tech little under the line the attachment of the PH gimbal withthe GoPro on, Ladinggear reciver Gps etc. Feeling wise i already know I am heavier under the blue line.Nano Tech 623g Gopro 94 grams Microtilt PH may 100g, receiver GPs extension plate etc. maybe another100 grams. Total 917grams

    CXA2.jpg


    2. Simple test am I right with my assumption japp do to the tilt backward there is more weightunder the blue line. Thus my COG has to be somewhere under the blue line.
    • Now to find this exact your frame might give you the chance to hold it on a different spot to balance it out or you might not have the chance to do that than just make an assumption.
    • You can always take the birfd down and increase or decrease the value and see if the flight behaviorget better
    CXA3.jpg

    3. So i persume that my CoG on the z-axis is probably were the yellow cricle is. I have a good starting point.
    • The IMU is what ever the measurement is over the CoG thus i need a negatice values. Lets say the length of the green line is 7 cm thus -7 I will enter in the WKM assistant. 7cm from the CoG to the middle of the IMU
    • The GPS on the Carbon plate is once the Lipo is on horizontal so the GPS modul is spot on on the CoG thusthe value 0.
    In those cases were 0 for the IMU z-axis value works the reason might be that the IMU is already close to the to the CoG. Check your bird you know the weight of your components make an assumption work front there.

    Boris



    ‚Äč


     
    Last edited by a moderator: Jun 23, 2012
  15. janoots2

    janoots2 Member

    Joined:
    Jan 22, 2012
    Messages:
    269
    Location:
    Western, PA
    I'm not saying just enter 0, but if you're unstable when you enter your cog in, it would be a solution to try. My cog is -11 cm from the imu, so its not even close to 0. Input this onto my system and it is very unstable. Keep the cog at the imu at 0 and it's rock solid. Works for me at least. Maybe ill try something in the middle to see if i can even get it better. It is a little twitchy in the wind, thats why i tried to dial it in better in the first place.
     
    Last edited by a moderator: Jun 23, 2012

Share This Page