Jul 15, 2012, 10:59 AM
Im the eye In the sky
Denmark, Region Zealand, Jyderup
Joined Mar 2012
251 Posts
Quote:
 Originally Posted by tritan So your CG is exactly in the middel of your multi like in the picture and not below ??? Come on now .... What happens if you turn your multi upside down - where is your CG now?
WE asspect that you have placed your IMU on COG without no weight and all! so the actual COG is where the IMU is...
But when u add weight then u move the COG, and that is why I am tempering with the numbers to tell the IMU that I have a new COG. ? Comprende?
Jul 15, 2012, 11:00 AM
Registered User
Joined May 2011
372 Posts
Quote:
 Originally Posted by nicnaimless I don't disagree with your statement here. But when you change the X Y Z values, you are telling the WKM that the IMU has been moved from the center of the craft by that much. You are not telling the IMU that you have uneven weight distribution. Like I said, DJI assumes you balanced your multi and they decided to call center of craft Center of Gravity. All the IMU understands is where is the center of your craft.
I think center of craft for the WKM is center of gravity because all vector forces acting will try to manipulate the attitude of the craft around this virtual fulcrum since there aren't any physical ones. The reason the WKM needs to know the distances of the IMU from the CoG so it can calculate what the input accelerometer/gyro values mean. the further away the IMU is from the fulcrum point of rotation, the more the acc vector value will be as the IMU moves more and faster the further it is from the center of rotation(arc distance = 2xPixIMU distance from center in the given axis * sin of the angle rotated)

Just a throught based on physics...i don't actually know how it works
 Jul 15, 2012, 11:00 AM DJI Testing / US Support United States, CA, Big Bear Joined Jan 2006 21,541 Posts Guys, he is a littel loss just drop it.. I dont want people to think it's ok to not input these values... If you are going to use the GPS feature of the WK-M please use the xyz values... If not then dont input them If it works for you then that's ok, but that does not mean on another airframe that it will work! Latest blog entry: DJI GO App Tutorial - Coach Mode
Jul 15, 2012, 11:01 AM
Registered User
United States, MI, Jenison
Joined Jun 2008
1,149 Posts
Quote:
 Originally Posted by jab They what about stabilization when flying in wind gusts and such things? Does the WKM think about stabilization as weights moving around the frame while flying? I agree with epedemi, and the math does to. A IMU does not care about weight distribution, it only sees rotation and acceleration and wants to know the center of the axes involved. The motor mix then has to compensate for any uneven weight distribution to keep the frame level, but the IMU could not care less about this.
You've got it Jab! That's where folks have been going wrong. It's DJIs fault for writing CoG in the manual.
Jul 15, 2012, 11:01 AM
Registered User
Joined Aug 2011
168 Posts
Quote:
 Originally Posted by jab They what about stabilization when flying in wind gusts and such things? Does the WKM think about stabilization as weights moving around the frame while flying? I agree with epedemi, and the math does to. A IMU does not care about weight distribution, it only sees rotation and acceleration and wants to know the center of the axes involved. The motor mix then has to compensate for any uneven weight distribution to keep the frame level, but the IMU count not care less about this.

Jab, when the wind affects the gyro, its actaully throwing the CofG off, thats instability, like adding weight to one end,therefore it compensates for that to level the craft, thats why its working harder in gustier conditions.
Jul 15, 2012, 11:01 AM
DJI Testing / US Support
United States, CA, Big Bear
Joined Jan 2006
21,541 Posts
Quote:
 Originally Posted by nicnaimless Yes, I adjust them accordingly to the center of the aircraft then I leave them alone unless I move the GPS mod or IMU. I do not change the values if I change the weight of the craft or add a piece of equipment such as different cameras.
Then you are doing it wrong

It's part of the flight codes logic.. There is no translation error, If you have been in the hobby that long you would know that they fully understand CoG!
Jul 15, 2012, 11:02 AM
Registered User
Joined Apr 2007
2,331 Posts
Quote:
 Originally Posted by epedemi WE asspect that you have placed your IMU on COG without no weight and all! so the actual COG is where the IMU is... But when u add weight then u move the COG, and that is why I am tempering with the numbers to tell the IMU that I have a new COG. ? Comprende?

This is what we all have been trying to tell you ..... we want to help you ok?
Jul 15, 2012, 11:02 AM
Registered User
United Kingdom, England, Blackpool
Joined Aug 2011
300 Posts
Quote:
 Originally Posted by jab They what about stabilization when flying in wind gusts and such things? Does the WKM think about stabilization as weights moving around the frame while flying? I agree with epedemi, and the math does to. A IMU does not care about weight distribution, it only sees rotation and acceleration and wants to know the center of the axes involved. The motor mix then has to compensate for any uneven weight distribution to keep the frame level, but the IMU count not care less about this.

Would DJI care to comment? Sidney?

One question, why do Aircraft/Drone Flight Controllers/Gyros need to know their craft Cgs.....
Jul 15, 2012, 11:05 AM
DJI Testing / US Support
United States, CA, Big Bear
Joined Jan 2006
21,541 Posts
Quote:
 Originally Posted by blade strike The xyz has NOTHING to do with atti or manual mode.. So you are right kind of. The senors dont care until you thorw the GPS / Compass in the Mix It's how DJI's fllight logic works.. If you want solid GPS(PH, RTH) you need these values to be close. If it works for you then thats great but dont tell people they are reading the manual wrong. There are MANY videos showing how to get CoG for the WK-M
1 more time
 Jul 15, 2012, 11:07 AM Registered User United States, CO, Denver Joined Dec 2002 939 Posts I'm not so sure Sidney wants to post in this thread any more
Jul 15, 2012, 11:07 AM
Im the eye In the sky
Denmark, Region Zealand, Jyderup
Joined Mar 2012
251 Posts
Quote:
 Originally Posted by tritan This is how I have understand it: your CG is below your multi right? Ok, lets say it is about on top of the red "Z" in your picture of assistant, about on the "roof" of the red "Z". So if you move everything , the CG white/black cirle, all red and green lines down so now the CG stands on the "roof" of the red "Z" your IMU is ABOVE the CG circle ok? Green is negative. Same for GPS, above which is green which is negative. I know of none who has a positive value for the GPS.
Yeah I understand now....
Sorry my english is not my native language...
I rather do math on a diagram- like i have shown u.
Jul 15, 2012, 11:10 AM
Im the eye In the sky
Denmark, Region Zealand, Jyderup
Joined Mar 2012
251 Posts
Quote:
 Originally Posted by R Hudson I'm not so sure Sidney wants to post in this thread any more
I would be scared shiiitless 2 ;-)
Jul 15, 2012, 11:12 AM
Gravity sucks :)
Indonesia
Joined Jul 2005
935 Posts
Quote:
 Originally Posted by snurre FW 5.12 for octa SAFETY ALERT I have just upgraded WKM firmware from 5.08 to 5.12. Octacopter "I", batt: 4S, ESC: CC Phoenix ICE 50 fw4.01 (set for MR, outrunner) "Old" type WKM controller (december 2011) After upgrading the fw I went through all tabs and reset /wrote the settings, new TX calibration included. ISSUE, on ground, transmitter on, all motors should be at rest / disarmed / quiet: Motors 7 & 8 periodically kick in . Very short bursts, but enough to turn props around a few turns (Youtube video upload in progress). During this workshop session the octa was supplied by a power supply featuring adjustable current limit. The current limit was set to limit current to just a few amps, as a safety precaution should any surprise occur. It seems like sometimes during these power bursts the current limitation kicked in (Spectrum telemetry signalled low batt). Therefore I do not know how much power would have been employed into the props if it had been supplied by the regular high power drive batteries. And I feel reluctant to try that out, since I now strongly suspect a FW bug in 5.12 for octas. With 5.08 this never occured (burst when all motors should be quiet). I cannot rate the severity of this suspected bug, but any safety related bug should be treated with great caution, even if not 100% verified. And we all know that 5.10 had a bug that made octas not work at all. Maybe that bug was not completely resolved and 5.12 was not thoroughly tested on octa? / Tomas
Hi, I see you're using CC, have you disable the "Power-on Beep" (connect the ESC to the Castle Link under the "other" tab) on every ESC?
Jul 15, 2012, 11:12 AM
DJI Testing / US Support
United States, CA, Big Bear
Joined Jan 2006
21,541 Posts
Quote:
 Originally Posted by epedemi Ok guys... I dont think perfect is a term in this business... But I think I have achieved something very close to it. Firmware 5,12 This is in windy conditions. Gps-atti mode. Frame = Dji F550 AUW = 2,5 kilos. Graupner 10x5 Maytech esc = 30amp. Zippy flightmax 5800 Torxpower ( tigermotor ) 2216 ( 900kv ) Gymbal enabled.

Those values are pretty close to mine now.. I did start to play with the (I) value to stop some unwanted wobble in fast decent.. It is set at 90, right now
 Jul 15, 2012, 11:12 AM Registered User United Kingdom, England, Surrey Joined Jan 2012 433 Posts O.K as I said yesterday, I would move my GPS unit up and recalibrate. Then try POI mode and see how it flew generally? Well it certainly made all the difference in the world.. It is so much better now, and everything works much more stable. So perhaps I was too hasty in blaming DJI? but if I had access to flight logs? I would have seen straight away where the problem came from.. Maybe something to consider? and it still isnt as stable as my friends Naza GPS which I set up for him?