After getting help finding your thread on yaw drift I need to know if when you were recalibrating the IMU with the basic adjustment or the advanced. I would like to see if I can improve on what my Phantom V3 is doing.
Thanks Joel

I'm sure he was calibrating the advanced imu.

Advanced calibration for sure.  FYI, after almost 3 months, numerous flights and the firmware upgrades, all is still good.  I have no need to calibrate the IMU, and won't unless something changes.

As I mentioned on yesterdays thread, the advanced IMU calibration helped my Yaw issue.  Good luck

Thanks guys, I only did the basic on a dead level surface. Will do the advanced tonight. Flew the 3rd battery today after the 3.12 update and all systems worked well. I wanted to make sure that there were not any issues with the update before I tried to correct the yaw issue. It's started raining now so it might be a few days before I get this done.
Joel

joelpoly@msn.co Posted at 2015-3-21 07:55
Thanks guys, I only did the basic on a dead level surface. Will do the advanced tonight. Flew the 3r ...

Please post your results.  The shimming to correct my drift required business cards stacked as such:

Left Front Pad    16  cards              Right Front Pad  0  cards

Left Rear Pad     19 cards               Right Rear Pad   3

And that was performed on a level surface.

I have been intending to pull out the out of kilter NAZA module and level it with the craft to see if that is the culprit, but this fix worked so well, I haven't messed with it.

This picture shows how out of level the NAZA module is to the Phantom.  A tungsten rod is laid across the NAZA module with a referenced level box behind.  You can see how the rod ends are off level with the motors, as the aircraft is off level due to the cards under the landing pads.

I "believe" the key is that the NAZA module needs to be level during calibration, and my NAZA module is not anywhere near level with the Phantom body.  So the card shimming appears to correct that, and subsequently the yaw drift.

I am disappointed that with all of the discussion of this issue, that someone from DJI has not stepped in and commented on this.  It would seem to be either "Yes, the NAZA Module needs to be mounted level with the craft, and the craft level when IMU calibration is performed.", or "No, here is the reason for the yaw drift, ........".

If you have one, a pool table is probably the best place to start if it was leveled when installed.

Jamie Posted at 2015-3-21 10:33
Joel,

Maybe you could edit or rename the Post Title to make it more relevant to the content.

Wow I didn't expect to see that big of a change, or that many cards.  When I did the basic calibration I used my phantoms case with a pain of glass on top and a spot level. I marked the case for the glass. I then marked the glass for the placement of the level and the phantoms landing gear so it would always be in the same place. I used index cards to level the glass on the case. Now planning to use cards to shim up left side to stop the drift to the right. Does that sound like it would work?

joelpoly@msn.co Posted at 2015-3-21 12:18
Wow I didn't expect to see that big of a change, or that many cards.  When I did the basic calibra ...

You can see how far my Phantom is out of level with the cards under the pads, yet the NAZA module is level with that shimming.  So if in fact the NAZA module needs to be level when calibrating the IMU, leveling the Phantom may do no good if the module is not mounted level with respect to a level aircraft.

I suspect that the NAZA modules are calibrated on the bench, then installed into the Phantom on the assembly line.  If the module is not installed level with respect to the Phantom, then the next IMU calibration will occur with an out of level NAZA module, hence the cause of the Yaw Drift.  I believe this is what happened in my case, and my problem with drift started after I did the IMU calibration.

But that is just my theory.

New guys....Don't calibrate the IMU unless the software tells you to, or you are experiencing this Yaw Drift or other IMU related problems.  Just because you can do something, or do something you think is cool... if it ain't broke don't fix it.

Joel, mine drifted 10 to 15 feet to aircraft Starboard/Right during a counter-clockwise pirouette (yaw turn) during a hover, and I had to shim up the Port/Left side to correct it.  It also drifted slightly forward, and the rear shims corrected that.

It took me around 12 to 15 shim trials and test flights, and a couple of hours to accomplish.  Track the changes in shims on paper as you go.  I did not the first half dozen trials or so and it got confusing.

I would be curious to know if anyone cursed with the Yaw Drift has just leveled the NAZA module in the craft, or shimmed up the craft to level the module, then calibrated the IMU with a successful conclusion.  That would save a lot of trial flights and calibrations.  Mine is flying, hovering, and performing pirouettes so well at this point, I dont' want to mess with it.

Im curious too as if they want the naza module level when calibrating the imu or the phantom itself.

It would be nice if the default values from the factory would be available, so that one could reset them, if needed.  Store them in memory, and if an advanced calibration doesn't correct the issue, there is something to fall back on.  RESET...

jtrager@carolin Posted at 2015-3-22 02:38
It would be nice if the default values from the factory would be available, so that one could reset  ...

Believe it or not, electronics do wear!, and every component in these birds has a tolerance which means that although they all look same...they are all different,,,,,a baseline as a last resort might be o.k, but certainly not accurate enough to rely on.