Unexplained Yaw Drift in GPS Mode

[mangoboom]

OK my Phantom 2 has started this unexplained yaw drift.

In GPS mode if I start to yaw when I get to around 180 degrees it starts to wander off in the direction of the yaw.I then let go of the stick and it returns to where it started the yaw. Previously I've had rock solid performance on yaw and GPS, Just locking onto a spot and sticking there but this will just start to drift quickly before I let go of the stick.

The only thing I can think it might be is a bad motor as one of them has a "grindier" feel than the others however all other aspects of flying are fine.

I've done all the firmware updates etc. Only thing I've recently done is add in mini iosd but if that were the issue then surely other aspects of a flight would crop up?

I've Googled it and found others with the same issue but no resolution from any of the threads I've read.

Any ideas?

Thanks

Matt

 

[dkatz42]

Was there much wind at the time? When you start yawing, the correction the autopilot applies for the wind goes off-kilter until it can come up with a new value, and my guess is that this calculation lags a bit so it drifts. Mine does this--it always drifts in the same way, downwind and at an angle depending on the direction and rate of the yaw (faster yaw yields faster drift at a sharper angle relative to the wind) which is what you would expect.

A bit of this is inevitable with control loops, because they figure things out by watching trends over time, and time, well, takes time--the correction value is always behind.

Full-scale airplane autopilots do the same thing--they hunt a bit after a turn because the wind correction they were using starts to drive them off course.

One could imagine tweaking the flight control algorithm by using the compass sensor (which is nearly instantaneous) to adjust the wind correction while yawing, and suppressing the control loop until the yawing stops, but apparently they don't do that (or at least not very well).

 

[tanasit]

I do realize that it's very difficult to yaw on the spot but in your case I believe the culprit is that defective motor.
The precision of the yaw depends directly on the efficiency of the opposing motors to generate the precise rpm for that.

 

[obiwan_pierogi]

This is one of the reasons I sent my Phantom out for repair.. It used to turn on a dime.. Then as soon as it started to turn, the copter would start wandering.. I also had a major problem with a strong left turning tendency, after a straight run.. GPS wasn't the problem, as I was locked onto 11 satellites.. I wish they could include some sort of sensor to inform of a bad motor.. I haven't heard back from my dealer yet-- 2 weeks now.. Will call him on Monday to see if he found out the problems with my copter.. Should have sent it directly to DJI in Calif., but the response from them (to tell me to ship it to them) took about 2 weeks..

 

[mangoboom]

I'm pretty sure it's the motor now.

I flew yesterday and around a minute into the flight the phantom started pitching so I immediately tried to land but it wouldn't go down straight so I ended up hitting pretty hard so that it cart wheeled and broke the Zenmuse baseplate.

Phantom seemed okay but I've got new bearings on there way so will try that for the next flight.

It's worrying that these things will fall hard if one motor fails.

Ill post how I get on with the yaw and new bearings...

 

[dkatz42]

It's worrying that these things will fall hard if one motor fails.

The price you pay for having only four motors--you can't get full maneuverability with only three. Go to six or eight and you can get redundancy as well as controllability.

The TED talk video had some interesting adaptations to failed motors, such as spinning continuously in yaw while being able to maintain position and altitude (four degrees of control freedom require four motors, so abandon a degree of freedom). In theory the Phantom control system could achieve this, but DJI hasn't done so.

 

[mangoboom]

DJI should license this.

http://www.uasvision.com/2014/03/14/onb ... ithm-demo/

 

[tanasit]

On the bottom I drilled a hole as seen below:

I lubricate the motor bearings with proper high speed bearing oil every 10 flights or so, After several hundreds flight, still no motor problems.

 

[Whiteram]

Great idea! What did you use for bearing oil ?


Sent from my iPhone using Tapatalk

 

[tanasit]

Great idea! What did you use for bearing oil ?


Sent from my iPhone using Tapatalk

Brushless motor oil below:

http://www.horizonhobby.com/products/br ... TRITEP5001

http://www.amazon.com/Brushless-Hi-Perf ... B0032HF79M

 

[Whiteram]

Ty


Sent from my iPhone using Tapatalk

 

[mangoboom]

Okay I just upgraded the motors to Antigravity and I'm still getting this Yaw drift issue in GPS and Atti mode.

The only possible thing I can think of is where I put the IOSD on top of the Can Bus inside the phantom.

See attached image.

What do you think ?

I'll strip it back down and remove the iosd mini and see what that does, but essentially I followed the instructions from here :

https://www.youtube.com/watch?v=TL8pqtEDuos

 

[len750]

Where does one send a Phantom out for repair? Mine works fine, but I would like someone to set up FPV and a gimbal correctly. Would DJI do it? How much would that cost and how long would it take?. I have the Phantom 1.1.1 stock with GoPro hero2.
Flying here in sunny Arizona.

 

[tanasit]

mangoboom, if you think the IOSD may interfere with the CAN BUS, you can try putting the copper tape between them.

I use the copper tape to protect the NAZA from interfering with the GPS in my Y6.

BTW did you check the CG?

 

[mangoboom]

Sorry what is the CG ?

 

[EMCSQUAR]

Sorry what is the CG ?

Center of Gravity

 

[Bigbells]

What is the general location (state, country, etc) where your flights take place?

In the Western US, the Phantom 2 drifts to the left when you release the sticks in GPS mode. In the Northeastern US, the Phantom 2 drifts to the right. This is due to magnetic declination.

In the following thread, started by forum member Ian Wood, MANY Phantom 2 owners are reporting the problem. The intent is to get DJI to address the very prevalent issue:
viewtopic.php?f=19&t=10317

 

[SlackerATX]

I'm having the same problem and I'm kind of losing my mind. I've had a couple crashes and once tweaked one arm. This is what I've done since.

1)Replace the shell, removed everything and put it in a new shell.
2)Verified the compass is calibrated and mod values in assistant are sane. (about 1250)
3)Took one motor apart that seems a littler grinder than the rest. Oiled it and it seems like the rest now.
4)Verified good GPS with 14 satellites.
5)Perfectly balanced all 4 props. (I mean perfect)
6)Verified up close that all 4 props are level while running.

I've thought about doing a few more things, but not sure how.

1)Play with the yaw gain in assistant. I've tried this, but it doesn't seem to change anything. Should I trying increasing up or down?
2)Test the motors for RPS?
3)Test/replace ESC?

I've bought 4 new motors and plan on replacing them tomorrow but I don't have much hope that will fix it. Has anyone had any success fixing this?

 

[EMCSQUAR]

You might try upping the YAW gain. Leave everything else as is.

 

[SlackerATX]

What is the general location (state, country, etc) where your flights take place?

In the Western US, the Phantom 2 drifts to the left when you release the sticks in GPS mode. In the Northeastern US, the Phantom 2 drifts to the right. This is due to magnetic declination.

In the following thread, started by forum member Ian Wood, MANY Phantom 2 owners are reporting the problem. The intent is to get DJI to address the very prevalent issue:
viewtopic.php?f=19&t=10317

I'm in Texas, I don't think I'm having this problem. Magnetic declination is only 4° 17' in Austin. I asked in that thread if they had difficulty yawing in place and the OP said he didn't have that problem.