Circular Hover Pattern in GPS Mode

Joined
Apr 30, 2013
Messages
14
Reaction score
0
Hi everyone,

First post here - been reading most of the posts. Great to see such an active, friendly community focused on the Phantom.

First some backstory --

A few weeks ago my phantom took a pretty nasty fall. One of the propellers separated from its motor mid-flight about 50 feet up in the air. The resulting crash damaged one of the motors. I replaced the motor, calibrated the IMU, and took her out for a test.

The phantom is flying well in ATTI mode, but in GPS mode it now tends to make 5-10 foot circles instead of hovering in one place. It doesn't matter what direction the drone is facing; it consistently makes clockwise circles with a radius of 5 to 10 feet. In attitude mode, if there is no wind, the phantom hovers in place.

Any thoughts on what could be causing the circular flying pattern in GPS mode?

Thanks in advance for all insight!
 
HurricaneProductions said:
Hi everyone,

First post here - been reading most of the posts. Great to see such an active, friendly community focused on the Phantom.

First some backstory --

A few weeks ago my phantom took a pretty nasty fall. One of the propellers separated from its motor mid-flight about 50 feet up in the air. The resulting crash damaged one of the motors. I replaced the motor, calibrated the IMU, and took her out for a test.

The phantom is flying well in ATTI mode, but in GPS mode it now tends to make 5-10 foot circles instead of hovering in one place. It doesn't matter what direction the drone is facing; it consistently makes clockwise circles with a radius of 5 to 10 feet. In attitude mode, if there is no wind, the phantom hovers in place.

Any thoughts on what could be causing the circular flying pattern in GPS mode?

Thanks in advance for all insight!

you may also want to double check your compass, as it may need to be re-calibrated also. you may also want to double check your craft for bent arms. this can be easily checked by the distance of the propellers' clearance at the tip from the shell. they should all be the same, if not, you most likely have a bent arm (or something similar to cause the propellers to not be "level"). other than that, you can double check the remaining 3 motors that have not been replaced to see if they are operating normally (no unusual noise or indication that it's hotter).
 
Are you getting a single blinking green led in gps mode ? or is it flashing yellow ?

If the Phantom is flying well in Attitude mode, but not gps mode, then you have two possible problems, compass, or gps unit.

You didn't mention if you had done an advanced calibration, so you might try that. Also look at the videos at the DJI website
regarding compass errors too large for the software to repair on its own. He uses a magnet to bring it back to a manageable
level.
The video regarding "toilet bowl effect" may be one to watch too.

Good Luck !
 
I've experienced that as well, often referred to as the "Toilet Bowl Effect"
do make sure nothing else has been damaged..and run the Advanced Calibration a few times and hopefully it'll go away.
 
Thanks for all the replies!

I have run advanced calibration a few times. Flight has improved but it is still giving me the "toilet bowl" problem.

The numbers I'm getting for the compass in NAZA are all within normal range.

I can say for sure that one of my Phantom's arms is definitely bent. The clearance between one of the propellers and the shell is far less than the rest, almost touching. I can visibly see where the arm is bent. I tried heating up the plastic and bending it back but that didn't work too well.

This arm must have been the first point of contact when it hit the ground.

I thought the bent arm was the problem when I first noticed it, but shouldn't I also be seeing this "toilet bowl" effect in attitude mode if that is what is causing it? Why does it only seem to happen in GPS mode?

The unit can be in ATTI, hovering well and as soon as I switch on GPS it starts to glide to the right and begin the clockwise motion.

--

I plan on replacing the shell once it comes in, I know the parts are on back-order everywhere.

Also, where in NAZA-M do you calibrate the compass? I thought that was taken care of during the IMU calibration?

Again, thank you for the replies. I appreciate the advice.
 
HurricaneProductions said:
I thought the bent arm was the problem when I first noticed it, but shouldn't I also be seeing this "toilet bowl" effect in attitude mode if that is what is causing it? Why does it only seem to happen in GPS mode?

Also, where in NAZA-M do you calibrate the compass? I thought that was taken care of during the IMU calibration?

Well of course it will never toilet bowl in ATTI mode, as the Phantom will do wherever the wind takes it.
the toilet bowl effect occurs when things are kinda skewed between the GPS and the NAZA and result is the Phantom (or other Quads) will never be satisfied in the same position and will continuously try to correct itself (kinda like a cat chasing it's tail)

the Advanced Calibration kinda covers it all (I think).
But it couldn't hurt to go outside and re-calibrate your compass the way you did it when you first bought your Phantom - toggle GPS/ATT and spin around a few times.
 
Are there any adjustments I can make in NAZA-M to compensate for the bent arm?

No matter how many times I run advanced calibration or re-calibrate the compass the toilet bowl effect remains. The diameter of the circle seems to fluctuate. Sometimes it's only 2-3 feet wide, sometimes the circle is 10-15 feet in diamater.
 
HurricaneProductions said:
Are there any adjustments I can make in NAZA-M to compensate for the bent arm?

No matter how many times I run advanced calibration or re-calibrate the compass the toilet bowl effect remains. The diameter of the circle seems to fluctuate. Sometimes it's only 2-3 feet wide, sometimes the circle is 10-15 feet in diamater.

nothing in the naza-m to my knowledge, although you could theoretically un-mount the motor and then add say something rubber underneath the motor on the side(s) closest to the body. that way, you can adjust the tilt of the motor by increasing the torque of the screws. this is good for minor adjustments and not meant to compensate for major differences in propeller levels. (you really don't want the screws to come undone in flight)
 
auck said:
nothing in the naza-m to my knowledge, although you could theoretically un-mount the motor and then add say something rubber underneath the motor on the side(s) closest to the body. that way, you can adjust the tilt of the motor by increasing the torque of the screws. this is good for minor adjustments and not meant to compensate for major differences in propeller levels. (you really don't want the screws to come undone in flight)

Nice tip! I think I'll give it a try.
 
I don't think the motor is causing the toilet bowl effect, if it flies okay other than when the gps is trying to work.

You must have taken the top shell off to replace the motor, is it possible that you put the top back on facing a different
direction ?

It would be easy to do, and would explain why it only flies different in gps mode. It's been done before.
 
Audaciter said:
You must have taken the top shell off to replace the motor, is it possible that you put the top back on facing a different
direction ?

Well, I used the red stickers as markers and they are lining up so I'm pretty sure the shell is on correctly.

I agree, if the circular flying pattern was caused by the tilt on the one motor it should be happening in all modes -- not just GPS. Still going to replace the shell when it comes in though. I'll report back with results.
 
I have bent the tip of an arm just recently also. Was able to almost bend it back level. I think that is the key. With one motor angled, any correction the GPS wants to make it's over correcting because of the arm angle, then the other motors push it around for another try to find stable hover point. Just my opinion but it seems logical.
 
Here in south america, we have seen some problems with the phantone du to the GPS reciver align. We use to correct it manually abou 20 to 45 degrees to the front-left side of the vessel. Its because the magnetic field here is different then there in europe and north america.

The toilet effect is very frecuent and all the phantom users have done this manual gps antena adjustment.

What i think.

in one of your crashes, maybe the GPS have rotated a few degres. Causing this 2-3 feet maybe 10feet circle effect. As the GPS signal is different due to its possition.

For that i recomend to you dissassembly the GPS protection and check if it is okay. And if it not, here is a video on how to fiz it.

In this case, this guy is on a very low region of brazil that why he adjusts 45º, here in Sao Paulo i had to adjusted in 31, and other places is 20 or less. In your case i think it would be a very feew degrees or maybe re-align it to 0 degress

HERE IS THE VIDEO https://www.youtube.com/watch?v=hrDdJKCzCBY

Is in portugues, if you have any question i can translate it for you.

Hope you the best man! GOOD LUCK
 

Members online

No members online now.

Forum statistics

Threads
143,066
Messages
1,467,359
Members
104,936
Latest member
hirehackers