Typhoon H Is Drifting and Doing the Toilet Bowl

QuadBart · Jun 11, 2016

**** if my Typhoon H isn't experiencing a problem; Drift and Toilet Bowling

The below video shows it doing the "Toilet Bowl" (I apologize for my scatterbrain commentary. I'm a little upset its defective. Price you pay for being an early adopter I guess)

I have the full unedited video that shows the entire flight, initially everything is fine. It holds position, it flys straight and works, but after several minutes it starts to drift, then when put in a hover it will intermittently start doing the "Toilet Bowl". After breaking the Toilet Bowl affect it will no longer fly straight. It will want to veer to the right. To fly in a straight line, I have to feed it a slight left feed to the right stick. And if I let go of the sticks rather than slowing and hovering it will drift to the right and backwards from 4-6' (shown in the unedited video).

I saw a video on youtube; where the guy has the same problem as mine does but he wasn't so lucky...

So details;

I've done at least a dozen compass calibrations (all blinking success) in 3 different locations with the same results. I've flown at this park dozens of times with my Q500 4k and P3 with zero issues.
I've done at least 3 IMU/accelerometer calibrations in different locations
I have the latest Firmware installed
I used UAV Forecast to ensure weather conditions were good for flying with a low Kp index
I have good GPS lock and enough Satellites to do follow me
I flew my Q500 4k in that same pl

So, I'm hoping my LHS will exchange my "H" rather than me having to send it back to Yuneec.... ****!

BobW55 · Jun 11, 2016

Only suggestion I have would be to go somewhere else. At least 5 miles away, and try it again.
If she flies stable in the second location, I would think possible RF interference. If she continues the toilet bowl
(good description) then I would think something wrong with the H.
In hover, it relies on the compass, barometer, and accelerometers to maintain a level/stable position.
I have always noticed an ever so slight drift/movements in hover on my H, but it pretty well stays where I leave it.
Hope you get it sorted out.

DroneClone · Jun 11, 2016

QuadBart said:
**** if my Typhoon H isn't experiencing a problem; Drift and Toilet Bowling

The below video shows it doing the "Toilet Bowl" (I apologize for my scatterbrain commentary. I'm a little upset its defective. Price you pay for being an early adopter I guess)

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

I have the full unedited video that shows the entire flight, initially everything is fine. It holds position, it flys straight and works, but after several minutes it starts to drift, then when put in a hover it will intermittently start doing the "Toilet Bowl". After breaking the Toilet Bowl affect it will no longer fly straight. It will want to veer to the right. To fly in a straight line, I have to feed it a slight left feed to the right stick. And if I let go of the sticks rather than slowing and hovering it will drift to the right and backwards from 4-6' (shown in the unedited video).

I saw a video on youtube;

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

where the guy has the same problem as mine does but he wasn't so lucky...

So details;

I've done at least a dozen compass calibrations (all blinking success) in 3 different locations with the same results. I've flown at this park dozens of times with my Q500 4k and P3 with zero issues.

I've done at least 3 IMU/accelerometer calibrations in different locations

I have the latest Firmware installed

I used UAV Forecast to ensure weather conditions were good for flying with a low Kp index

I have good GPS lock and enough Satellites to do follow me

I flew my Q500 4k in that same pl

So, I'm hoping my LHS will exchange my "H" rather than me having to send it back to Yuneec.... ****!

What about a stick calibration, sticks could be off centered, so when releasing sticks, they could be giving slight input?

PatR · Jun 11, 2016

Perhaps once more performing a compass and accelerometer cal, this time starting with the front of the H facing north would help. Toilet bowling is always caused by one of two things. Incorrect gain settings and lack of/poor compass calibration. A level "toilet bowl" as yours was doing is compass induced. Where do you store your H when not in use? Hopefully not near some magnetic source. Are you one that has a powerful stereo system with large speakers and strong magnets in your vehicle?

DroneClone · Jun 11, 2016

PatR said:
Perhaps once more performing a compass and accelerometer cal, this time starting with the front of the H facing north would help. Toilet bowling is always caused by one of two things. Incorrect gain settings and lack of/poor compass calibration. A level "toilet bowl" as yours was doing is compass induced. Where do you store your H when not in use? Hopefully not near some magnetic source. Are you one that has a powerful stereo system with large speakers and strong magnets in your vehicle?

Exactly, Have you adjusted any of the gains/expo settings, either accidentally, or intentionally? Good note there Pat !

CAPTAINDRONE · Jun 11, 2016

QuadBart said:
**** if my Typhoon H isn't experiencing a problem; Drift and Toilet Bowling

The below video shows it doing the "Toilet Bowl" (I apologize for my scatterbrain commentary. I'm a little upset its defective. Price you pay for being an early adopter I guess)

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

I have the full unedited video that shows the entire flight, initially everything is fine. It holds position, it flys straight and works, but after several minutes it starts to drift, then when put in a hover it will intermittently start doing the "Toilet Bowl". After breaking the Toilet Bowl affect it will no longer fly straight. It will want to veer to the right. To fly in a straight line, I have to feed it a slight left feed to the right stick. And if I let go of the sticks rather than slowing and hovering it will drift to the right and backwards from 4-6' (shown in the unedited video).

I saw a video on youtube;

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

where the guy has the same problem as mine does but he wasn't so lucky...

So details;

I've done at least a dozen compass calibrations (all blinking success) in 3 different locations with the same results. I've flown at this park dozens of times with my Q500 4k and P3 with zero issues.

I've done at least 3 IMU/accelerometer calibrations in different locations

I have the latest Firmware installed

I used UAV Forecast to ensure weather conditions were good for flying with a low Kp index

I have good GPS lock and enough Satellites to do follow me

I flew my Q500 4k in that same pl

So, I'm hoping my LHS will exchange my "H" rather than me having to send it back to Yuneec.... ****!

I had the same problem with a brand new Phantom Pro recently (video has been posted). The factory compass & IMU calibration should be perfect. When we do our own calibration and IMU calibration, numbers change & it "normally" should not be a problem. However, sometimes it becomes a problem and the GPS to IMU to Compass gets out of whack & the next thing you know you have the "toilet bowl" effect. I wish these drones all had a FACTORY RESET option for IMU/COMPASS so that one could start all over.

glider · Jun 11, 2016

GPS is used for horizontal positioning estimates. If you are a good pilot, fly with GPS off. If the toilet bowl stops, it's the GPS glitching.

It could be any combination of issues, including vibration.

How much data is available to diagnose problems?

RBC · Jun 11, 2016

glider said:
GPS is used for horizontal positioning estimates. If you are a good pilot, fly with GPS off. If the toilet bowl stops, it's the GPS glitching.

It could be any combination of issues, including vibration.

How much data is available to diagnose problems?

I did not think the H has the option to turn off GPS like the Phantom 4 does in Atti mode. Am I wrong on this?

CAPTAINDRONE · Jun 11, 2016

glider said:
GPS is used for horizontal positioning estimates. If you are a good pilot, fly with GPS off. If the toilet bowl stops, it's the GPS glitching.

It could be any combination of issues, including vibration.

How much data is available to diagnose problems?

You never want to have the toilet bowl effect. Because this means your GPS and compass (true & magnetic north) are not correct. RTH has a limited chance of ever working in this situation.

glider · Jun 11, 2016

Yes it can be disabled, but not sure once in the air if it can be re-enanabled or if RTH is rendered useless.

glider · Jun 11, 2016

CAPTAINDRONE said:
You never want to have the toilet bowl effect. Because this means your GPS and compass (true & magnetic north) are not correct. RTH has a limited chance of ever working in this situation.

Believe me, working with early versions of Arducopter I'm all too familiar with toilet bowling, but it is not limited to just GPS and compass. It would be good if the telemetry reports all sensor data but I have not had the H long enough to get into it that deep

Steve Carr · Jun 11, 2016

Do you have the Obstacle Avoidance turn on? If so, that can cause an issue. Does it occur in both Angel and Smart mode?
A stable hover is achieved mainly by the accelerometer/gyro. GPS doesn't have the accuracy needed for exact station keeping.

QuadBart · Jun 11, 2016

My LHS exchanged out my H for another one! One of the reasons why I support my LHS

To answer some of the questions;

Yes I checked to make sure my "sticks" weren't out of "whack". What also leads me to believe this was not an issue with the "stick rates (gains/expos)" is that the first few minutes, the bird flys and holds exactly the way it should. It degrades over time. If it was the "stick rates", I'd expect the issues from the very beginning...
Happens in both Angle and Smart Mode
Collision Avoidance off
Even though it doesn't say to; I always faced the "H" North during Compass Calibration
During IMU/Accelerometer calibration I even took a "level" to my H to ensure it was indeed "level"
I did multiple dozen-ish compass calibrations all were successful based on the lights and sound
I downloaded/reinstalled/verified the firmware and flew it again to ensure it wasn't a bad download. Same bad results
I've flown the "H" in 3 different locations where I've flown my Q500 4k and Phantom 3 with no issues

Here's to keeping my fingers crossed this new one works correctly!

Steve Carr · Jun 11, 2016

QuadBart said:
My LHS exchanged out my H for another one! One of the reasons why I support my LHS

To answer some of the questions;

Yes I checked to make sure my "sticks" weren't out of "whack". What also leads me to believe this was not an issue with the "stick rates (gains/expos)" is that the first few minutes, the bird flys and holds exactly the way it should. It degrades over time. If it was the "stick rates", I'd expect the issues from the very beginning...

Happens in both Angle and Smart Mode

Collision Avoidance off

Even though it doesn't say to; I always faced the "H" North during Compass Calibration

During IMU/Accelerometer calibration I even took a "level" to my H to ensure it was indeed "level"

I did multiple dozen-ish compass calibrations all were successful based on the lights and sound

I downloaded/reinstalled/verified the firmware and flew it again to ensure it wasn't a bad download. Same bad results

I've flown the "H" in 3 different locations where I've flown my Q500 4k and Phantom 3 with no issues

Here's to keeping my fingers crossed this new one works correctly!

It sounds like you definitely had all the bases covered. I know that some of these issues are caused when the bird is not level or gets moved during bootup. Since I have seen several posts of this issue I really wonder if the IMU has a glitch. Did the LHS have any comments about a possible cause???
In any case, let us know if the new bird flies as expected. It' great to have a store that takes care of these problems.

QuadBart · Jun 12, 2016

Follow up - I got the new H home, updated the firmware, compass calibration, and IMU/Accelerometer calibration. Wind had picked up by the afternoon as a thunderstorm was rolling in, so didn't get a chance to fly it till this morning. Today, no wind, and UAV Forecast showing "Good Flying Conditions".

IT WORKED PERFECT! The new H flew exactly as expected. And now that I have something to compare the Defective H (from this point forward, Defective H referred to "A", and new working H as "B"), I can say there is/was something wrong with the GPS/Compass board on "A".

One thing I noticed when doing a Compass Calibration on the "A" is that almost every time, before I finished the CC (on the last rotation), it would flash the success lights and sound the success tone even before the bird was returned to level. In all the CCs I performed, every single one was a success (based on the tone and lights). The "B" did not flash the success lights or play the tone until the last rotation was completed and returned to level. Not having anything to compare "A" to, I thought this was normal. I see now that was probably an indicator something was wrong.

Also flying "B" around it acts/performs just like any of my other "drones", its stable, goes exactly where I want it, and stops like I would expect (including momentum) I realize now that "A" was never really ever stable, almost as if GPS was always partially turned off and would degrade over time. But it doesn't explain when the sticks were let go why it would drift in semi circle, clockwise till it reached the 4 o'clock position in a 4-6 foot semi-circle (even in calm skies).

Another thing "A" did was "twitch" it would randomly just "twitch" it would momentarily dip in any direction briefly and return to level. It wouldn't change location, it would simply "dip a wing" and return to center... It wasn't constant but a few times a flight. "B" does not do this.

Again, thanks to my LHS for exchanging "A" for "B"!

Brandon B · Jun 12, 2016

...Were you flying near California? and what day?.... there is the the GPS jamming out here 500 miles surrounding China Lake NWS:
http://gizmodo.com/faa-warns-of-gps-outages-this-month-during-mysterious-t-1780866590

QuadBart · Jun 12, 2016

Nope. I saw this article last week. My video was shot on 11 June and I'm not near California and out of range of the the circles shown...

Brandon B said:
...Were you flying near California? and what day?.... there is the the GPS jamming out here 500 miles surrounding China Lake NWS:
http://gizmodo.com/faa-warns-of-gps-outages-this-month-during-mysterious-t-1780866590

CAPTAINDRONE · Jun 12, 2016

QuadBart said:
Follow up - I got the new H home, updated the firmware, compass calibration, and IMU/Accelerometer calibration. Wind had picked up by the afternoon as a thunderstorm was rolling in, so didn't get a chance to fly it till this morning. Today, no wind, and UAV Forecast showing "Good Flying Conditions".

IT WORKED PERFECT! The new H flew exactly as expected. And now that I have something to compare the Defective H (from this point forward, Defective H referred to "A", and new working H as "B"), I can say there is/was something wrong with the GPS/Compass board on "A".

One thing I noticed when doing a Compass Calibration on the "A" is that almost every time, before I finished the CC (on the last rotation), it would flash the success lights and sound the success tone even before the bird was returned to level. In all the CCs I performed, every single one was a success (based on the tone and lights). The "B" did not flash the success lights or play the tone until the last rotation was completed and returned to level. Not having anything to compare "A" to, I thought this was normal. I see now that was probably an indicator something was wrong.

Also flying "B" around it acts/performs just like any of my other "drones", its stable, goes exactly where I want it, and stops like I would expect (including momentum) I realize now that "A" was never really ever stable, almost as if GPS was always partially turned off and would degrade over time. But it doesn't explain when the sticks were let go why it would drift in semi circle, clockwise till it reached the 4 o'clock position in a 4-6 foot semi-circle (even in calm skies).

Another thing "A" did was "twitch" it would randomly just "twitch" it would momentarily dip in any direction briefly and return to level. It wouldn't change location, it would simply "dip a wing" and return to center... It wasn't constant but a few times a flight. "B" does not do this.

Again, thanks to my LHS for exchanging "A" for "B"!

There really needs to be a fix for this that users can do without returning the drone. I had a new Phantom recently that did this and I researched everything to resolve it myself. I even tried the magnet trick (you rotate a magnet around the compass in horizontal and vertical circles). This will demagnetize the compass slightly from being attracted in one direction (kind of a reset). You then run the IMU and it should reset the compass. It works for some, but it did not work for me. I have several Phantoms and compared the IMU compass calibration numbers between them. The one that was doing the same as your Typhoon had all the wrong numbers and the IMU was desperately trying to compensate (hence toilet bowl and or twitching).

The normal rule is that if you calibrate your compass once and everything is fine, DON'T keep doing it as you'll potentially end up with a problem (bad calibration/IMU or the physical compass moves from all the spinning). Only do it when it needs it. The same is true after a Firmware upgrade. Most will advise to recalibrate the compass after a firmware upgrade. But not every firmware upgrade requires a compass recalibration.

In summary... If it happens again with your new drone, there really is not exact way to resolve it, it's all trial and error or send it back.