Problem Solving Help - P4 Partial Loss Of Control

[Syazoo]

I bought a used P4 about 6 weeks ago. It looked to be in excellent condition physically. Only mod is the addition of a maxxuav. About two weeks ago I was flying in an open field outside of any control zones. The P4 suddenly flew straight up without any stick input. over about a two minute period I was able to slowly get it to decend a little at a time and then to land. I do not have the logs to share from that flight. I have since calibrated the IMU, sticks and compass. The DJI Go4 app indicated -.03 VPS every time I turn it on. I have been on several flights since and it has behaved perfectly. Not as nice to control as my mavic pro but overall very well. I need to do some video work today and so I took it out for a test run. I tried the flight in DJI Go 4 and I did not like the way it was hovering. I decided to try Litchi for no other reason than I just bought the program and want to become familiar with it. Litchi makes managing the logs with AirData very convenient. I am sharing a link for what I hope is all the data available for a flight I just took. I took off and let the P4 hover for a minute to warm us as I usually do. Then I wanted to test its behavior by going up and down close to myself. The app reported 100% battery with the enabled speech. Suddenly the P4 started ascending and I was not able to take total control. The app reported 80% battery, down from 100% less than a minute previously. Then it reported 100%. The app went back and forth between those two voltages a couple times. It would drop then accend. I tried going to sport mode so that I would be disabling avoidance and hopefully have more control. That strategy did not do anything to help me gain total control. I finally was able to get the drone to drop to the ground. It was not an eloquent landing but at least it was not a fly away. I logged in to AirData to see if there was anyting obvious reported. My limited understanding only shows me that the log is reporting everything working perfectly. I will learn to forensic this data eventually folks but for now I need some help. I hope this link is adequate. I checked pretty much everything to be allowed as shared. UAV conditions were good to fly..... about 16 mph steady and 27 gusts. I have had this drone up many times in the same conditions and it has been a warrior. Please let me know if there is anyting else I may provide for anyone who is generous enough to give this some time. Thanks.


Drone in question:

 

[Meta4]

The P4 suddenly flew straight up without any stick input. over about a two minute period I was able to slowly get it to decend a little at a time and then to land. I do not have the logs to share from that flight.

I'm confused already.
Is the Airdata record you attached from a different flight or the one you are asking about?

I have since calibrated the IMU, sticks and compass.

None of those would be expected to affect the issue you described.

I tried the flight in DJI Go 4 and I did not like the way it was hovering.

Your drone will hover exactly the same, whether you use DJI go 4 or any other app.

I decided to try Litchi for no other reason than I just bought the program and want to become familiar with it. Litchi makes managing the logs with AirData very convenient

I haven't tried analysing a Litchi log for a long time.
I'd forgotten how terrible they are for incident investigation.
It's really hard to work out what, if anything is showing in the data you've included.

Suddenly the P4 started ascending and I was not able to take total control. The app reported 80% battery, down from 100% less than a minute previously. Then it reported 100%. The app went back and forth between those two voltages a couple times. It would drop then accend.

My first suspicion is the VPS sensors/data but I can't tell if they are OK or not from this data.
The battery doesn't look like it was fully charged at the start of the flight (or it's in poor condition)

Please let me know if there is anyting else I may provide for anyone who is generous enough to give this some time. Thanks.

Try doing a test flight using DJI Go 4 and the data will be a lot easier to make sense of.

To share flight data from DJI Go4...
Go to DJI Flight Log Viewer | Phantom Help
Follow the instructions there to upload your flight record from your phone or tablet.
That will give you a detailed report of the flight.
Come back and post a link to the report it gives you.
Or .. just post the txt file here.

 

[PhantomWetSuits]

Check to make sure you Battery has the same updates as the drone , Since these are intelligent batteries they have
the ability to really effect how the Phantom 4 Flys, I say this because you purchased used so your probably in the dark on the battery updates.

Phantomrain.org
Gear to fly your Phantom in the Rain and float on Water.

 

[Meta4]

Check to make sure you Battery has the same updates as the drone , Since these are intelligent batteries they have
the ability to really effect how the Phantom 4 Flys,

There's no way that battery firmware could make any difference to the way a drone flies.

 

[PhantomWetSuits]

There's no way that battery firmware could make any difference to the way a drone flies.

We have found that the Battery can and does make a Difference in the behavior of the all the DJI drone not just the Phantom.
In the Phantom 4 having the wrong update can cause your Phantom to just take off , which is what the OP describes.
With the Mavic 2 Air not having the Right battery Update can stop you cold at 150 ft distance just for examples

Than you have to consider that he bought it used and was excited to test it out with out checking all the updates.

So Yep , they sure can.

Phantomrain.org
Gear to fly your Phantom in the Rain and Float on Water.

 

[Meta4]

We have found that the Battery can and does make a Difference in the behavior of the all the DJI drone not just the Phantom.
In the Phantom 4 having the wrong update can cause your Phantom to just take off , which is what the OP describes.
With the Mavic 2 Air not having the Right battery Update can stop you cold at 150 ft distance just for examples

Again ... there's nothing at all in the battery firmware updates that could possibly make any difference at all to the way the drone flies.

I don't know what the details are for the incident you mention, but whatever it was, the cause was not related to battery firmware.
It's just impossible for it to have been.

 

[ArnoldLeVine]

Guys, what possible harm can come from having the battery firmware up to date, since whether it could/can affect flight is debatable, why not go with battery firmware that matches your bird and RC, and eliminate that from the possibilities.

 

[PhantomWetSuits]

Guys, what possible harm can come from having the battery firmware up to date, since whether it could/can affect flight is debatable, why not go with battery firmware that matches your bird and RC, and eliminate that from the possibilities.

True, but its not Debatable and here is why anyone that owns a Mavic Air 2 that was stopped at 150 Ft was because they did not update the Battery Firmware,
Now with the Phantom 4 we had a Phantom 4 Pro where every time it started it just took off and it was the battery firmware.

Hopefully the OP will check on that .

Phantomrain.org
Gear to fly your Phantom 4 Pro in the Rain.

 

[Syazoo]

When I bought the Phantom, the drone required a firmware upgrade to .810 according to DJI Go app. I did that. It was not until later that I learned that the battery firmware update was a part of the drone firmware update. Learned from calling DJI. They said I should be using the assistant software so I installed it. I did a refresh two times, once with each of my batteries. I was expecting to see the middle two LED's flashing if the battery was being updated. Neither did. I assumed that they had been updated.

Thanks for looking and trying to help me solve this. Over this past weekend I recalibrated my VPS using assistant. I took my drone out and flew it around in an area where there is nothing around at all. No buildings, nothing just open Saskatchewan fields. The phantom flew perfectly. I brought it to my shop again today where I had the problem that prompted this thread for me. I started the drone up in Litchi. I got a message saying strong interference. I decided not to take off. Besides the problem that prompted this thread, I had the same control issue one other time in Calgary. In Calgary, I got an error saying strong interference. I am sharing that flight as well. I lost control. The Phantom continued to ascend after I let off of the stick. I tried to change modes during the flight. Looking at the flight playback I see that it is not showing the down stick I was doing while the Phantom continued to ascend. I was able to get it down a little at a time. My stick input looks so erratic as I watch the flight. I took off fast, up and away and the drone instantly felt out of my control. Admittedly, my eyes were on the drone for the most part and I did not see all of the errors. Something I should add is that there is a lot of steel buildings around my shop. The yard I was at in Calgary is open but I know that the driveway I took off from is loaded with rebar. This is a dji Go 4 log. I bought Litchi because of all the positive ranting about it online. Frankly, I prefer the dji Go 4. I only do not like how much storage space it consumes on my 16 GB Ipad Mini.

My Mavic Pro has not showed any sign of having problems at either spot the Phantom did. Again, I really appreciate the constructive help. I love this Phantom when it is right. I have such beautiful footage and I love the way it responds to control input. The Mavic Pro is more fun, but this one is a great tool.

 

[Syazoo]

I just thought for whomever is looking at this drone data, here is the flight I mention above that I took over the weekend in an open field which went perfectly. So same drone, different location after VPS calibration.

 

[Meta4]

I lost control. The Phantom continued to ascend after I let off of the stick. I tried to change modes during the flight. Looking at the flight playback I see that it is not showing the down stick I was doing while the Phantom continued to ascend. I was able to get it down a little at a time. My stick input looks so erratic as I watch the flight. I took off fast, up and away and the drone instantly felt out of my control.

This flight was affected by a yaw error.
You were lucky it wasn't a big one, most involve high speeds and end badly

Admittedly, my eyes were on the drone for the most part and I did not see all of the errors.

Look at this display of the flight data and the number of error warnings are there to see now.

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

The yard I was at in Calgary is open but I know that the driveway I took off from is loaded with rebar.

Launching from reinforced concrete is a big risk and many drones have been lost to yaw errors caused that way.
The compass always knows where north is, but the gyro sensor doesn't when it's powered up and takes its heading info from the compass when it starts up.
But if the compass is deflected by the unseen steel that's right below the compass sensor, the gyro sensor gets bad initialisation data and an incorrect idea of where north is.
On launching when the compass is away from the steel, it reads correctly, but the IMU doesn't.
The data conflict between the two sensors causes the drone to "chase its tail" just trying to stay still.
If the difference between the actual heading data and the falsely initialised IMU is small, the drone can often be recovered.
If the difference is a bigger angle, the result is not pretty.

 

[Syazoo]

Thanks for taking a look and the responses. I see the errors. I was not aware of the amount of rebar in that pad until later. I learned something. I suspect that the bad flight I had at my shop was caused by the same issue of interference by a tremendous amount of metal. My company is a street sweeper manufacturing company. You may well imagine that we have a tremendous amount of steel within our steel building. I have had no such problem with my mavic pro or mavic air. Are these drones less sensitive? I have a couple thousand feet logged flying my air in my building. I shoot marketing footage within the shop. Covid has made video an integral part of my marketing strategy.

So to what I hope is the end of my queery about the my loss of control, would it be fair to say that the loss of control is operator error and not a drone hardware malfunction? I have perfect flights in area's without metal. In Calgary in the same yard I had previously launched from the grass and the flight was without event. The flight in Calgary that went wrong did not give me any indication of interference when I fired up the P4. I do get a strong interference message at my shop when I fire up the P4. I just have to heed the warnings.

 

[Meta4]

would it be fair to say that the loss of control is operator error and not a drone hardware malfunction?

It's the operator's choice of a bad startup location

The flight in Calgary that went wrong did not give me any indication of interference when I fired up the P4. I do get a strong interference message at my shop when I fire up the P4. I just have to heed the warnings.

You won't always see any warning in an spot where you could have a bad yaw error.
It doesn't take much steel - a small amount, close to the compass can do it.
The compass sensor will give a warning if the magnetic field is stronger than the earth's normal magnetic field, but one that's a similar strength won't set off any warnings.
But it can cause real problems if it gives the compass a false reading at startup.

You can check that this won't be a problem by comparing the direction the drone is facing with the direction the drone icon on the map is facing.
If the drone is facing.
If they are both facing the same direction ( east and east, or S & S etc) you are good to go.
But if they face different directions, that's an indication that the compass is being deflected by an outside influence.

 

[Syazoo]

Ok, I understand. Great tip regarding direction checking. So in my case it is necessary to shoot in and around my shop. Would I have a solution possibility by launching away from my shop a few hundred yards where I know there is very little metal and flying to my shoot location? Or does the metal around my building still a potential problem after I am in the air? If the solution is reasonable, can I safely choose the home location to be at my shop for landing purposes and termination of the flight session?

 

[Syazoo]

I got to go out a few days ago. I had a successful flight and then after a battery change the gremlin returned. I was at what I beleived to be a spot that should be free of interference. In discussions to this point we have thought that I was launching at spots where interference could be the reason for the control loss.

Flight 1:

DJI Flight Log Viewer - PhantomHelp.com
or airdata log:
Flight that went wrong. I started the Phantom up and let it run and warm up for a little while. Then I launched and moved to the left to be away from myself. It immediately accelerated away and was flying away. I was able to stop the away movement. I tried to land and it went toward the ground very fast so I went up and down trying to regain control. I switched to sport to disable VPS. Then I went in to the menu and disabled within go4 so that I could go back to P mode and try to move slower. I managed to get it on the ground without any hard landing or any incident at all. I have quite a few logs up here now so hopefully we can see something more obvious. All That I can see is that there seems to be an awful lot of signal loss.

DJI Flight Log Viewer - PhantomHelp.com
or airdata:

 

[StPeteFlorida]

Love Airdata

 

[assemblysolutions@outlook]

I bought a used P4 about 6 weeks ago. It looked to be in excellent condition physically. Only mod is the addition of a maxxuav. About two weeks ago I was flying in an open field outside of any control zones. The P4 suddenly flew straight up without any stick input. over about a two minute period I was able to slowly get it to decend a little at a time and then to land. I do not have the logs to share from that flight. I have since calibrated the IMU, sticks and compass. The DJI Go4 app indicated -.03 VPS every time I turn it on. I have been on several flights since and it has behaved perfectly. Not as nice to control as my mavic pro but overall very well. I need to do some video work today and so I took it out for a test run. I tried the flight in DJI Go 4 and I did not like the way it was hovering. I decided to try Litchi for no other reason than I just bought the program and want to become familiar with it. Litchi makes managing the logs with AirData very convenient. I am sharing a link for what I hope is all the data available for a flight I just took. I took off and let the P4 hover for a minute to warm us as I usually do. Then I wanted to test its behavior by going up and down close to myself. The app reported 100% battery with the enabled speech. Suddenly the P4 started ascending and I was not able to take total control. The app reported 80% battery, down from 100% less than a minute previously. Then it reported 100%. The app went back and forth between those two voltages a couple times. It would drop then accend. I tried going to sport mode so that I would be disabling avoidance and hopefully have more control. That strategy did not do anything to help me gain total control. I finally was able to get the drone to drop to the ground. It was not an eloquent landing but at least it was not a fly away. I logged in to AirData to see if there was anyting obvious reported. My limited understanding only shows me that the log is reporting everything working perfectly. I will learn to forensic this data eventually folks but for now I need some help. I hope this link is adequate. I checked pretty much everything to be allowed as shared. UAV conditions were good to fly..... about 16 mph steady and 27 gusts. I have had this drone up many times in the same conditions and it has been a warrior. Please let me know if there is anyting else I may provide for anyone who is generous enough to give this some time. Thanks.


Drone in question:

View attachment 121719



View attachment 121720

never read much past went straight up. Chech return home Alt