Old drone, new pilot, new crash, new crash course in learning the ropes

waterwin · Friday at 4:32 PM

Hello all,
Building experience with a P4P Obsidian, waiting for my first spare parts after an impressive and heartbreaking crash. So now I have all the time to investigate the logs, pick your brains and look at YT videos showing how to replace the landing skids. Thank you for having me.
Erwin

GFields · Friday at 7:34 PM

Greetings from Birmingham Alabama USA, welcome to the forum! We look forward to hearing from you!

Starz · Saturday at 9:33 AM

waterwin said:
heartbreaking crash.

Share some details?

waterwin · Saturday at 10:07 AM

Details of my crash.
My P4P Obsidian is not a new drone that I acquired from a family member (acquired by him second hand and never flown by him). I had flown the last couple of days after passing the obligatory French UAS exam and other administrative procedures. So at most a few hours experience in total. It needed an electronic ID and I acquired the Airbot Data. As I had 2 3-d printed bars to connect small flashlights to the landing gear (came with the drone) I used these to put at the back side of the landing skids next to each other as a small platform for the Airbot. This appeared to be a good place giving a good sight for the gps-receiver in the Airbot and not blocking the gps in the AC itself. Almost 3 minutes in my flight a warning appeared Downward sensor error. Aircraft unstable at low altitude. Fly with caution. which I might have missed on the screen of the GL300E as I tend to fly looking at the drone and not on the screen of the controller, but the message is in the Flight log. At the end of the flight I used most of the battery (still has over 14% in it) and pushed the RTH button. Bird reached the 50 meter home flight altitude, cruised to the H location and started to descent. I thought it went really fast and maybe 5 meters above ground moved the left stick upwards which should have stopped the descent (I think). Well, it did not and crashed very hard on the concrete. Both landings skids are totally broken off. No damage to the camera and gimbal, no other damage to the drone. After the 'landing' the battery was unlocked and a bit outside the battery compartment. I have studied the controller flight data. At the moment I can not access the AC flight data as I took of the camera housing which has the usb port in preparation for the ordered new landing gear. When replaying the controller flight data file I do see the above mentioned message and some more like reaching the maximum altitude, starting the home flight, the battery warnings etc. But the file ends at 33 meters of altitude and a downward speed of 3 m/s, so about 10 seconds before hitting the ground. If I understand correctly, the controller flight data is registered every 10 seconds and the AC flight data file should have a lot more detail and a lot more data points. This could explain why I miss the last 10 seconds of this flight.
Looking back at the current data, I think the 2 3D-printed bars+Airbot are in the way of the downward sensors that sit at the back of the drone. So I am also trying to find a better place for this electronic ID box.
I still do not understand why the drone did not react to my stick movement by stopping the descent. During replay I verified that the stick movement is indeed left stick up.
I have flown (many hours) a small RC heli mostly inside and on a flight simulator before starting with this drone.

BudWalker · Saturday at 10:25 AM

waterwin said:
Details of my crash.
My P4P Obsidian is not a new drone that I acquired from a family member (acquired by him second hand and never flown by him). I had flown the last couple of days after passing the obligatory French UAS exam and other administrative procedures. So at most a few hours experience in total. It needed an electronic ID and I acquired the Airbot Data. As I had 2 3-d printed bars to connect small flashlights to the landing gear (came with the drone) I used these to put at the back side of the landing skids next to each other as a small platform for the Airbot. This appeared to be a good place giving a good sight for the gps-receiver in the Airbot and not blocking the gps in the AC itself. Almost 3 minutes in my flight a warning appeared Downward sensor error. Aircraft unstable at low altitude. Fly with caution. which I might have missed on the screen of the GL300E as I tend to fly looking at the drone and not on the screen of the controller, but the message is in the Flight log. At the end of the flight I used most of the battery (still has over 14% in it) and pushed the RTH button. Bird reached the 50 meter home flight altitude, cruised to the H location and started to descent. I thought it went really fast and maybe 5 meters above ground moved the left stick upwards which should have stopped the descent (I think). Well, it did not and crashed very hard on the concrete. Both landings skids are totally broken off. No damage to the camera and gimbal, no other damage to the drone. After the 'landing' the battery was unlocked and a bit outside the battery compartment. I have studied the controller flight data. At the moment I can not access the AC flight data as I took of the camera housing which has the usb port in preparation for the ordered new landing gear. When replaying the controller flight data file I do see the above mentioned message and some more like reaching the maximum altitude, starting the home flight, the battery warnings etc. But the file ends at 33 meters of altitude and a downward speed of 3 m/s, so about 10 seconds before hitting the ground. If I understand correctly, the controller flight data is registered every 10 seconds and the AC flight data file should have a lot more detail and a lot more data points. This could explain why I miss the last 10 seconds of this flight.
Looking back at the current data, I think the 2 3D-printed bars+Airbot are in the way of the downward sensors that sit at the back of the drone. So I am also trying to find a better place for this electronic ID box.
I still do not understand why the drone did not react to my stick movement by stopping the descent. During replay I verified that the stick movement is indeed left stick up.
I have flown (many hours) a small RC heli mostly inside and on a flight simulator before starting with this drone.

Can you provide the flight logs that you have?

waterwin · Saturday at 11:23 AM

Attached ..

BudWalker · Saturday at 11:48 AM

waterwin said:
Attached ..

Nothing unusual about the .txt log except that it stops abruptly. This is usually caused by a battery that comes loose. The .DAT from the P4 will have info that will allow us to determine if this what happened. If the P4 is like the P3 then the desired USB port will be in the nose, not in the camera body. The P4 will need to be powered up. Take a look at these instructions.

How to retrieve a V3 .DAT File from the AC

waterwin · Saturday at 12:17 PM

BudWalker said:
Nothing unusual about the .txt log except that it stops abruptly. This is usually caused by a battery that comes loose. The .DAT from the P4 will have info that will allow us to determine if this what happened. If the P4 is like the P3 then the desired USB port will be in the nose, not in the camera body. The P4 will need to be powered up. Take a look at these instructions.

How to retrieve a V3 .DAT File from the AC

The nose of the P4P is clean .. My micro usb slot is in the camera / gimbal / plate assembly.

BudWalker · Saturday at 6:09 PM

waterwin said:
The nose of the P4P is clean .. My micro usb slot is in the camera / gimbal / plate assembly.

View attachment 126507

Bummer. Was the battery on when you retrieved the P4?

waterwin · 2025-04-20T04:58:04-0400

Quoting from my description above :" ... After the 'landing' the battery was unlocked and a bit outside the battery compartment. ... "
I assembled the drone a little to gather the data files from the AC. I have files created by Assistant 2 Log Export but also the .DAT files directly from the AC memory using the Flight Data option from Assistant 2. The internal Assistant 2 Open Data Viewer function can open files but the screen remains empty and exported GPS.txt file is empty as well. All files are too big to attach here.

waterwin · 2025-04-20T05:23:42-0400

So I ran DatCon to create the csv and log.txt files from first part FLY107.DAT and second part FLY108.DAT. I can see the downwards sensor issues

168.849 : 11686 [L-FMU/FSM]not near ground
169.084 : 11698 [L-FMU/FSM]near ground
169.180 : 11703 [L-FDI]ULTRASONIC: fault on, stuck

but also :

176.760 : 12081 [L-API]motor state on, cann't wirte param to flash ||
repeating many times

file ends with

960.340 : 51260 [L-API]motor state on, cann't wirte param to flash ||
984.065 : 52447 [L-FDI][CTRL]: fault on , height_ctrl_fail
984.085 : 52448 [L-RC]craft ctrl failed!!!
984.185 : 52453 [L-FDI][CTRL]: fault off, height_ctrl_fail
986.105 : 52549 [L-RC]craft ctrl failed!!!
988.125 : 52650 [L-RC]craft ctrl failed!!!
990.145 : 52751 [L-RC]craft ctrl failed!!!
990.420 : 52764 [L-SYS]set user priv 0xff
990.420 : 52764 [L-API]motor state on, cann't wirte param to flash ||
992.164 : 52852 [L-RC]craft ctrl failed!!!
1020.520 : 54269 [L-SYS]set user priv 0xff
1020.520 : 54269 [L-API]motor state on, cann't wirte param to flash ||
1050.620 : 55774 [L-SYS]set user priv 0xff
1050.620 : 55774 [L-API]motor state on, cann't wirte param to flash ||
1080.720 : 57279 [L-SYS]set user priv 0xff
1080.720 : 57279 [L-API]motor state on, cann't wirte param to flash ||
1110.800 : 58783 [L-SYS]set user priv 0xff
1110.800 : 58783 [L-API]motor state on, cann't wirte param to flash ||
1140.900 : 60288 [L-SYS]set user priv 0xff
1140.900 : 60288 [L-API]motor state on, cann't wirte param to flash ||
1163.645 : 61426 [L-FLYMODE][Ctrl<8>] REQ_BATTERY NAVI_GO_HOME ctrl_navi_go_home
1171.000 : 61793 [L-SYS]set user priv 0xff
1171.000 : 61793 [L-API]motor state on, cann't wirte param to flash ||
1190.245 : 62756 [L-FLYMODE][Ctrl<8>] REQ_BATTERY AUTO_LANDING ctrl_auto_landing
1190.265 : 62757 [L-FLYMODE][GO HOME] reset go home status
1191.045 : 62796 [L-RC]Start landing. rc vib!
1197.901 : 63139 [L-FDI]COMPASS(0): disconnect
1197.902 : 63139 [L-FMU/DM]Busy Device Changed. Type:compass0, <ID:6 idx:0-->ID:7 idx:1>, change times:
1197.921 : 63140 [L-FDI]COMPASS(1): disconnect

My zip file that has the not allowed extension like csv and kml, kml is quite empty though, is too large to attach.

BudWalker · 2025-04-20T10:20:26-0400

Can you provide FLY107.DAT and FLY108.DAT using Dropbox or GoogleDrive? Often it's unclear exactly what significance eventLog entries have.

waterwin · 2025-04-20T10:54:03-0400

Thank you for your time and effort !

Dropbox

www.dropbox.com

Dropbox

www.dropbox.com

BudWalker · 2025-04-20T12:36:33-0400

This is a bit different than several other P4 incidents where there was an abrupt loss of power. Most look like this

Beginning to hate phantoms ...

In your incident the voltages fluctuated some before collapsing

I don’t have a compelling explanation why one of the voltages increased to the max battery voltage.

I think the most likely explanation is that the battery dislodged

waterwin · 2025-04-21T09:38:52-0400

I have looked at the end of the FLY7 event log again.

1163.645 : 61426 [L-FLYMODE][Ctrl<8>] REQ_BATTERY NAVI_GO_HOME ctrl_navi_go_home
1171.000 : 61793 [L-SYS]set user priv 0xff
1171.000 : 61793 [L-API]motor state on, cann't wirte param to flash ||
1190.245 : 62756 [L-FLYMODE][Ctrl<8>] REQ_BATTERY AUTO_LANDING ctrl_auto_landing
1190.265 : 62757 [L-FLYMODE][GO HOME] reset go home status
1191.045 : 62796 [L-RC]Start landing. rc vib!
1197.901 : 63139 [L-FDI]COMPASS(0): disconnect
1197.902 : 63139 [L-FMU/DM]Busy Device Changed. Type:compass0, <ID:6 idx:0-->ID:7 idx:1>, change times:
1197.921 : 63140 [L-FDI]COMPASS(1): disconnect

It starts a home flight when the battery is low, auto-landing, start landing, vibrate controller, ...

6 seconds later it still registers that the compass becomes disconnected.

The compass disconnects because my landing gear breaks upon ground impact.

That is the end of the crash and the battery can not be failing before that moment or being dislodged before, as in that case it would not be registering the failure of the compass in the flight log.

The battery max voltage can be a result of me trying to stop the dropping drone with a left stick up movement that I can also see when looking at stick movements on screen. But I was too late to stop the action and the drone from touching ground.
The battery level might have been too low already to do such a braking manoeuvre and could not deliver/sustain the needed current.

Motors speed graph from just before Return Home to Compass Disconnected

BudWalker · 2025-04-21T13:42:33-0400

waterwin said:
I have looked at the end of the FLY7 event log again.

1163.645 : 61426 [L-FLYMODE][Ctrl<8>] REQ_BATTERY NAVI_GO_HOME ctrl_navi_go_home
1171.000 : 61793 [L-SYS]set user priv 0xff
1171.000 : 61793 [L-API]motor state on, cann't wirte param to flash ||
1190.245 : 62756 [L-FLYMODE][Ctrl<8>] REQ_BATTERY AUTO_LANDING ctrl_auto_landing
1190.265 : 62757 [L-FLYMODE][GO HOME] reset go home status
1191.045 : 62796 [L-RC]Start landing. rc vib!
1197.901 : 63139 [L-FDI]COMPASS(0): disconnect
1197.902 : 63139 [L-FMU/DM]Busy Device Changed. Type:compass0, <ID:6 idx:0-->ID:7 idx:1>, change times:
1197.921 : 63140 [L-FDI]COMPASS(1): disconnect

It starts a home flight when the battery is low, auto-landing, start landing, vibrate controller, ...

6 seconds later it still registers that the compass becomes disconnected.

The compass disconnects because my landing gear breaks upon ground impact.

That is the end of the crash and the battery can not be failing before that moment or being dislodged before, as in that case it would not be registering the failure of the compass in the flight log.

The battery max voltage can be a result of me trying to stop the dropping drone with a left stick up movement that I can also see when looking at stick movements on screen. But I was too late to stop the action and the drone from touching ground.
The battery level might have been too low already to do such a braking manoeuvre and could not deliver/sustain the needed current.

Motors speed graph from just before Return Home to Compass Disconnected

View attachment 126512

You're right. The abrupt power loss was caused by the impact with the ground. The impact is seen in the accelerometer data

Looks like the impact started at about 1197.865 secs, just 0.07 secs before the .DAT ends.

At 1197.242 full throttle was applied. The P4P had been increasing motor speeds by about 1500 RPM/sec. As a result of full throttle, the P4P then increased that rate to about 4500 RPM/sec.