Slow-Mo tip over after landing

[h-elsner]

Typhoon H has a couple of firmware problems than never were solved and never will because the H is discontinued. With PX4 autopilot FW the H performs very smooth landings.

Microsoft OneDrive

The first problem is landing detection. This depends on many parameters and I guess not all are optimal in FW. Unfortunately we cannot change parameters. Another problem is that the H did not automatically disarm as the H Plus and all other PX4 based drones do. So, all problems (stick calibration, channel settings, stick poti cleanness etc.) have influence on landing detection.

The accelerometer in the IMU drifts during lifetime. This may end up in a defect flight controller. Many discussions here about this case. It's absolutely neccessay the calibrate the accelerometer from time to time to overcome this situation. I do it since I had the problem that the H did refuse to set down. It happened never again.

My first approve would be to do an acc calibration which is easy to do. The next step is clean and calibrate the sticks in ST16. The potis in the sticks are not the best, we all know. Then try again...

br HE

 

[DroneRanger]

Typhoon H has a couple of firmware problems than never were solved and never will because the H is discontinued. With PX4 autopilot FW the H performs very smooth landings.

Microsoft OneDrive

The first problem is landing detection. This depends on many parameters and I guess not all are optimal in FW. Unfortunately we cannot change parameters. Another problem is that the H did not automatically disarm as the H Plus and all other PX4 based drones do. So, all problems (stick calibration, channel settings, stick poti cleanness etc.) have influence on landing detection.

The accelerometer in the IMU drifts during lifetime. This may end up in a defect flight controller. Many discussions here about this case. It's absolutely neccessay the calibrate the accelerometer from time to time to overcome this situation. I do it since I had the problem that the H did refuse to set down. It happened never again.

My first approve would be to do an acc calibration which is easy to do. The next step is clean and calibrate the sticks in ST16. The potis in the sticks are not the best, we all know. Then try again...

br HE

Excellent background info h-elsner. Much appreciated. I anticipate uploading the relevant files later today for group analysis. Thanks again everyone!

 

[WTFDproject]

Please be aware that we need the Telemetry, Remote, and RemoteGPS files for that flight to do a proper analysis. For the issue you are having the Remote.csv is going to be a must along with the Telemetry.csv.

Keeping the folder structure by following the instructions gives us more than that flight, but also allows us to see a trend in control issues from previous flights that can occur with the ST16.

The attached document gives some background into factors affecting control.

There are lots of good comments in this thread. And lots of valid recommendations. But as valid as they are, the recommendations are speculation based on similar experiences and related history. And there are LOTS of possible causes in that history.
When we get the flightlogs, we can go from speculation to observation. You have two paths you can follow. Speculation or Analysis. Either may lead to success. Either may lead to nowhere. It's your choice. But I would choose flightlog review.

 

[DroneRanger]

There are lots of good comments in this thread. And lots of valid recommendations. But as valid as they are, the recommendations are speculation based on similar experiences and related history. And there are LOTS of possible causes in that history.
When we get the flightlogs, we can go from speculation to observation. You have two paths you can follow. Speculation or Analysis. Either may lead to success. Either may lead to nowhere. It's your choice. But I would choose flightlog review.

I believe that I followed the instructions correctly. Judging by its size, I would imagine that there is more than just the final flight recorded in this log file. The date of the crash was 2/18/2023.

Thanks in advance!

 

[WTFDproject]

I believe that I followed the instructions correctly. Judging by its size, I would imagine that there is more than just the final flight recorded in this log file. The date of the crash was 2/18/2023.

Thanks in advance!

You did good. Everything is there.
It is too early to tell much yet, and other eyes will see other things. But I did notice an odd feature.
None of your control stick signals (including the throttle stick) ever went anywhere near zero or much above neutral during the whole flight. That could be a really long coincidence of pot corrosion. Or it could be a common cause, such as an improperly performed channel calibration.
Have you recently done a Controller Calibration or recently entered the Channel Settings menu for any reason?

Note the signals looked normal on the last two actual flights of 02/13/23 (flights 139 and 140).
The first "flight" that day looks weird, but it may not have been an actual flight.

 

[DoomMeister]

@DroneRanger we need you to power just the ST16 and enter the Final Output Display by double tapping the Volts at the upper left of the display. It will look like this:

Place the Rate slider on the right side of the controller to full up Rabbit position and record the values of CH1, 2, 3, and 4 at neutral position of the sticks then at full stick deflections for J1/CH1, J4/CH2, J3/CH3, and J2/CH4. Also record the Pan Knob K1/CH8 and Tilt slider K2/CH7.

Now adjust the Rate slider to full down Turtle position and repeat recording the values of the stick/channel combinations.

When finished please report your results. This will help us with what we are seeing in the telemetry files.

 

[DroneRanger]

@DroneRanger we need you to power just the ST16 and enter the Final Output Display by double tapping the Volts at the upper left of the display. It will look like this:
View attachment 29311
Place the Rate slider on the right side of the controller to full up Rabbit position and record the values of CH1, 2, 3, and 4 at neutral position of the sticks then at full stick deflections for J1/CH1, J4/CH2, J3/CH3, and J2/CH4. Also record the Pan Knob K1/CH7 and Tilt slider K2/CH6.

Now adjust the Rate slider to full down Turtle position and repeat recording the values of the stick/channel combinations.

When finished please report your results. This will help us with what we are seeing in the telemetry files.

Copy that! I can get that info after work today. Thanks very much!

 

[DoomMeister]

Copy that! I can get that info after work today. Thanks very much!

Thank you. I have a good idea of what the issue is and this will help confirm or debunk my ideas.

 

[DroneRanger]

You did good. Everything is there.
It is too early to tell much yet, and other eyes will see other things. But I did notice an odd feature.
None of your control stick signals (including the throttle stick) ever went anywhere near zero or much above neutral during the whole flight. That could be a really long coincidence of pot corrosion. Or it could be a common cause, such as an improperly performed channel calibration.
Have you recently done a Controller Calibration or recently entered the Channel Settings menu for any reason?

Note the signals looked normal on the last two actual flights of 02/13/23 (flights 139 and 140).
The first "flight" that day looks weird, but it may not have been an actual flight.

I did not do a Controller Calibration, but I may have looked at the Channel Settings page, however I did not make any changes to anything there. Also, It looked to me like you attached a graphic of some sort to your post because I saw a thumbnail of it in the forum email notification I received and you seem to make reference to it, but I do not see it attached here. Strange.

 

[Steve Carr]

This is the graphical display you should see to check outputs.
Don't do a controller calibration until you verify the sticks are working properly.

 

[DroneRanger]

This is the graphical display you should see to check outputs.
Don't do a controller calibration until you verify the sticks are working properly.

Copy that. I'll check those readings and report the results a bit later today.

 

[WTFDproject]

...., It looked to me like you attached a graphic of some sort to your post because I saw a thumbnail of it in the forum email notification I received and you seem to make reference to it, but I do not see it attached here. Strange.

It was a limited snapshot of the controller status at the fraction of a second the tilting event began. It was part of a different discussion. Both the other discussion and the graphic were deleted due to low value added. I have no idea why the remnant of the image still appears.

 

[DroneRanger]

It was a limited snapshot of the controller status at the fraction of a second the tilting event began. It was part of a different discussion. Both the other discussion and the graphic were deleted due to low value added. I have no idea why the remnant of the image still appears.

Ah, thank you. That explains it!

 

[DroneRanger]

@DroneRanger we need you to power just the ST16 and enter the Final Output Display by double tapping the Volts at the upper left of the display. It will look like this:
View attachment 29311
Place the Rate slider on the right side of the controller to full up Rabbit position and record the values of CH1, 2, 3, and 4 at neutral position of the sticks then at full stick deflections for J1/CH1, J4/CH2, J3/CH3, and J2/CH4. Also record the Pan Knob K1/CH7 and Tilt slider K2/CH6.

Now adjust the Rate slider to full down Turtle position and repeat recording the values of the stick/channel combinations.

When finished please report your results. This will help us with what we are seeing in the telemetry files.

OK...here it is:
Rabbit Mode (Neutral / Full Deflection)
Ch1 - 50% / 100% & 0%
Ch2 - 0% / 99% & -100%
Ch3 - 0% / 99% & -100%
Ch4 - 0% / 79% & -79%
Ch7 - 0% / 100% & -101%
Ch8 - 0% / 100% & -101%

Turtle Mode (Neutral / Full Deflection)
Ch1 - 50% / 80% & 20%
Ch2 - 0% / 60% & -60%
Ch3 - 0% / 60% & -61%
Ch4 - 0% / 48% & -48%
Ch7 - 0% / 100% & -101%
Ch8 - 0% / 100% & -101%

Note: On my ST16, the Tilt was mapped to Ch7 and the Pan was mapped to Ch8. However, both pan and tilt controls operate the camera as expected.

Also, from what you can see so far, is there any evidence that I made any inadvertent control inputs that could have caused this tilt incident?

Thanks again all!

 

[DoomMeister]

Your Tilt slider and Pan Knob are mapped properly. I’m going to edit my earlier post so I don’t confuse anyone else. I wrote the wrong channel number on those. Your values from the Final Output Display appear to be at the proper ranges.

Flights from before February of this year seemed to have a normal range of values for the stick inputs to the aircraft. Those in February did not appear to have that range especially for full throttle down that is needed to cause the idle down of the motors after touchdown. The flights on Feb 13 had values close to 1100 for full down throttle (should be around 956 in Turtle or 683 in Rabbit). The flight on Feb 18 only got to a little over 1300.

The sticks at neutral have a value of 2048 in the FlightLog for the Remote. J1, J3, and J4 range in value from 683 to 3412 over their full range when the Rate slider is at Rabbit. J2 ranges from 956 to 3140 when the Rate is at Rabbit.
As for the issue you had with flight 144 on 02/18/2023 for whatever reason the signal for full down throttle did not have a full enough range to allow the motors to go to idle. When the motors do not go to idle the aircraft remains too buoyant and performs the classic skip and tip that you experienced. Once the tilt of the aircraft gets too large you cannot disarm the motors (I’ve had this happen to my TH landing on uneven ground). I always recommend having the Rate slider in Rabbit position for both takeoff and landing so the full range of stick input is available.

At this point I would suggest cleaning the potentiometers in the ST16. Control Stick Gimbal Maintenance.pdf

I would also suggest creating a new model and binding to it as shown in this video.

After binding to the new model I would also suggest performing the Accelerometer calibration.

Then try another test flight and see how it goes.

 

[DroneRanger]

Your Tilt slider and Pan Knob are mapped properly. I’m going to edit my earlier post so I don’t confuse anyone else. I wrote the wrong channel number on those. Your values from the Final Output Display appear to be at the proper ranges.

Flights from before February of this year seemed to have a normal range of values for the stick inputs to the aircraft. Those in February did not appear to have that range especially for full throttle down that is needed to cause the idle down of the motors after touchdown. The flights on Feb 13 had values close to 1100 for full down throttle (should be around 956 in Turtle or 683 in Rabbit). The flight on Feb 18 only got to a little over 1300.

The sticks at neutral have a value of 2048 in the FlightLog for the Remote. J1, J3, and J4 range in value from 683 to 3412 over their full range when the Rate slider is at Rabbit. J2 ranges from 956 to 3140 when the Rate is at Rabbit.
As for the issue you had with flight 144 on 02/18/2023 for whatever reason the signal for full down throttle did not have a full enough range to allow the motors to go to idle. When the motors do not go to idle the aircraft remains too buoyant and performs the classic skip and tip that you experienced. Once the tilt of the aircraft gets too large you cannot disarm the motors (I’ve had this happen to my TH landing on uneven ground). I always recommend having the Rate slider in Rabbit position for both takeoff and landing so the full range of stick input is available.

At this point I would suggest cleaning the potentiometers in the ST16. Control Stick Gimbal Maintenance.pdf

I would also suggest creating a new model and binding to it as shown in this video.

After binding to the new model I would also suggest performing the Accelerometer calibration.

Then try another test flight and see how it goes.

Thank you for the analysis and suggestions. To summarize, there was no inadvertent stick push on my part and no software issues. The crux of the problem is the ST16's pot components, materials, and construction. I recall doing a Hardware Monitor check before this flight and all the controls were looking good and smooth in their full range. I also recall going into Turtle Mode from Rabbit Mode during the flight and I probably landed in Turtle Mode. I won't do that again.

Having opened up both the H and the ST16 for various reasons, I will say that the component and materials choices do not seem on par with, say, DJI. While the engineering of the H seems sound and well thought out, I don't know if the same can be said for choices made by the design and manufacturing teams. Perhaps they increased robustness in subsequent models and iterations, but the H seems to have some shortcomings in this area.

After I get the replacement parts sourced, I will follow your preflight suggestions and report back to the forum on how return-to-flight goes.

Finally, I have noticed that the "soft-touch" coating surrounding the control stick assemblies has degraded and become sticky. This is true with a lot of devices these days and I wish manufacturers would stop using this material or use higher quality stuff. It is so annoying to have otherwise good gear become unusable due to this pervasive problem. I have a pair of stabilized Canon binocs that are now so sticky that you don't need a neck strap to hang on to them when you're not looking through them. Frustrating!

Anyway, thanks so much for your time and expertise in this matter and thanks to everyone else in the forum who chimed in on this incident. It has been very helpful!

 

[DoomMeister]

I had an ST16 recently that had become sticky around the stick gimbals. I used cotton swabs and isopropyl alcohol to clean them up. It took quite awhile to get past the stickiness, but it has been fine for the 2 months since.

 

[h-elsner]

Your FW is old and was never updated. Flight controller should be 1.35 and is 1.3 (from 11/2016). I do not recall what was changed, but I feel good with latest FW which was released in 2017 and is widely used.

Firmware Version
Kamera 3.2.26(A)
Gimbal 1.25
Autopilot 1.3

It takes 4.7sec from motor 5 stop to flight mode 16 (landed), no emergency mode visible - this is strange. But motor cut is after 3sec throttle zero as expected. The first part was interrupted by a short neutral position gap which restarts the 3sec timer.

Pulsing in the roll axis co incidents with stick pulses.


Throttle stick was never down, max -54%. This is not enough to start landing detection system in the flight controller.


br HE

 

[DroneRanger]

Your FW is old and was never updated. Flight controller should be 1.35 and is 1.3 (from 11/2016). I do not recall what was changed, but I feel good with latest FW which was released in 2017 and is widely used.

Firmware Version
Kamera 3.2.26(A)
Gimbal 1.25
Autopilot 1.3

It takes 4.7sec from motor 5 stop to flight mode 16 (landed), no emergency mode visible - this is strange. But motor cut is after 3sec throttle zero as expected. The first part was interrupted by a short neutral position gap which restarts the 3sec timer.
View attachment 29325View attachment 29324
Pulsing in the roll axis co incidents with stick pulses.
View attachment 29326

Throttle stick was never down, max -54%. This is not enough to start landing detection system in the flight controller.
View attachment 29323

br HE

I was holding the left stick down hard against the stop, so perhaps the limit at -54% is due to mis-calibration?

I will check firmware versions and update them before return-to-flight. Thanks.

 

[WTFDproject]

I was holding the left stick down hard against the stop, so perhaps the limit at -54% is due to mis-calibration?

Perhaps. But it would be better to establish a good baseline before going into calibration. It would be good if you run this sequence and upload the resulting new set of information.
1.) Remove the props.
2.) Start the ST16.
3.) Start the TH.
4.) Wait for bind.
5.) Start the motors (but do not increase throttle with motors running).
6.) Stop the motors.
7.) Ensure rate Slider is in "Rabbit".
8.) Cycle all control devices at least once.
9.) Ensure Rate Slider is in "Turtle".
10.) Cycle all control devices at least once.
11.) Shutdown the drone.
12.) Get a Screenshot of the Yaw (Rudder) EXPO page. Refer to the attached document.
13.) Shutdown the ST16.
14.) Restart the ST16.
15.) Upload the new flight log and screenshot for review.

Edit: Added new steps 7, 9 and 10.