My Trex suddenly jumps in the air and then slams back on the ground

I just got back from flying.
I have a
T REX 600.

I equipped it with a Pixhawk 4 mini a few weeks ago.

Since then, I’ve been around 10 flights with this TREX / Pixhawk. Problem-free.
For tuning, I put ATC_ANG_YAW_P on channel 6 so that I can set it.

After I finished setting I turned off channel 6. That was the only change in the parameters that I made before this last flight.

Today I wanted to fly again:

I increased the rotor speed.
The rotor speed was reached.
I slowly increased the grades from minus to positive. When I was about zero degrees, the helicopter jumped about a meter into the air and was immediately pushed back to the ground with full force. Everything happened in a fraction of a second.

Then I switched off the engine to be on the safe side.
The only damage that can be seen is that it has turned the tail rotor a little around its axis on the tail boom. all servos seem to be working normally.

I really doubt whether I can trust the FCs.

Here are the parameters and the log file.!AjSq727ChpJzdpEXNWfhVvSR5FY?e=MgFRl8!AjSq727ChpJzdSKOQDwVlrZEYEU?e=UjuSef

I only saw that the RCIN shows input values ​​that I have not entered myself. How can that be?

I’m very frustrated because I’ve bought and installed two FCs in the past few weeks.
Both are not working properly or I am making mistakes.
And our flying saw blades can also be dangerous when they react so unpredictably.

I hope you can help me.



@heri I am sorry to hear about the issues that you have been having with your FC installs. I will try to look at your data but probably won’t be able to look at that right away. I will look at your parameters when I have a few minutes to spare and maybe I can find something there.

@heri what version of firmware are you using? From the param file, it appears your are flying master which I would not recommend.

I took a quick look at this flight. I did see inputs but I wasn’t sure which were intentional and those that were not.
@picoflug has been having issues with inflight shutdowns with the same FC(PH4-mini) on arducopter 4.0.6. The Developers have some concern with this behavior.
What RC receiver are you using? The log said it was connected via SBUS? Is that true? I also saw that you are using the passthrough option on servo 7. What are you using this for?

Are you using any special mixing in your RC transmitter? Mixing one channel with another? Or using the helicopter setup in the transmitter? I recommend that users use the plane setup in their RC transmitter because all of the setup unique to helicopters is done by the flight controller. There are no special helicopter functions that are needed in the RC transmitter.
Any more information that you can provide will help me understand the issue.

Hello Bill,

you are a real treasure!

To your questions:

  • My flight:
    I increased the rotor speed. The collective rotor blade angle was negative.
    Then I slowly increased the angle of the blades. When it was around zero, I had the violent reaction of the helicopter. I didn’t consciously move the other sticks. If anything, just a little.

  • When I started experimenting with the Pixhawk at the end of 2020, the version was 4.0.5. the stable version. And I assumed I had loaded the stable version. From 4.0.5. I haven’t upgraded to a higher version yet.
    I’m still a beginner at ardupilot. Maybe I made a mistake while loading the firmware.

One question: what is the “master” version? I didn’t load such a version on purpose.

  • Therefore I do not assume that I have the same problems as @picoflug, because I do not have version 4.0.6. . At least I think so.

  • I use a Futaba remote control (T14SG) with a “Fasstest” receiver. The receiver is connected via the S-Bus.

  • I set RC7_OPTION to “4”. With a rocker switch on my remote control I trigger the “RTL” mode with channel “7” on my radio. I don’t know why Servo7 is set to “passthrough”. Presumably this is the standard value for this parameter?

  • No, I don’t mix any functions in the transmitter.
    Yes, I use the helicopter setup. But all mixers are deactivated. I’ve been using traditional FBL systems for several years.
    But you’re right. In the future I will also use the airplane setup. Then you don’t have to deactivate the mixer.

It’s now 11 p.m. in Germany. Since my last post, I’ve “flown” the TREX for an hour with the engine disconnected on my workbench. Just to see if the twitch comes back. There was no twitch.

Tomorrow I will strap the helicopter to a heavy table and do a test flight with the engine running to see if this triggers a reaction.

Is there a general problem with the PH4-mini?

I can only thank you again for your help.




how or where can I find out which software version I have on my board.

You made me feel insecure.

In the last few weeks I have also installed a PH4-mini in my Vario Bell 230. This has version 4.0.7. . I believe!

It can easily happen to mix up something when tuning on two helicopters at the same time.




Maybe I found something.

I looked at the following log parameters:

  • RCIN C8 This is the switch for me to increase the engine speed.
  • RCOUT C8 is the reaction of the PH4-mini as it slowly increases the speed, within 20 seconds.
  • POWR Vcc is the on-board voltage.

You can see me flipping the switch on the remote control. RCIN C8 goes to the full value.
Then the value for RCOUT C8 increases slowly. While the value for RCOUT C8 rises, you can see how the on-board voltage oscillates and drops below 5 volts. For whatever reason.

I suspect the following: The FC briefly set the swashplate servos to neutral (PWM 1500) when there was undervoltage. PWM 1500 means 5 degrees positive in my setup.
A fraction of a second later the FC set the servos back to my controlled values at this moment, therefore probably one degree minus.

I had often observed this behavior when setting up my TREX on the workbench. As a beginner of ardupilot, I didn’t think anything of it.

Can you see more about this in the log files?



The version is found in the log under the MSG label. That is one of the first messages written to the log. You were running 4.0.5. So you were not running master. This is consistent with Holgers firmware version where he was having issues.


Hello Bill,

Thanks for your hint where I can find the version.

I am now pretty sure that I have found the cause of the strange behavior of my TREX.
And I have to be honest and say “mea culpa”.

The sinking of the Titanic had many small causes and not just the iceberg.

That’s the way it is with me and I want to briefly describe it as a warning to the other pilots.

When I bought the PH4-mini, there was also a power board in the set. However, this is only designed for a 10s lipo battery and my TREX flies with a 12s lipo battery.

Therefore, I did not supply the power board with power from the drive battery, but with the additional battery for the receiver and the servos.
A 2s lipo battery.

Unfortunately, I had exchanged this 2s lipo battery for a 2s LiIon battery.

These have a slightly lower voltage.

The already low supply voltage for the power board was again a little lower.

When the rotor speed is increased, the helicopter wobbles back and forth on its runners. As a result, the swash plate servos work hard, as the PH4-mini tries to compensate for the movements of the helicopter.

As a result, the servos need a lot of power, which puts a further strain on the voltage of the receiver battery.

And all of these small causes lead to the PH4-mini power supply dropping below 5 volts. Only minimally, but that’s enough.

Holypro writes in its manual that the PH4-mini can withstand an undervoltage of less than 5 volts, but it is not recommended.
The manufacturer should warn more clearly here!

Everything else was my fault. You never stop learning.
Even after decades of helicopter flying, people still make careless mistakes.

Thank you very much for your help.