3.5-rc6 Pixracer DSM problem

Hi. I’m having serious issues with my DSMx receiver if I update from 3.4.6.
Setup: AUAV PixRacer + AUAV Power module, DSMx receiver (OrangeRx R615X DSM2/DSMX) running in DSMx mode, OrangeRX transmitter module, sending 8 channels (does not affect outcome).

It works well for a few seconds, then 10+ channel starts to move randomly, rapidly. First reproduced in-air, close to the ground. Turned off the radio, radio failsafe kicked in, landed.
Latest 3.4.6 release works well with the same setup.
First I observed this a few weeks ago, about 2 RCs ago. But I had some other problems to solve, so reverted to stable.
3.5-rc6-pixracer-dsm-issue.zip (330.6 KB)

1 Like


Thanks for the report.

We discussed this issue on the dev call today but didn’t come to any conclusions except that we may need to try and reproduce the problem ourselves to get to the bottom of it.

There was also some discussion by @LuisVale that there were some “known issues” with the receiver although we didn’t get into the details.

So I’m not sure exactly how we’re going to get to the bottom of this, perhaps we need to buy the exact transmitter/receiver that you have and test (if you have links to where we could buy them that would be helpful) or somehow try and debug with your help (i.e. send you binaries with extra logging, etc).

Anyway, just wanted to let you know that we haven’t forgotten about this and with other issues being resolved daily, this is creeping up the to-do list.

Thanks, good to know :slight_smile:

Transmitter: Turnigy 9XR pro
Transmitter module: OrangeRX DSMX / DSM2 Compatible 2.4Ghz Transmitter Module (JR/Turnigy) HK link
Receiver: Spektrum compatible DSMX (ebay)

On sunday, I was accidentally flying it again. The problem reproduced. The drone flew away moving randomly, as it picked up the noise from the RC input, until I turned off the transmitter, crashed into the ground.
Right, this is my fault, forgot to revert the FW.

Now, I only have 3, zero sized logs on my SD, so sadly I cant upload it this time.

If any of the team members live in the area, I would happily borrow my drone for testing. Otherwise I can do some more flying to produce logs, but only with a wire to avoid escapes :slight_smile:

Can you please suggest a dsmx compatible satellite receiver? I’d need one for my other drone. This would be a good chance to test any differences.

I bought a said-to-be-real Spektrum receiver, this one: https://www.amazon.com/gp/product/B01ABV7K5A/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1

I can reproduce the very same problem with it. I’ve found no relation between radio settings and behavior.


Ok, thanks for the feedback. We haven’t forgotten about this issue, it’s actually holding up the official release to some extent. We have two developers who use spektrum in the dev team but they haven’t been able to reproduce the problem.

One odd thing is that then transmitter is sending 8 channels you say but we’re seeing a lot more channels in the RCIN message.

One idea that came up on the last dev call was that perhaps some other extra information (telemetry data?) is being encoded by the transmitter in the upper channels.

You don’t happen to have a way to connect the receiver to your computer? Perhaps using a serial->USB converter (aka FTDI cable)? We somehow need to see the stream of characters that’s being sent from the receiver to the flight controller.

I’ve gone ahead and ordered the same transmitter and receivers that I think you have. Once it arrives (hopefully within the next week) I may need to ask you for details on the transmitter configuration:
transmitter: https://hobbyking.com/en_us/turnigy-9xr-pro-radio-transmitter-mode-1-without-module.html
transmitter module: https://hobbyking.com/en_us/orangerx-dsmx-dsm2-compatible-2-4ghz-transmitter-module-v1-2.html
receiver: http://www.ebay.com/itm/311416121989

There are quite some settings on the transmitter, will send the full config once i’m at home.
I’ve played around with transmitter settings (sending only 1-8 channels; 1-12channels; also played with timing 11ms-35ms as far as I remember) none of those seemed to make any difference.
I have an FTDI and maybe a digital logic analyzer. I’ll try to produce some test data for you, but I may need a bit of time for that.

"perhaps some other extra information being encoded"
i would be surprised if there are more data being sent. Although sometimes it seemed some channels were duplicated instead (roll seemed to move to channel 10 and 12, sometimes somewhere else). It also possible this is completely wrong, just observed Mission Planner Radio page, trying to see some relation between stick movements and display.
I still suspect this is some timing / parsing / threading issue.

Moreover, yesterday I tried to flash the latest PX4, there i constantly got RC loss/found messages. May be this is coming from there…
Please do not believe me, I’m just brainstorming with these weird ideas.
I’ll update you ASAP to have facts instead.

I had problems with FTDI, what are the correct settings for it or what tool should I use?
My Radio settings (only the External RF settings)
Mode: PPM
PPM Frame: 24.5ms 500u - (there is a minus at the end)
there’s nothing else.

Recorded the communication with a Saleae logic analyzer. Exported to various formats, but one of the files can be opened with it’s own free sw (https://www.saleae.com/downloads).
I dont really trust its serial analyzer with autobaud.

1-9ch_ppm24.5-500u-.zip (532.7 KB)


Sorry for no reply for so long. It looks like Tridge has found and fixed the problem. It was a bit tricky - it had to do with some DMA changes we had made to improve performance (this includes changes to allow the 8kHz update rate from our IMUs).

I plan to push out AC3.5.1-rc1 on Monday so people can test the fix. Either that or I’ll make a binary available separately for initial testing and then follow up with AC3.5.1-rc1 if that goes well.

Thanks for your patience!


Here’s a .zip file with the fixes included on top of AC3.5.0.

OlliW has found the same problem and he’s testing over in this other thread. Copter-3.5.0 has been released

Hi @rmackay9 ,
Checked, cannot reproduce the problem anymore.
Thank you very much!

Sorry for late response.

1 Like


NP, thanks for checking and posting the results!


This question is perhaps slightly off topic, but I’m having this DSMX issue with an AC3.5.dev version of the firmware, but no DSMX issues when I load stock AC3.5.5. Is there any way in Mission Planner to determine what FW version this particular dev branched out from? Also, does anyone know if I should expect the same behavior with the X2.1 from mRobtoics? I do not have this problem with Pixhawk running the same dev branch, so it would seem that it’s a combination of hardware and software that creates the issue.

There should be no difference between the Spektrum port on the Pixhawk1 and the X2.1
Same hardware.

Your dataflash logs should identify which commit the firmware was built from. Given that, you can use git to look at the history for that commit.

Have same problem with mRo x2.1 simple or 777 version with DSMX receivers ( Spektrum SPM4649T or OrangeRx R111XN ) for very long time and still can`t use any of them, binding ok but channels not moving.

Checked all latest and older arducopter pilot rover firmwares, no lucky

SPM4649T looks like SRXL receiver, so powered from separate power pin 5V and anyway not working.