@Nikos-Mavr You're welcome! And can you please start a new forum post for this new issue? Thanks!

@Cliff-Wong 267ff340-42c9-4c97-86a8-3b0b61ff7b4b-image.png

I am experiencing a severe issue with trajectory execution in the SLAM/Mapper mode where the drone exhibits an inverted and cumulative 'mirror' behavior regarding position setpoints.

As shown in the attached image:

If I set a first waypoint directly in front of the drone (1st Point), the drone flies backward instead, covering the intended distance but in the exact opposite direction. From that new incorrect position, if I command the drone to return backward by the same distance (2nd point, aiming to return to the original starting location), the drone does not return. Instead, it flies forward, doubling the distance in the opposite direction from the newly commanded point.

It appears as though the drone is accumulating a starting error or interpreting the coordinate frame in a completely mirrored way (FRD vs NED mismatch), where both the direction and the origin reference point are inverted. The dashed lines in the image represent the intended path, while the solid arrows represent the actual path flown. I am on SDK 1.6.3 with factory settings.

Hi @austin-c ,

You can find the latest M0138 firmware with regen completely disabled here

Also, I added the esc config that i've been using for testing here

For step tests, i have been using pretty aggressive step command:

./voxl-esc-spin-step.py --id 2 --rpm 2000 --step-amplitude 4000 --step-frequency 2.0 --timeout 2.0 --cmd-rate 2000 --enable-plot 1

In order to make the ESC not de-sync (due to long demag time of the motor), the following params were adjusted:

max_rpm_delta set to 4000 (you can go lower), to cap the maximum rpm transitions. Although in my test, the step is 4000, so this had no effect timing_advance: 45 -- this is pretty high commutation advance to trigger commutation earlier so that there is more time for back-emf sensing -- high advance wiil reduce motor efficiency and will cause the motor to heat up a bit more sense_advance: 30 -- delay back-emf sensing so that we don't sample back-emf during the motor coil demagnetization more details : https://gitlab.com/voxl-public/voxl-sdk/utilities/voxl-esc/-/blob/dev/voxl-esc-tools/doc/low_kv_motor_tuning.md

I had another thought.. What if you use a small battery to absorb the voltage spikes? Basically hook up a 6-7S battery with a beefy Schottyky diode (10A) set up with Cathode (negative) pointing to the battery.

you will have to make sure the power supply voltage never exceeds battery voltage + diode drop. you will also need to make sure that you don't over-charge the battery with repeated spike absorption.. It is possible to have some sort of self-discharge circuit to slowly discharge the battery to a certain voltage (using a Zener diode + resistor or something similar).

Anyway, just an idea..

Try out the no-regen firmware..

Alex

@Myles-Levine , if you are masking out 90% of down-facing camera, then it is probably useless any movement of the features will get it out of the unmasked region very easily and feature will be dropped.. It may hurt VIO as there may be features going in and out, only trackable for a few frames (just adds to the complexity).

You should try to move the down-facing camera to free up it's FOV or even have it facing slightly angled to the back of the drone..

Alex

@Dan-Jennings I'm also guessing that the driver may not be detecting the hardware so is not able to get the data for the battery_status. Have you tried swapping some other hardware to see if the problem is associated with a particular unit?

@emarcphera @Eric-Katzfey thank you so much after upgrading to voxl-mainline im getting GPS readingScreenshot from 2026-02-28 01-45-10.png

Thank you for the quick reply.
I will try doing as you instructed and post the results.

Thank you.
Kim

@RyanH ,

Sometimes there is small pieces of random magnetic debris that gets stuck in the motor.

First, you could try using compressed air to clear out the motor.

Additionally, It could help to spin the motor up and down using voxl-esc command line tools a few times to clear it out (without prop is best, just for safety). Spinning the motor without prop a bunch of times could get the debris dislodged and actually broken up by the moving motor parts, so that the debris is no longer large enough to cause a problem. Reversing motor spin using voxl-esc test tools can help dislodge the debris (you can commend negative power/rpm).

You can find useful info regarding motor testing here : https://docs.modalai.com/voxl-escs/faq/

Alex

@bendraper
Hello,
I hope this quick clarification helps for warnings:

CMake Policy warnings: OV internally (external dir in the repo) still uses boost and references an older cmake version -- it's a harmless warning for devs (as you are building from scratch)

Yaml-cpp lib warnings: This is a third party dependency that gets compiled from scratch during the make/build process. We use C++ 17 but the library is based on an older C++ version; nevertheless, it compiles fine on C++ 17 with minor warnings. -- it's a harmless warning

TrackOCL warning: ModalFlow is a core library written in C++ 17 that allows function call usage as done in TrackOCL, i.e., the narrowing conversion is harmless in this case.

That being said, we are actively working on the dev branch, so all flight--impacting warnings are hopefully caught early. We currently have no flight-impacting warnings/bugs identified, but if you found one warning/bug that caught your attention, please feel free to share.

All the best,
ZBFT

@restore , the change to enable the maximum resolution raw output for OV64B was made at the end of October 2025, so you would need at least SDK 1.6.0, I believe. There was a change in the camera pipeline (not the camera driver) to allow such a large image size. Are you able to test on a newer SDK (just test the latest if you can)?

Additionally, the resolution 9248x6944 cannot be directly used in MISP debayering (which uses OpenCL). There are some special requirements on the image stride, which width of 9248 does not match. So what will happen is the cpu will realign the image before feeding it to the gpu, but it is a lot of data to copy for a 64Mpix sensor. So I made a small change in width from 9248 to 9216 (just cut off 32 pixels) and it can be fed into the gpu directly. That change is not published yet, but i can share it.

Lets first confirm that you can use the 1.6.x SDK to start the camera server at 9248x6944 and you should be able to save raw bayer and view the misp output.

you can always double check the list of available raw resolutions using voxl-camera-server -l

by "start the misp output stream", i mean that you need to have a client that subscribes to the misp output stream, so that the frames start going and AE can work, sending exposure and gain updates to the camera, so that the image is properly exposed. For example, viewing the stream in voxl-portal or just using voxl-inspect-cam hires_misp_color to get the data flowing, then you can save the raw bayer, which will have the proper exposure and gain applied.

Alex

@restore There are a few examples of code that listens for Mavlink messages in our SDK. For example, in voxl-mpa-tools, take a look at the source code for voxl-inspect-mavlink.c. https://gitlab.com/voxl-public/voxl-sdk/utilities/voxl-mpa-tools/-/blob/master/tools/voxl-inspect-mavlink.c?ref_type=heads

Probably would be a good idea to run voxl-inspect-mavlink on the mavlink_onboard and mavlink_to_gcs pipes to see which one has the desired Mavlink message in it. Then create your own application that listens for the Mavlink message on that pipe.

@Alex-Kushleyev The Alfa Network AWU2036

@qubotics-admin First you need to determine the version of PX4 running. With ssh or adb you can run the following at the command line: px4-ver all. You will see PX4 version and Vendor version. So, for example, for px4 version 1.14.0 and vendor version 8.134.3 then you would checkout the tag v1.14.0-8.134.3-dev from the repo to see that commit.

@ndwe Whether you use a VOXL or Jetson as a companion computer shouldn't make any difference unless you want to use the VOXL SDK. The version you run is totally up to you. If our v1.14.0 based version works for you then use it. We do plan to transition to v1.17.x in the near future. I don't think the issues you are seeing are related to something specific in the flight core v2 hardware setup. I would think that if you swapped it out for some other brand you would see the same issue. I would ask on the PX4 forums to see if others have experienced the same issues.

Hi @Alex-Kushleyev, sorry for the late reaction, it took some time before i could update to the new SDK (1.6). the system now works normally with the resolution 2704x1540, and downscaled to 1920x1080 for the MISP pipeline. Thanks for the help!

@syamala-kotireddy ,

If the ESC is not showing any signs of life (no blue led blink), it probably means the ESC power regulator is not working properly. However, you mentioned that VOXL2 is actually booting fine? Can you measure the voltage on the connector / pads that provide 5V power from M0129 ESC to VOXL2?

Please see the following post, where a capacitor on the ESC next to the main voltage regulator was knocked off, resulting in ESC not starting up. https://forum.modalai.com/topic/4151/voxl-mini-4-in-1-esc-missing-capacitor/ . However, in that post, i think VOXL2 was not booting either. In your case the issue may be something different.

It is worth carefully inspecting the ESC.

Also, if you have a power supply that can measure current, you can set it to 8V and measure the current draw of the ESC alone, it should be very small (maybe 20-30mA) but not zero.

Alex

@Piyush-Singh , can you please provide more details

how exactly all cameras are connected to VOXL2 mini (including interposers, flexes), etc. A diagram would help a lot which kernel variant are you running on your voxl2 mini (2.0.0 or 2.0.2).

Are all the other cameras working if you disable the TOF sensor in the camera server config?

Alex

@cbay voxl-fpv-px4 is based on voxl-px4 but has some custom tuning code / parameters for our fpv drones that isn't in voxl-px4. So, it's kind of a superset. But there isn't anything related to low level i2c drivers so I have no idea how that could be causing this issue.

@Jetson-Nano , since we currently have no plans to do the MIPI integration of this sensor, you always have the option to use USB connection and voxl-uvc-server to publish images via MPA. You would need to check if the sensor supports the standard UVC interface and try it out...

Alex

@tom Thank you!