Hi,
is my assumption correct that only one camera + imu is used for the position estimation? Is it possible to include more tracking cameras?
Background:
I have 4 tracking cameras one at each side of my drone. Including these into the qvio-server should improve the position estimate right?
Re: VOXL2 USB-C Port for accessory/peripheral
As mentioned in the post above it seems like using the usb-c port on the voxl2 for peripheral devices is not possible. I can't find anything in the docs regarding this (perhaps I'm blind). I tested this "theory" by connecting an orbbec astra stereo u3 over the usb-port and the devices was not recognized. I plugged it into the usb2 port of the 5g modem and everything worked just fine. Could someone please give me an explanation for this or a link that explains reason for that behavior?
Hi,
is my assumption correct that only one camera + imu is used for the position estimation? Is it possible to include more tracking cameras?
Background:
I have 4 tracking cameras one at each side of my drone. Including these into the qvio-server should improve the position estimate right?
@Zachary-Lowell-0 Okay maybe this seems to work with foxy but as i mentioned I'm using humble and it seems like there is a mismatch between the version of the px4_msgs. I can provide a docker image to recreate the problem (at least with my voxl-px4 build).
To recreate use this docker image:
run with
docker run -it --network=host learningsql/voxl-px4-hector:tagname
Do:
cd ros2_ws
colcon build
source install/setup.bash
ros2 topic echo /fmu/out/vehicle_status
There should appear no msgs (at least on my system)
Fix this with:
rm -rf src/px4_msgs/msg
unzip /px4_firmware_msgs/msg_voxl_px4.zip
colcon build
source install/setup.bash
ros2 topic echo /fmu/out/vehicle_status
Now there should be msgs on the screen.
Idk if this is relevant for your px4_msgs fork but at least i found a solution and maybe this thread can help other people.
Thanks for you help!
@Zachary-Lowell-0 I'm using a newly flashed sdk 1.2.
voxl2:~$ voxl-version
--------------------------------------------------------------------------------
system-image: 1.7.6-M0054-14.1a-perf
kernel: #1 SMP PREEMPT Wed Mar 20 19:48:52 UTC 2024 4.19.125
--------------------------------------------------------------------------------
hw platform: M0054
mach.var: 1.0
--------------------------------------------------------------------------------
voxl-suite: 1.2.0
--------------------------------------------------------------------------------
And I'm using the repo you've send me in another post (https://github.com/modalai/px4-firmware/blob/voxl-dev/src/modules/microdds_client/dds_topics.yaml) and with that i followed your instructions from your github(https://gitlab.com/voxl-public/voxl-sdk/services/voxl-px4).
For ros i use ros2 humble in a docker container running the arm64v8/ros:humble-ros-base image with the px4_msgs from this repo https://github.com/PX4/px4_msgs.git.
I'm investigating a possible problem with the px4_msgs version in your px4-firmware fork and the repo i mentioned above. Maybe this rings a bell.
Is there anything else i can provide you with?
Hi,
I have the weird problem where the running agent is not publishing on all topics specified in the clients dds_topics.yaml. E.g. on the topic /fmu/out/senor_baro and /fmu/out/vehicle_odometry everything works fine but nothing is published on /fmu/out/vehicle_status and /fmu/out/vehicle_gps_position even though everything is specified in the .yaml with the corresponding px4_msgs. The vehicle status msg is published by uorb (at least "listener vehicle_status" in the px4 shell gives me a msg) and the microdds agent created a publisher, datawriter and topic. I've rebuild voxl-px4 and the microdds agent to no avail. I honestly dont know what to do next so any advice would be great!
@Alex-Kushleyev every esc should be correct and was tested using qgroundcontrol's actuator tab. I was able to get the drone to lift off in the "position mode". Idk why it did a backflip in manual. Since it is flying now i think i don't need further assistance. Thanks for your time!
Hi,
I'd like to know how i can add new topics to the uxrce_client. According to the px4 documentation i only have to add the topic in this yaml file and then build the client (included in px4) with it.
My problem is that in the forked version of px4 by modalai i am unable to locate the file since the above mentioned directory does not exist.
Any hints as to where the yaml file is located in the modalai fork are highly appreciated!
@Alex-Kushleyev
Thanks for your advice.
I'll look into the broken esc once my time schedule cleared up a bit and when my professor gives me the green light for the RMA since we need to allocate funding for that.
As for the new ESC. I flashed the new firmware modalai_esc_firmware_m0117_1_v0_39_RC10_eb6fb500 found herelink text. I also uploaded and verified the newly calibrated esc parameters. But the problem remains.
Using the voxl-esc-tools i can confirm that every motor spins in the right direction and that every motor spins at the given rpm.
Still when arming with q-ground-control 2 of my 4 motors only spin up to about 4000 rpm while the other 2 motors sit at 14000 rpm (giving more throttle at this point leads the drone into doing a backflip). Do you have any advice on this? Sadly i wasn't able to find any related forum posts (if you have one in mind a link would be great)
Hey @Vinny,
sorry for not updating you on this. I have remeasured the voltage yesterday and today with a battery. I suddenly had 12 V again so idk what that was about.
Sry i copied the topic from some other thread and made an edit error ^^ I'm using a voxl 2 chip with a M0134 esc and a power module v2 kit.
Today i have switched the esc with the only spare part and it powered up again. But it didn't give enough power for all rotors to spin at the same rpm, leading the drone to do a backflip and break one of my rotor blades 
- maybe a software error (do you advice to recalibrate the esc params again after switching the esc?) that i made today. I'll check again tomorrow.
Any advice on what to do with the orignial esc? Maybe i should test it with a laboratory power supply?
Hope this helps!
Hi @Vinny,
thanks for your advice.
I've tested the chip with a 4s battery but nothing changed.
The drone was (to my knowledge) never armed with the power plug - only with a 4s battery. For test purposes i spun only 1 rotor at a time at 100 rpm for max for 3-5 secs in order to see if they are working and if the configuration in qgroundcontrol is correct. Could this really be the reason the esc is broken? During a subsequent flight everything was working fine and after that the drone wasn't armed again for 3 months.
I measured the resistance between the two pads and it converges to:
Since you mentioned these values should be above >100 kohm i assume i did something wrong? The drone is part of a scientific research project at my university thus we have only limited funding available and hence only one spare esc here right now.
Is there anything you can deduct from the resistance values?
Best wishes!