Aerial Robotics Testbed » telekyb3 » rotorcraft-genom3

Thanks Anthony for the quick reply!

I agree that we need to upgrade the firmware, but since I have never done it before, can you guide me through the process? I just got this tutorial from my colleagues, and I am not sure if it is clear enough for me.

update about this point:

Ayham Alharbat wrote:

1. A temporary solution would be to downgrade the rotorcraft-genom3 component back to version 3.1, how can I do that?

With the help of Gianluca Corsini I tried doing the following:

# in robotpkg_path/robotpkg/robots/rotorcraft-genom
make checkout
cd work.Name/rotorcraft-genom3-3.3/
git checkout rotorcraft-genom3-3.1
cd ../..
make update .

Here is the terminal output:

===> Updating for rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
=> Building package update list
Scanned 5 packages
===> Rebuilding for rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
===> Deinstalling for rotorcraft-genom3
robotpkg_delete: Couldn't remove /home/ayham/ut/openrobots/lib/genom/pocolibs/plugins/rotorcraft-3.1.so
robotpkg_delete: Couldn't remove /home/ayham/ut/openrobots/lib/genom/pocolibs/librotorcraft-c-client-3.1.so
Removed rotorcraft-genom3-3.3pl0117172219~codels+pocolibs-client-c+pocolibs-server
===> Rebuilding for rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
===> Reinstalling for rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
libtool: warning: relinking 'rotorcraft.la'
=> Adding run-time search paths to pkg-config files
=> Registering installation for rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
Added genom3-pocolibs-1.21 to the requirements of rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
Added genom3-2.99.40 to the requirements of rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
Added openrobots2-idl-2.2 to the requirements of rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
Added pocolibs-3.1 to the requirements of rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
===> Done install for rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
===> Done update for rotorcraft-genom3-3.3pl0117172523~codels+pocolibs-client-c+pocolibs-server
make: Nothing to be done for '.'.

Now: I can connect to the robot successfully, where I see in the terminal:

rotorcraft-pocolibs: connected to chimera1.0, /dev/ttyACM1

but I cannot get any data to/from the robot. For example in Matlab:

• rotorcraft.get_sensor_rate().result.measured_rate return zeros
• rotorcraft.get_battery().result.battery.level also returns zero

Is it possible that I need to downgrade other component(s) to make this work?

On Tuesday 17 Jan 2023, at 17:56, Ayham Alharbat wrote:

I agree that we need to upgrade the firmware, but since I have never
done it before, can you guide me through the process? I just got
this tutorial from my colleagues, and I am not sure if it is clear
enough for me.

That tutorial seems obfuscated to me.
You just need to install tk3-paparazzi-1.2, e.g. from robotpkg in
hardware/tk3-paparazzi and tk3-flash in sysutils/tk3-flash.
Then, `tk3-flash flash chimera-N.0 <prefix>/lib/tk3-firmware/chimera`
where N is the chimera id of your board and <prefix> where you
installed stuff.

1 A temporary solution would be to downgrade the rotorcraft-genom3
component back to version 3.1, how can I do that?

With the help of Gianluca Corsini I tried doing the following:

1. in robotpkg_path/robotpkg/robots/rotorcraft-genom make checkout cd
   work.Name/rotorcraft-genom3-3.3/ git checkout rotorcraft-genom3-3.1
   cd ../.. make update .

Why don't you try one the methods that I suggested? What you did will
typically create some issues in your setup. The checkout command is
not meant to downgrade packages ...

Thanks Anthony! I will try to update the firmware then.

But is there a way to know/retrieve the ID of the board?

Anthony Mallet wrote in #note-4:

where N is the chimera id of your board and <prefix> where you
installed stuff.

On Tuesday 17 Jan 2023, at 18:53, Ayham Alharbat wrote:

But is there a way to know/retrieve the ID of the board?

The ID is normally printed on a label on the board. You also have it
in the kernel logs (dmesg) when you plug it. Something like:
[242628.082141] usb 1-2: Product: tk3-paparazzi
[242628.082144] usb 1-2: Manufacturer: LAAS/CNRS
[242628.082146] usb 1-2: SerialNumber: chimera-6

You can also use the tty device directly, e.g. /dev/ttyACM0

It worked!

For reference, this is what I did, after connecting the board to PC with USB:

tk3-flash flash /dev/ttyACM0 ~/ut/openrobots/lib/tk3-firmware/chimera

output:

Please power on or reset the device....
Probed STM32 bootloader
STM32F76xxx/77xxx bootloader version 9.3
Retrieving load data from /home/ayham/ut/openrobots/lib/tk3-firmware/chimera
Found data segment 0, paddr 0x8010000-0x80101ff, 512 bytes
Found data segment 1, paddr 0x8010200-0x8016f13, 27924 bytes
Found data segment 2, paddr 0x8016f14-0x80175b7, 1700 bytes
Found settings segment 3, paddr 0x8008000-0x80080bb, 188 bytes
Preserving settings in 0x8008000-0x80080bb
Reading data from 0x8008000-0x80080bb, 188 bytes
tk3-flash: outdated settings data, please flash with option -i or -u

so I did:

tk3-flash -u flash /dev/ttyACM0 ~/<path>/openrobots/lib/tk3-firmware/chimera

output:

Please power on or reset the device...
Probed STM32 bootloader
STM32F76xxx/77xxx bootloader version 9.3
Retrieving load data from <path>/openrobots/lib/tk3-firmware/chimera
Found data segment 0, paddr 0x8010000-0x80101ff, 512 bytes
Found data segment 1, paddr 0x8010200-0x8016f13, 27924 bytes
Found data segment 2, paddr 0x8016f14-0x80175b7, 1700 bytes
Found settings segment 3, paddr 0x8008000-0x80080bb, 188 bytes
Searching for existing settings
Reading data from 0x8000000-0x8000003, 4 bytes
Reading data from 0x8008000-0x8008003, 4 bytes
Reading data from 0x8008000-0x800ffff, 32768 bytes
Found settings data at 0x8008000-0x8008093, 148 bytes
Updating settings in 0x8008000-0x80080bb
Erasing flash page 1.
Erasing flash page 2.
Writing data to 0x8010000-0x80175b7, 30136 bytes
Writing data to 0x8008000-0x80080bb, 188 bytes

--- Settings --------------------------------------

serial          : chimera-12.0
boot-timeout    : 5                (bootloader timeout (s))
boot-addr       : 132              (boot address (muliple of 16kB))
boot-lock       : 0                (set to 1234567890 to update bootloader)
motors          : 6                (number of controlled motors)

I upgraded rotorcraft to 3.4, and then tried to connect again, and it works!

For testing, I can get the sensor rate, but, not all of the data is there, as in:

>> rotorcraft.get_sensor_rate().result.measured_rate

ans = 

  struct with fields:

        imu: 452.1847
        mag: 0.9900
      motor: 0
    battery: 0

and when trying to do:

>> rotorcraft.get_battery().result.battery.level

ans =

  0×1 empty cell array

I am using MATLAB 2019b, and here are my robotpkg_info:

nhfc-genom3-2.2~codels+pocolibs-client-c+pocolibs-server Near-hovering flight controller for rotorcrafts
pom-genom3-3.4~codels+pocolibs-client-c+pocolibs-server State estimation combining multiple independent sources
tk3-paparazzi-1.2   Autopilot for the Paparazzi hardware
tcl-genomix-1.8     Interactive genomix TCL client for controlling GenoM3 components
genom3-2.99.40      Software components generator for robotics architecture.
genomix-1.7.2       HTTP server for controlling GenoM3 components
pkg_install-20110805.12 Package management and administration tools for robotpkg
mrsim-gazebo-1.2    Gazebo simulator plugin for tk3-mikrokopter multi-rotors
eltclsh-1.18        Interactive shell for the TCL programming language.
optitrack-gazebo-1.0 Gazebo simulator plugin for OptiTrack Motion Capture
libedit-3.1r20160903 Generic line editing and history functions.
tnftp-20151004~ssl  The enhanced FTP client in NetBSD
tk3-mikrokopter-1.8.5 Autopilot for the Mikrokopter hardware
joystick-genom3-1.0~codels+pocolibs-client-c+pocolibs-server Export joystick readings
mrsim-genom3-2.1.1~codels+pocolibs-client-c+pocolibs-server Simulator for tk3-mikrokopter based quadrotors
optitrack-genom3-2.3~codels+pocolibs-client-c+pocolibs-server Export motion capture data from an Optitrack system
tk3-flash-1.4       Programming utility for Mikrokopter hardware
libmrsim-1.3        Simulator library for tk3-mikrokopter based quadrotors
arduio-genom3-1.3~codels+pocolibs-client-c+pocolibs-server Control an Arduino board running the arduio software
maneuver-genom3-2.1~codels+pocolibs-client-c+pocolibs-server Simple maneuver planner for aerial robots
rotorcraft-genom3-3.4r1~codels+pocolibs-client-c+pocolibs-server Low level controller for tk3 quadrotors
genom3-pocolibs-1.21 GenoM3 template for pocolibs-based components
digest-20080510     Message digest wrapper utility
openrobots-idl-1.6  Common IDL interfaces for genom3 components
pocolibs-3.1        Real-time communication library used by GenoM modules.
matlab-genomix-1.9  matlab client for controlling GenoM3 components with genomix
libkdtp-1.3         Spline-based Kinodynamic Trajectory Planner
openrobots2-idl-2.2 Common IDL interfaces for genom3 components

Another question:
Do I have to worry about the 'outdated settings data', I know that -u preserved the old settings to the new firmware, but do I need to set some of the new settings?

Thanks again, I really appreciate your help!

Also, rotorcraft.start() gets stuck in MATLAB and does not return.

On Tuesday 17 Jan 2023, at 19:44, Ayham Alharbat wrote:

For testing, I can get the sensor rate, but, not all of the data is
there, as in:

rotorcraft.get_sensor_rate().result.measured_rate

imu: 452.1847
mag: 0.9900
motor: 0
battery: 0

That's a totally different issue.
imu and mag number seem odd. What are you asking for, regarding sensor
rates? What is the output from rotorcraft?

Anthony Mallet wrote in #note-9:

What are you asking for, regarding sensor rates? What is the output from rotorcraft?

I did not fully understand your question, but let's say that before doing get_sensor_rate I do set_sensor_rate of the following values:
s_rate_imu = 1000;
s_rate_mag = 1;
s_rate_motor = 100;
s_rate_battery = 1;

And when I get_sensor_rate this is the output:
imu: 452.1847
mag: 0.9900
motor: 0
battery: 0

On Wednesday 18 Jan 2023, at 15:30, Ayham Alharbat wrote:

s_rate_imu = 1000; s_rate_mag = 1; s_rate_motor = 100;
s_rate_battery = 1;

OK, I was surprised by the 1Hz magnetometer, as this is a bit useless,
but at least this matches.
And what are the messages printed by rotorcraft after you connect?

So the main issue seems the motor data. Are the motor powered? What
can you see on the 7 segments display of the paparazzi?

Anthony Mallet wrote in #note-11:

And what are the messages printed by rotorcraft after you connect?

rotorcraft-pocolibs: connected to chimera1.2, /dev/ttyACM1

Are the motor powered?

Yes, and the ESCs make the startup tones.

What can you see on the 7 segments display of the paparazzi?

'0'

On Wednesday 18 Jan 2023, at 19:28, Ayham Alharbat wrote:

And what are the messages printed by rotorcraft after you connect?

rotorcraft-pocolibs: connected to chimera1.2, /dev/ttyACM1

If it prints just that, it might be that you are connected to the
auxiliary tty of the chimera. You should connect to the main tty,
i.e. endpoint 0, probably /dev/ttyACM0 if you don't have other serial
devices connected. Or just use the serial id of the board (chimera-N).

What can you see on the 7 segments display of the paparazzi?

'0'

OK, this means that the chimera sees 6 motors, so you should be fine.

Anthony Mallet wrote in #note-13:

If it prints just that, it might be that you are connected to the
auxiliary tty of the chimera. You should connect to the main tty,
i.e. endpoint 0, probably /dev/ttyACM0 if you don't have other serial
devices connected. Or just use the serial id of the board (chimera-N).

That was exactly the problem! Now I can connect to it as usual and I was able to fly it with no problems.

What can you see on the 7 segments display of the paparazzi?

'0'

OK, this means that the chimera sees 6 motors, so you should be fine.

Out of curiosity, I have a question, what does the output of the 7 segments display represent? is there a table that I can refer to in case of future problems?

Many thanks, Anthony for helping me fix this issue! I will try to compile what I learned from you here into some sort of documentation that can be shared with the team.