Machinekit On BeagleBone Black

[Anıl Eren]

Hello,

I am working at a university student club and we are trying to make a CoreXY 3D printer with BeagleBone Black and the CRAMPS cape. We are new to this things. We installed the machinekit image from this link: elinux.org/Beagleboard:BeagleBoneBlack_Debian#BBW.2FBBB_.28All_Revs.29_Machinekit

The machinekit program works well with sample configurations but the problem is that, we need CoreXY configuration with CRAMPS board. We tryed the CoreXY configuration that Charles Steinkuehler made on the github but I cant give the link because it is not exist now. We copied only the config files to machinekit folder and opened it from the MK Launcher. The configuration gave an error and we realized that the configuration files is out of date.

After that, we try to start a configuration whose link is given :https://github.com/machinekit/machinekit/tree/master/configs/ARM/BeagleBone/CRAMPS. Firstly, we copied all the machinekit folders to the machinekit folder that BBB has in it. Then we started the configuration from the MK Launcher and it gave an error. It was an setup.sh error and we solved that problem by using the command "sudo chmod +x setup.sh" at the directory which has the setup.sh file. Then we started that configuration again and hal_pru_generic error appeared. This problem is solved by two steps :

1. In the pru_stepper.ini file, the hal_pru_generic.bin location was shown uncorrect. We find hal_pru_generic.bin by using the command "sudo find /-name pru_generic_bin" and we changed the file location to the correct location.

2. In the CRAMPS.ini file, the xenomai kernel option was commented since we use 3.8 xenomai kernel. The RT_PREEMPT one is commented and the xenomai one is uncommented.

After these steps, the configuration started without error but after rebooting the BBB, all the configurations (sample configurations included) were not started and any error didn't appear. After killing the halcmd task on the task manager, the program gave an error. The debug message on the error, we found that the system waiting the file in the /sys/class/uio/uio0. Howerver this file is not exist in that directory. Everytime we reinstalled the image and tryed the steps given above again and again but this problem always occurs after rebooting the BBB.

In summary, we need to learn how to install a working CoreXY configuration on the BBB. Also this configuration should start after rebooting the BBB. 

Note : The CRAMPS board didn't  manufactured. It's pcb design is revised to manufacture it by our equipment easily.

Thanks

Anıl

[Charles Steinkuehler]

On 1/8/2019 7:17 AM, Anıl Eren wrote:
> Hello,
>
> I am working at a university student club and we are trying to make a
> CoreXY 3D printer with BeagleBone Black and the CRAMPS cape. We are new to
> this things. We installed the machinekit image from this link:
> elinux.org/Beagleboard:BeagleBoneBlack_Debian#BBW.2FBBB_.28All_Revs.29_Machinekit
>
> The machinekit program works well with sample configurations but the
> problem is that, we need CoreXY configuration with CRAMPS board. We tryed
> the CoreXY configuration that Charles Steinkuehler made on the github but I
> cant give the link because it is not exist now. We copied only the config
> files to machinekit folder and opened it from the MK Launcher. The
> configuration gave an error and we realized that the configuration files is
> out of date.

I just updated my CoreXZ printer over the holiday break and got it
working with newer versions of Machinekit. The latest configuration
files are here:

https://github.com/cdsteinkuehler/machine-configs/tree/master/configs/CoreXZ

I don't think the CoreXY.comp is included in the mainline Machinekit
builds so you'll need to compile and install it using the comp command.

Note that my printer is a CoreXZ (X and Z combined, independent Y) vs.
the more typical CoreXY (X and Y combined, independent Z), but you can
use the same HAL component for both. The difference is in which axes
you connect to the component:

https://github.com/cdsteinkuehler/machine-configs/blob/master/configs/CoreXZ/CoreXZ.hal#L65-L73

Also note Nicholas Seward's CoreXZ is designed with different scales
for the X and Z axes. You can change this as needed, I think most
CoreXY have 1:1 scale vs. 1:3 on the CoreXZ:

https://github.com/cdsteinkuehler/machine-configs/blob/master/configs/CoreXZ/CoreXZ.hal#L75-L76

> After that, we try to start a configuration whose link is given :
> https://github.com/machinekit/machinekit/tree/master/configs/ARM/BeagleBone/CRAMPS.
> Firstly, we copied all the machinekit folders to the machinekit folder that
> BBB has in it. Then we started the configuration from the MK Launcher and
> it gave an error. It was an setup.sh error and we solved that problem by
> using the command "sudo chmod +x setup.sh" at the directory which has the
> setup.sh file. Then we started that configuration again and hal_pru_generic
> error appeared. This problem is solved by two steps :
>
> 1. In the pru_stepper.ini file, the hal_pru_generic.bin location was shown
> uncorrect. We find hal_pru_generic.bin by using the command "sudo find
> /-name pru_generic_bin" and we changed the file location to the correct
> location.
> 2. In the CRAMPS.ini file, the xenomai kernel option was commented since we
> use 3.8 xenomai kernel. The RT_PREEMPT one is commented and the xenomai one
> is uncommented.
>
> After these steps, the configuration started without error but after
> rebooting the BBB, all the configurations (sample configurations included)
> were not started and any error didn't appear. After killing the halcmd task
> on the task manager, the program gave an error. The debug message on the
> error, we found that the system waiting the file in the /sys/class/uio/uio0.
> Howerver this file is not exist in that directory. Everytime we reinstalled
> the image and tryed the steps given above again and again but this problem
> always occurs after rebooting the BBB.

The uio devices are created when loading the universal device tree,
which can be done at run-time (typical on Jessie) or via the boot
loader (typical with Stretch and the latest 4.x -RT kernels).

--
Charles Steinkuehler
cha...@steinkuehler.net

[Anıl Eren]

8 Ocak 2019 Salı 16:40:09 UTC+3 tarihinde Charles Steinkuehler yazdı:

Firstly, I want to say thank you very much for your explanation. According to your explanation, we solved the configuration problem and the motors that we tested are worked correctly, however we have another problem. We have .stl files that we need to print them but the machinekit supports .ngc type files so how can we solve this problem?

Thanks

Anıl

[Charles Steinkuehler]

On 1/18/2019 8:37 AM, Anıl Eren wrote:
> Firstly, I want to say thank you very much for your explanation. According
> to your explanation, we solved the configuration problem and the motors
> that we tested are worked correctly, however we have another problem. We
> have .stl files that we need to print them but the machinekit supports .ngc
> type files so how can we solve this problem?

https://reprap.org/wiki/Resources#CAM_Tools

--
Charles Steinkuehler
cha...@steinkuehler.net

[Anıl Eren]

Hello,

We have solved the .ngc problem by using Cura with extension NGCWriter. However, we have another problem with the message :

Near line 1 of /home/machinekit/Desktop/der/CFFFP_XY_Idler.ngc:
Named paramater
#<ini[fdm]velocity_extrusion_enable> not defined

Also we have different problem with some different .ngc files. The error message is :

G-Code error in comp311_2.ngc

Near line 21 of

/home/machinekit/machinekit/nc_files/examples/comp311_2.ngc:

Requested tool 4 not found in the tool table.

How can we solve these problems? 

Thanks.

Anıl EREN