Advanced Search

Hello, I am a beginner trying to learn LinuxCNC. Please note that not only that I am totally new to LinuxCNC, but I am also totally new to working with CNC machine, G code, and the C programming language, so please forgive me if I ask stupid questions. Since I don't have any lathe or mill that I can experiment with (without messing things up), I am playing with LinuxCNC on a virtual machine (Debian inside VirtualBox). I have successfully managed to build LinuxCNC using run-in-place method as instructed in this guide linuxcnc.org/docs/master/html/code/building-linuxcnc.html .

Right now, I want to understand how .c kinematics files work in LinuxCNC. So, as a first step, I did a little experiment by modifying the inverse kinematics function in trivkins.c (located at src/emc/kinematics/) as shown.

My understanding is that the inverse kinematics translates the position of the tool into joints variables, so by putting a negative sign in front of pos -> tran.x before passing that to identityKinematicsInverse function, I expect the machine to move in -x direction whenever I command the machine with G code to move in +x direction (like G0 X10). 

After I have modified trivkins.c, I compiled the file and rebuild LinuxCNC by going into /src and use  ./autogen.sh, then ./configure, and then make. 
However, as you can see in the photo attached   , when I open axis.ini config (which I believe it uses trivkins.c), the machine still do engraving in the +x direction. Nor did it go to X-10 when I command it with G X10.

I tried modifying trivkins.c in different ways, such as setting some directions to a constant, but I still experience the same result. So what did I do wrong here? Why I see no changes in AXIS after I modified the kinematics files. I am pretty sure the trivkins.c was compiled because when I deliberately put syntax error line in there, make threw an error

This is looking amazing!   I am impressed with how smooth that boomerang video looked!  Absolutely fantastic work my friend!  Thank you very very much for your hard work! 

Chris

Hi all,

This is my first time playing with electronics. I had almost zero knowledge but during the last 3 months I understand quite a bit now. I was quite afraid of it (especially Linuxcnc, I was keeping Acorn ahead because of the ease) but I am quite enjoying it now. I have wired my lathe previously with a printer BOB and it ran successfully. Now I am getting it hooked with a mesa 7i96s card with nema 23 motors and a spindle encoder. Getting a mesa board here in India is quite hectic so I will ask some questions frequently so that I dont mess anything up. 

1. What are the 5V, GND pins on each stepper axis of the Mesa card used for?

2. I heard that you need shielded wires for steppers. So do I need to shield the pulse and direction wires from the mesa card to the drivers along with the wires from the drivers to the motor or just the latter is enough?

3. Even though I searched quite a bit about the enable signal of the drivers, I couldn't land on a decision on how to wire it. I understand that leaving it unconnected or having the pins low enables them and having them high disables them. I really dont want the steppers to be enabled all the time, maybe F1, F2 from Linuxcnc will disable-enable them. Any wiring suggestions?

4. There is a Frame GND on the Mesa card. Do I need to ground it?

Thanks in advance..


EDIT: I now understand what the 5v and GND is for on the mesa stepper axis segments. I somehow missed it on the manual..

My suggestion is to give up ecat-io and look for other solutions, such as mesa card or fpga or arduino (I don't know how fast they are )

Hi. Yes. I'm from RSA.
I have only used the rotary inductosyn  stuff as shown above, but the principle is the same as a normal resolver.

The linear/rotary inductosyn type has a low inductance (as far is I understand) and the mesa or pico products doesn't work for them, bwich is why I went for the farrand stuff.

But as noted by jmelson above, there are products for the normal wound type resolvers from pico, and mesa electronics.

so it depends very much on the rest of the hardware in the machine and if you can re use that. 

the sa-cnc club meetings are mostly online, so you don't have to travel to Johannesburg  to join. 

Reference routing (inspiration) for the above wiz part schematic. I doubt all that analog/digital isolation insanity is mandatory, but some does claim those motion control being noisy [rolleyes], and so best practice always just add them in if you are routing this. All what's left to do is the 3V3~5V translation and hooking up 50 gpios. Tomorrow. Maybe. 

My Riocore Kit looks forward its own faraday cage (EMP warfare protection [rolleyes x2[). Let's get in touch, Olivier - I'll send you one (or two!). 

You may already know this but here is the function qtvcp calls for tool diameter probing (probe_tool_z_diam):
github.com/LinuxCNC/linuxcnc/blob/master...obe_routines.py#L314

Can you describe, using this code where the problem is?
If I understand you properly, you are saying the tool doesn't lower enough to touch the probe on the side of the tool when soide probing in X+ direction.
Does that mean when doing X- probing the tool _is_ low enough?

Chris
 

Can you post the result of

ip a

Maybe  a routing/ip address issue?

onceloved wrote:

However, it outputs the count value and has no direction. The encoder forward counter value increases, and the encoder reverse counter value also increases instead of decreasing.

Seems odd
0x2020 (or 0x2040 depending on where you look) shows a signed 32 bit integer for encoder.
I was able to translate the manual into English with Google Translate
drive.google.com/file/d/1GwtOZH0HaHGVrvv...SEO/view?usp=sharing

Most people are running Debian 12 and Linuxcnc 2.9.3 becasue the ISO on the downloads page also allows ethercat to be installed from repositories as per the sticky in this section.

Pico Systems has a single channel resolver to quadrature converter for US $150, that is compatible with most so-called brushless resolvers. It is not compatible with variable reluctance resolvers, they need too much excitation current. Typical resolvers it works with are Kearfott/Singer and Harowe Controls Harosyn.
See pico-systems.com/osc2.5/catalog/product_...ath=5&products_id=31
For more info.

AndyDM01 wrote:

I am a little bit confused about what to download, the one Rodw put on google drive or the one that was posted on 03/08
 

The ISO on the downloads page has been updated with my image.. 
Get it from here becasue its now official
linuxcnc.org/downloads/

Installing linuxcnc on a standard Debian instance is described in the docs here
linuxcnc.org/docs/stable/html/getting-st...th_preempt_rt_kernel

PCW, thanks for your answer!Yes, indeed. If I start the PC, my auto internet connection via RTL8111 is online but no websites reachable. eno1 connection to the Mesa card is working but with slower ping. If I switch to a faster ping with "sudo ethtool -C eno1 rx-usecs 0 ", my auto internet connection is online, but I can't reach any website as before. That's only possible if I close the eno1 connection.
I had to install lshw first. The feedback is - not the same irq for both connections.   

I just installed OBS to make a video of that phenomenon.  Maybe I do that tonight. Nope. OBS doesn't recognize my mic. I have to research.

I am shy to install this machine completely new, because I have installed several things and packets up to this point and did not document them. Beginning with libre office, Atril, Firefox, Chrome, Telegram, Geany, Visual Studio (all no problem to install again) and several packages to help with all of these Debian/LCNC problems on my way. Is there any tool or command to show me these not included packages in Debian 12? That would help me a lot.

Thank you for your time and effort...! 

Just like other stepper systems for LinuxCNC the position feedback is from the step generator. For Remora this is the raw step count from the controller board. The LinuxCNC Remora component then calculates the position in mm by using the joint scale value.

Like my earlier comments, the smallest move / positioning accuracy can only be as small as 1 step.