Advanced Search

Exactly what I was looking for.

Thank you

Hi Forum,

I am trying to get Classic Ladder working with ]LinuxCNC RIO.

I try to get my VFD working with Classic Ladder. This is what works on the same machine with soft stepping.

loadrt classicladder_rt numS32out=100
addf classicladder.0.refresh servo-thread
loadusr classicladder --modmaster vfd.clp

When I start LinuxCNC with the FPGA connected I get the following error:Shared memory conflict or RT component not loaded!

ERROR CLASSICLADDER-   Ladder memory allocation error

I tried to remove the numS32out=100, but no success. I can load the program manually with the ladder editor, but it is still not in modmaster mode and won't work.

Someone an idea what the cause would be and how to fix it?

Hello Tommy 
I am using the following hardware.
qty 1   7I92M ethernet Anything I/O
qty 1    7I77D daughter card
qty 1    Lenovo think center M715q
I am re using the MSI servo drives, SEM servo motors and Heidenhain encoders, I plan on adding a A axis later.

Regards
John

Not sure If I bring anything but :
once you start machine and load tool z offset is present ?
I have problem with G43 on the first time after starting machine. Even my macro for tool change issue G43 after program is executed z offset disappears. Once I manually issue G43 everything is working than fine.

I have no idea how I can track what is causing this problem
my workflow :

Start machine
home all XYZ
load Probe (M6 T99) , during subprog execution z offset is visible for split of the second and than disappears)
G43 manually from MDI - offset now visible 
Probing part and zero G54
M6 T1 -all offsets are correct, if I don't issue manually G43 all the offsets with all the tools are messed up later on.


 

You could verify that the 7I97 analog output voltage makes sense
(it should be 10V*pwm_value/pwm_scale)

Maybe the drive needs some calibration

Also you might try adding a bit of series resistance in the analog output
for stability (say 47 ohms) in case the spindle drive input has a large capacitance

Take your pick- Amazon, eBay, Ali Express- just search for "accordion way cover", "accordion bellow", etc. and you'll find it. Random examples I'm not specifically suggesting:

www.amazon.com/uxcell-Machine-Rubber-Acc...Pleats/dp/B0087ZCY86
www.ebay.com/itm/315803295816

Also, not related to the jog question but you can avoid the debounce
by setting the "slow" parameter for the assosciated inmux  pins.

Sadly the choke didn't seem to change the behavior of the spindle. Maybe I am using the wrong style? It is salvage from some e-waste a few years back. I picked it of the few I have since the wire diameter was similar to the 20awg I was using to wire up the mesa card.  I'll attach a photo.

In an attempt to re-scale the spindle output, I was mapping out the actual speed vs the commanded speed. That was when I noticed some odd behavior. It has me wondering if I still have something wrong in the hal or ini.When commanding a speed below around 500 rpm, the spindle runs in reverse. Ex. M3 S100 runs the spindle counter clock wise at 370 rpm. When that command should have run the spindle clock wise at 100 rpm. After I start commanding S500 and up, the spindle resumes the correct rotation. Although, M3 S500 results in 20 rpm. Every commanded increase in rpm after that point is very nearly equal to actual rpm. Ex. Increasing S500 to S600 results in a reported speed of 120 rpm.

The spindle driver board has a wire for both forward and reverse. I control this with a SPDT relay. I checked this relay just in case, but it was not active while going in reverse below M3S500. When an M4 is commanded the relay does become active as it should. M4 commands behave identically to M3, but reversed. Commanded speeds below S500 are clock wise, and speeds at or above S500 are counter clock wise.

When I say actual rpm, this is read from showhal on spindle-vel-cmd-rpm. I have verified it in two ways, 1) the spindle driver has a tachometer on it, 2) I counted the rotations in one minute while the spindle was running at about 20-22 rpm.

My spindle PID are all set to 0. Is this normal for an ac spindle servo?

I will re-attach the hal & ini

I gave it a try today, and with my example script it's behaving better, but if you speed the loop up, it still shows up. This is with the following loop:

as you can see below, there is still something checking and reverting back to manual mode in between commands.
I have a branch set away with a sim config for my actual project I'm setting up that this bug reproduces with the actual UI I've been playing with, if if that helps (VERY work in progress ): github.com/digiexchris/qtpyvcp_grinder_t...ee/flexhal-mdi-issue
should be able to cd into rpi_grinder_touch and execute linuxcnc ./rpi_grinder_touch.ini and it should run it for you. Just take it out of estop, power on, and hit run. homing is optional to duplicate the issue. the script that's running as a loadusr module is in grinder-backend.py

Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: Manual
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: MDI
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: Manual
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: MDI
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Must be in MDI mode to issue MDI command
Current mode: Manual
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: MDI
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: Manual
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: MDI
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Must be in MDI mode to issue MDI command
Current mode: Manual
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: MDI
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Must be in MDI mode to issue MDI command
Current mode: Manual
Loop
Current mode: Manual
Current mode: MDI
Current mode: MDI
Current mode: MDI
Loop
Current mode: Manual
Current mode: MDI

Hi,

Maybe we can add a node configurator.

This then can be used to simulate the ethercat devices, by adding controls, like buttons, etc. for interaction.
And eventually it could be used as a plc controller, when adding logic just as classicladder.

It seems beckhoff has twincat for linux, running on debian. But it is not open source.
Did they write the ethercat master for linux themselfves?
They say in linux you can run multiple instances of twincat.

Make sure that you have a jog speed set, for example:

setp halui.axis.jog-speed 100

Posting some additional debug details in anticipation of getting some resolution.
This is from a single job that I ran yesterday. Exact same job run 3 times, as you can see there were 2 failures
and finally on the third try I was successful.

The second FAIL and the PASS are nearly identical.

Please chime in if you have any clues. I am using the subroutine from qt_auto_probe_tool.ngc

  

Aciera do you have any idea if there is more detailed description available about the C channel patterns of Yaskawa encoders?

Hey guys,

I first installed Linuxcnc 2.7 a few years ago just to have a look at it. Recently I finished the mechanics of my machine, which has dual Y steppers.
I found a good tutorial to add the second stepper in the HAL and INI, but I found out that 2.7 is quite old and the method described in the tutorial wouldn't work. I decided to do a fresh install of 2.9 per download page (LinuxCNC 2.9.3 Debian 12 Bookworm PREEMPT-RT ISO).

I did a latency test and got terrible results (120000ns while I got 15000ns on the previous install of 2.7). From what I can gather, this is due to the PREEMPT kernel. I can see a pinned post on the forum for "RTAI deb", but I have no idea how I would install this on my machine since it is not connected to the internet.

As you may be able to tell, I'm a bit out of my depth here. What I'm looking for is idiotproof instructions on how I can get 2.9 with the RTAI kernel on my PC, preferrably in a way that is as simple as installing the ISO.

Hope you guys can help.

HelloI've been working for a while on retrofitting a “small” Mecanumeric / Charlyrobot milling machine (the second one is the real manufacturer but they merged) at my work.

I work in a fablab in a high school in France. This machine has been at the school for many years, and originally ran on a Windows XP PC with specific software.The motherboard in the machine had a number of problems which I won't go into now, including a slightly scratched tool sensor.....I suggested modernizing it, making it more upgradeable and easier to repair.

So I had a few items purchased, including drivers and a Mesa 7i96S card. I reverse-engineered the wiring and managed to implement it pretty much as I wanted. The control panel was made from MDF using the big CNC machine, an Arduino using the great Arduino Connector project and a salvaged industrial touchscreen. I've added the ability for LinuxCNC to detect the opening of the cover to prevent the execution of a program. And to deal with minor oversights, I've added a relay to activate and deactivate the vacuum cleaner.
Finally, I've designed a small, unpretentious tool probe (which I can show if you're interested) with a microswitch and 3D printing.

 

More photos coming soon