Advanced Search

Here is my latest EtherCAT card, the EaserCAT 6000, with 12 digital inputs, 4 digital outputs, and one frequency counter input for THCAD output.
 
It's the one down in the corner. There are three RTelligent ECT60 EtherCAT stepper motor drivers as well.
The inputs are limit switches for X, Y, Z, e-stop, floating, break-away and arc-OK.
Plasma-on is controlled by output 1.
There is one frequency counter input. Come to think of it it wouldn't be difficult to add another frequency counter for ohmic sensing, but I don't think my plasma cutter supports that, a Herocut 55X.
More details are here, including all schematics and source code.
github.com/MetalMusings/MyOwnEtherCATDevice
I haven't been cutting a whole lot with it yet. The worry has been the frequency counter that had some interference, seems to have been due to poor cable choice. With that sorted, voltage reading is now stable and THC works just fine.

Hello everyone,
As it's my first message here, let me first introduce myself.
Feel free to skip this part

So... I'm a 40+ years old electronics engineer. I'm currently self-employed, and in long term partnership with a French CNC/3D-printer vendor. I'm using Linux since kernel 2.4.18 (~20+ years), and currently using Linux exclusively.
Most of the time, I develop custom solutions to specific problems, that includes board design, firmware development, computer software, web interfaces... I also worked on few CNC-based 3D printers. (few keywords for the curious: PIC, AVR, C, C++, python, MySQL, php)
I'm quite familiar with Beckhoff and Bosch-Rexroth, as these are the most common in my area (there's Siemens too, but Siemens being Siemens, I just can't...).

======

Read this as a quite general discussion about PLC synchronization, I'm currently working on a project to eject and replace any closed-source Windows-based software for CNCs (i.e. no more TwinCAT and IndraWorks, for runtime and development), and my feelings are telling me that I can easily replace existing PLC tasks (that are synchronized) with some python script (that won't be synchronized). Note, you may also think of C as a replacement for python, it's not the point here.

I'm planning to use python as a base to write PLC programs, but as I see it, it won't be synchronized with servo-thread.
Does anyone has any arguments on the critical conditions that any PLC program running has to be synchronized with servo-thread, or the opposite ?
As far as I'm concerned, I don't see anything that has any real impact of both the servo-thread and PLC task are not synchronized (read it as: I has no side-effect for the servo-task to get an action from the PLC on next cycle), but I may miss some cases.

Ofc, I don't talk about critical procedures like homing or emergency stop. I'm talking about procedures like, let's say T3 M6, this has no obligation to be synchronized, so having a python task that run independently is fine, as long as it has the correct checks and waits, and acknowledge the procedure after everything is done.

I build a small pyvcp side panel and did some gcode coding to build a test program. Not finished yet, but I think it will be handy for testing.

Random values and one can change the limits on the go via the panel.

Sorry for the bad video and picture quality. 
Video

oooh

All pins have GPIO capability (primary function) 
Secondary functions are PWM,Stepgen etc

Does anybody have a post processor that will work with Fusion 360 for LinuxCNC's  GUI QTPlasmac
Thanks

Thanks for looking into things.
I am not sure if I understand correct.
If I type mesaflash --device 5i25 --readhmid I get this



Is I/O port and GPIO refering to the same thing. In the screenshot I/O port 8 says it has no sec. (secondary?) function.


 

linuxcnc.org/docs/stable/html/config/ini...ome_absolute_encoder

if using multiturn absolute servo encoders there isn't any homing, the encoders use batteries to maintain their position and on startup of linuxcnc those encoder positions are written to linuxcnc. this is all setup inside the config files, check the docs for using absolute encoders for the correct ini settings. they are unique and special. with this once started and the encoder positions update the dro's the machine is set in the homed state. linuxcnc does not store the previous positions on shut down as they could be useless on startup even for a non absolute encoder machine which may move slightly when powered off or could be moved while powered off giving an incorrect assumed position to actual machine position.

in the sidebar tab under the WCS tab is a button named limits override but its use is somewhat finicky from my experience. it is just commanding the same functionality available via linuxcnc so the docs will have all of the use case information for how/when to use it.

Thanks for the fast reply, you are right, I imported dragon on the new install to double check and indeed also resets the position to zero. Dragon config is copy/pasted.. so I guess is due to the new install or LCNC 2.9.4? How are stored the encoder positions?

Maybe didn't notice due to edit I had BONUS QUESTIONS : qtdragon allowed me to manual jog outside machine software limits (with home all in red) is this possible with PB? How can I force absolute encoders position to zero? I don't have limit switches on the machine

is your new config correctly setup for absolute encoders? the dro's in probe basic are just displaying the axis data from linuxcnc so if it worked previously in dragon, it should also work the same in probe basic. it sounds like the absolute encoders are not correctly updating linuxcnc at startup. so i would start by double checking your ini settings for homing section maybe?

he needs to update the config files so the user parameter entries are correctly stored in the var file. once he gets his configs sorted the parameter values will be stored persistently and will be pulled into the subroutines correctly. probe basic does tool setting a little different than qtdragon, so there may be some confusion if he tries to interperet G38 statements and translate them into probe basic.

other reasons for probe triggered during a non probe move can be that the hal is incorrectly setup to handle probe and tool setter inputs if one of the circuits is in the triggered state, ie the touch probe unplugged when attempting a tool setter function. but without his configuration setup its difficult to tell.

The distance from the corners to the spline is the exact G64 P[x] value.

Yay!

Hey guys,

I switched to probe basic due to the user friendly interface, previously using dragon GUI.
To avoid losing my actual config I performed a fresh installation on a new SD card ( using Raspi5 )

I'm heading to this weird behavior:

move the machine to X Y Z position -> restart  Probe basic (no servo power off) keeps the position info
move the machine to X Y Z position -> restart  Probe basic (servo power off) X Y Z are back to zero

I guess the cause could be:

1) Probe basic
2) The fresh install is missing something

Raspi5 flashed 2.9.2
Upgraded to LCNC 2.9.4
Installed Ethercat
Installed PB
Merged my actual config with PB

BONUS QUESTIONS : qtdragon allowed me to manual jog outside machine software limits (with home all in red) is this possible with PB? How can I force absolute encoders position to zero? I don't have limit switches on the machine

Thanks in advance for the help

hi, i am studying Linuxcnc and I need to enable coolant on my lathe with an external switch,

my external switch connected to pin as:

an output coolant pin connected as:

how connect  toogle to run flood?
i find this code but understand how it works in linuxcnc


and I find this lesson but don't understand how edit button code to google switch
all my attempts led to errors,

but this code is work good to spindle


and my copy to coolant show error...



And I have a pneumatic spindle chuck and need connect 3 state toggle to open chuck and close chuck,
to open chuсk I receive input but need run 2 step(enable relay(M64 P0) and disable after 3 second(M65 P0))
to close chuсk I receive input and need run 2 step (enable relay(M64 P1) and disable after 3 second(M65 P1))
Maybe this type of chuck is turned on and off differently?
thank you