Advanced Search

Oops, just thought. the tension should be able to be done using home offsets

Once the fabric is tensioned, linuxcnc should be able to use the joint axes to control 4 motors in unison. Normally, it is just used for 2 on both sides of a gantry but I know its Been used on a 4 joint Z axis where all 4 corners are lifted..

I think you should use halcompile because this also lets you document your component via a man entry and the web site in the same file.
For new contributions, you should be using 2.10 (master) branch as 2.9 is frozen.
Using C would just add extra work. Not sure on user space comps though. I have never written one.

Both is fine.
The .comp format is more compact and easier to understand and maintain particularly for people not totally familiar with pure c.

Note that the component should remain compatible with older configurations.

I would think that your biggest issue is going to be the stretching of the fabric and that partly cut sections are going to get bunched up as you move the material back and forth.

I have picked up 4 new motor drivers and ac servos. I am trying to understand how I can use these with LinuxCNC..
I've been searching but can't even figure out how to start..  Has anyone used the Omron G5 hardware?

Thanks for any help.

Personally I would add the compensation to the kinematics model but if you feel more comfortable using a hal solution then why not give that a try.

There is also the external offsets feature:
www.linuxcnc.org/docs/devel/html/motion/external-offsets.html

One easy way is to use a mux8:
linuxcnc.org/docs/2.5/html/man/man9/mux8.9.html

This way you can net your three spindle signals to the three selections of the mux and you can define your modbus values to the corresponding inputs of the mux.

I am using mb2hal (modbus) and it exposes a pin `mb2hal.00.00.int` that is an input, the values can be 1, 2, or 5. I need to set the values based on other bit pin:

if spindle.0.forward = true then mb2hal.00.00.in = 1
if spindle.0.reverse = true then mb2hal.00.00.in = 2
if spindle.0.on = false then mb2hal.00.00.in = 5

How do I create this logic in the hal?

Cool idea. I also think this shoul be possible.

I am thinking of servos instead of steppers. Reason is build in encoders (ok cl steppers have it too) and the possibility to change to torque mode. So the two pulling servos in position mode doing the move and the two servos in torque mode can build up adjustable tension.

This setup than could be used as an initial meassure of the layer thickness and correct values for that.

I can imagine some kind of tape at the sides for all corrections. Like the tape on the shaper woodworking cnc handheld machine (dont know the exact name right now). This way you can use a camera like system an be on track all the time I guess.

Cool, just ping me in here when you have time and need help getting things setup!

Most probably, yes.
Stuff helpful to know:
-what is the maximum Y length in one go?
Stuff randomly wandering through my tired brain:
-would be easier to use ESP-CAM for side monitoring and adjustment
-encoders can be tracking the material on both sides as LinuxCNC can use 2 or more encoder feedback's for the same joint/axis
-see "bottle labeling machine" for the type of sensor used for aligning the labeling foil (stuck my finger between the cutting and gluing rollers on one of those 2 weeks ago, what a mess! )

Funny you mention that, as that is on my to-do list. It's been a little bit of a challenge. Once I have the mill 90% I'll try to read my notes and get some docs done. I also need to contribute code too. I also get that there has been less EtherCAT used in the past, which may change as the cost comes down. I was able to retrofit with 1.2KW AC servos for a little more cost than keeping 30-plus year old steppers, with new drivers, power supplies, and an interface card.

Glad you got it sorted. The docs are only giving probe basic specific information, the assumption on configuration building is that the main lcnc docs are used to obtain a working config. In that regard the connections for the probe basic UI dialog are what's being connected here in lieu of lcnc stock dialogs. This gives probe basic control of tool change ques. with the limited number of people working on probe basic, typically 2 or fewer at any given time it's too much to cover what's already available in linuxcnc. This is why we express the importance of qualifying a working configuration in axis before beginning the transition to probe basic to eliminate any general linuxcnc configuration problems without the complexity being compounded by the probe basic requirements.

For your situation it's 100% uncharted territory because none of the devs use Ethercat. If you would like to contribute a docs section for Ethercat users it would be greatly appreciated. The sphinx docs source files in the repository are where the doc pages reside and you can write up the specifics needed and format in the .rst format and submit a pull request and I will add them to the main docs page. I'm sure future Ethercat users would be very grateful for some guidance on how to avoid pitfalls the standard config docs can cause.

Let me know if you need assistance getting started.

Thanks, I was able to figure it out.

I did read the words.  In fact I reread, prior to posting here.   To be honest, all the documentation is light, I would rather follow the docs than be a pia, yet we all are humans and make mistakes.
If you were asking if I removed all the hal_manualtoolchage, yes there is no hal_manualtoolchange in any of my hal files.

The issue is that the other tool-change nets were not defined. So while I followed the words and removed the red, it also implied to add the other nets.  I am using EtherCAT so my machine hal was not built from a pnconfig, etc.

For anyone in the future, these nets need to be in your machine hal.