Advanced Search

Thats some really good points about the gantry.
Appreciate your help.

Which driver would i need for this? Just the standard cia402?

I dont think, that is possible.

If you wire the switches to the servo drive, you can use one of the drives internal homing methods. They allow a combination of home switch and encoder zero position. That will give you very high repeatability.

If you dont wire the switches to the servo drive, you are forced to the linuxcnc internal homing. Same with limit switches. This will reduce your repeatability to the repeatability of the home switch (you will need high precision home switches).

However if you use the servos internal homing in a gantry setup, you have to wire both gantry drives to the same switches. Servo based homing does not know about the mechanics of your gantry. You do not want two servos fighting each other and bending something.
Also maybe limit the torque during homing (PDO 6072)

My home switches are wired to the IO Modules. I only have one switch per joint
If i trigger the switches manually they getting recognised correctly in the hal show.
Do you know by a chance if its also possible the other way around?
So telling the driver like pin xy was triggered over the ethercat way?

My brain blockade right now is the following:
I have home switches configured like i did with steppers before. When triggered, shouldnt LCNC tell the Servo stop moving? Why do i need to home hooked to the drives.

If you need startup jogging and you need to tune the servo ... you need to turn off Lag monitor under the  Axis X.. Parameter ... Monitoring: Position Lag Monitor = False ...

If you have reduction factor ... you will determine it in the nominator and denominator rows somewhere in the drive manager ...

Yes one motor revolution == one encoder revolution ..

Yes HD should reverse you motion .. every odd 

This setting you generally free enable movement... next tune or you will need firmware update to start autotunnig the servo system ..

How did you wire your home and limit switches?
If you push the siwtches by hand, do you see the "pins" change in halshow?

I would recommend to wire the limit and home switches to the servo drives DI inputs.
Then you can see the signals in PDO 60FD.

e.g. add this to your ethercat-conf.xml for each A6 servo:

<pdoEntry idx="60FD" subIdx="00" bitLen="32" halType="complex">
<complexEntry bitLen="1" halPin="limit-neg" halType="bit"/>
<complexEntry bitLen="1" halPin="limit-pos" halType="bit"/>
<complexEntry bitLen="1" halPin="home-switch" halType="bit"/>
<complexEntry bitLen="13" />
<complexEntry bitLen="1" halPin="din1" halType="bit"/>
<complexEntry bitLen="1" halPin="din2" halType="bit"/>
<complexEntry bitLen="1" halPin="din3" halType="bit"/>
<complexEntry bitLen="1" halPin="din4" halType="bit"/>
<complexEntry bitLen="1" halPin="din5" halType="bit"/>
<complexEntry bitLen="1" halPin="din6" halType="bit"/>
<complexEntry bitLen="1" halPin="din7" halType="bit"/>
<complexEntry bitLen="1" halPin="din8" halType="bit"/>
<complexEntry bitLen="1" halPin="din9" halType="bit"/>
<complexEntry bitLen="7" />
</pdoEntry>


I use negative logic. You have to configure this on your drive to match the limit/home switches. This is permanently saved in the drive.
e.g.
# set operating mode
ethercat upload -p 2 --type int8 0x6060 0x00
ethercat download -p 2 --type int8 0x6060 0x00 8

# set "active low" DI input logic
ethercat upload -p 2 --type uint16 0x2004 0x02
ethercat download -p 2 --type uint16 0x2004 0x02 1

ethercat upload -p 2 --type uint16 0x2004 0x06
ethercat download -p 2 --type uint16 0x2004 0x06 1

ethercat upload -p 2 --type uint16 0x2004 0x0A
ethercat download -p 2 --type uint16 0x2004 0x0A 1

You also have to set homing mode  (PDO 0x6098) - see manual 

Have a look at monokrom. it might do this already as the designer did the ramp piercing and pushed to qtp. It has a proper SQL database

as an aside, how do i mark this as answered ??

Thanks that fixed it. Unsure what exactly happened, but happy to have it fixed

 

le_potato wrote:

What about removing the package from the image file?

The installer supresses it and its not installed but as soon as there is an apt update, it comes back. Very annoying. This was present in the original release of Debian Bookworm and it was fixed in the first update but the update were never pushed to the Debian Builder we use to make the iso (which is what we should use to follow the Deban way)

Hey there,
im hare to look for some help on getting my Servos to work. Which currently ignore my endstops and would "crash" into them if i wouldnt stop
The Servos are Stepperonline A6 400W Ethercat ones, combined with an Beckhoff EK1100 and some IO Modules.
I also got the absolute encoder cables, but so far thought to just use them like a "stepper drop-in" and then if working, to go the next step and do the "integrated homing" way.

I just setup everything i think i will need:
- installing the Ethercat
- installing the cia402 driver

So hopefully theres people here who might be able to "get me on the right way" with my Servo setup and explain me my mistakes.
PrintNC_LinuxCNC/EtherCAT at main · Luc1luc/PrintNC_LinuxCNC

2.9 pre is years old and buggy. You can't depend on it. You really should upgrade to 2.9.4 by following the steps in the Getting Linuxcnc docs or adopt Debian Trixie pre-release as it should be out as the official Debian release next month.

I've been busy on this machine. Got my B axis alive.
I will throw some pictures up of the other progress tomorrow.

youtube.com/shorts/gAclUIYUk1Y?si=PEWtBkNokJtn4JyQ

In other news, my new pc build runs much better generally and I don't have the real-time delay error warnings anymore. And I've been busy making mechanical progress on my machine.  

youtube.com/shorts/gAclUIYUk1Y?si=PEWtBkNokJtn4JyQ

I've been messing with this again. I have an intel 2 ports pcie lan card that I tried to use. I got the ip address from ip a command and put it in the etc/ethercat.config. for some reason only the lan on the motherboard works even with the address changed. But the servos were crunchy sounding and it was disconnecting randomly. I put it back to the correct address for the motherboard port and its smooth. I have not been able to run ethercats on the pci card.

Got the saw back home. She's definitely rough. Good news is it looks possible to make the saw miter in the opposite direction by removing one support bar. The way the shuttle seems to work is it has a ball screw with a limit switch on it to control the length that the hydraulic retracts. Once it hits the limit switch it stops the hydraulic from further feeding forward. I may rework the vices as they currently have a very shallow depth of hydraulic travel. The bearings in the saw wheels feel sloppy but smooth. In the morning I am pressure washing the whole machine and may try to put power to it to see if anything still works. If I can reuse hydraulic pump and band saw motor that would be awesome!