Advanced Search

instructions for that are in the readme

Not sure what the issue is, it works for me:

 

Built a 7I96S bitfile, flashed to 7I96S and ran it with LinuxCNC

( This is on Mint 20.3 = Ubuntu 20.04, Efinity 25.1 patched with the July 24,2025 patch )

deedee wrote:

Hi,

i've got the AX5106 and A5203 up and running. db1981 helped al lot regarding the homing and pos_mode issues.
The project itself is still work in progress, but the main features are working. See attached files. These are the relevant snippets from *.hal: [..]

 

Hi deedee: Are the referenced ax5206.xml and ax5203.xml in the ethercat-config.xml the files attached or are they different ones?

 

started a proper configurator tool in python, early version, if you want to try it clone the test repository.
early sample, most of the functions are not working yet, you can move the components by grab it, make connections by click the pin and click to another pin, and delete the connections using the list in the right side and the "Del conn" button.

I'll try with 2025.1 today

Did you apply the 2025.1 patch?
 

I tried efinity 2025.1.When I try to open the project seveni96s.xml i get the error:

The previous synthesis run has failed. To avoid potential problems or unreliable data, please run it again.

I click O then I get the message:
The family or device maybe invalid. Click OK to proceed with default family and device. Click Cancel to close project and abort.

Then I click OK the next message:
Encounter internal error. Please get help from support team

What does this all mean ? Do I have to choose a family or device first ?
I`m working on xubuntu 24.04

Do you have any suggestions ?

When I press any of the keys X0 YO ZO X this error Unknow command M: M130 appears. What could be the cause?

They all connect to the VFD COM pin
(Spindle-,Out4b,Out5B --> VFD COM)

for the overwrite stuff i prefer encdoders:

www.youtube.com/shorts/jnna9jJnXfs

Modbus speed test:
www.youtube.com/shorts/ZcQjW4829R0

In a rather unfortunate and quite embarrassing turn of events: The actual gearbox section was missing from that drive without anyone noticing. Turns just fine now;).

Yep cool, I am thinking of things like feedrate override, jog speed, rapid override etc nothing "realtime". We have noticed Modbus is slow with general IO, horses for courses I guess.

I've had more electronic parts turn up today so I plan to get busy on the weekend and start testing Riosub over the next few weeks. I've got 4 and 6 axis closed loop controller boards ready to test, both with dual Modbus connections and shift register out, so I can trial multiple combinations... will keep you posted. 

Cheers,
Russell.

you are right, it was quite simple indeed.

i opened the file interfaces file with
$ sudo open /etc/network/interfaces

commented out every line under #wireless

(managed to save the file this time since i was with sudo previlages.. unlike previously where i opened the file manually through the file manager)

$ sudo reboot

now wifi shows up under network manager gui, and it can be turned on and off easy.

thank you!

hello, 
in the nc files you have set the actual x and z positions as starting point after you do a tool change (M6).
In my case it moves to the tool changing position and than the startingpoint is wrong.
I adapted it for myself, but I just wanted to ask if there was a specific reason why the order was chosen that way.
I'm also interested in why the cycle isn't used for thread cutting.

Thanks

Yes one often need some type conversion.
Maybe conv_float_u32? linuxcnc.org/docs/html/hal/components.html

It might be better to use an absolute encoder on the C axis in your case. I also have a 5-axis head/head design with ±185° of rotation, and the unwinding is handled by the postprocessor in CAM.

Looking at your design, did you integrate harmonic drives on these two rotary axes?