Advanced Search

I have a Mks dlc32 board connected as described in previous posts but I am not sure how to assign x pulse via i2s channel and not gpio25. Is there a a way to have pulse and direction via i2s? I am using this :

github.com/wezhunter/ESP32_LinuxCNC_MotionController_RealTime

Edit:  I found the answer, the software is not designed to work with pulse assigned to i2s pins. I use the A4988 stepper drivers and wanted to use them without any mods.

  I had seen that as well. To me it's like opening up an all you can eat buffet on  the Titanic while it's sinking.
 

"I think you can use any spindle with an encoder. The spindle doesn't have to be exact."

On the linuxcnc side yes, technically you can spin the spindle by hand and it would work. But to have high dynamics I need a closed loop drive as well because if you use a generic VFD the spindle will have minimal torque at 0 rpm and will take a relatively long time to accelerate. With open loop generally this spindle isn't supposed to run below about 1000rpm.

Basically, I want closer to speedio level tapping than 1986 haas level

thanks

in cia402.comp, to the last row, I added
and compiled again,and all axis are moving now,

thanks to you all

"Jianken has a servo driver they recommend (pre-program) for this type of spindle"

Jianken does not offer a drive for this spindle that can accommodate tapping (encoder interface). The drive they offer is just an open loop vector drive. The company they use (farwide) does offer a servo drive, but I cant get one.

Additionally, that drive can no longer be exported because its over 599hz.

I think I caught a clue.

first photo is taken when the motors working, and second is taken when the motors not,
according to L7N's manual P on first digit is meaning servo is in Position mode,
How can I take the servo to P mode?

Thanks Aciera,
That requirement eluded me when researching this subject.

John

I think you can use any spindle with an encoder. The spindle doesn't have to be exact.
The axis follows the spindle (from encoder readings) so a good/normal Z axis is what you need.

I could repeat that.
It was MaxAcceleration that had a too low value 1e6, now 1e8.
This only matters if one gets gravel sound from the motors.
Setting MaxAcceleration slightly above real max acceleration reduces that significantly.
Otherwise, there is no harm to set it very high. Set too low and it can't accelerate up to speed fast enough.
Pushed a commit, please update the eeprom. Firmware is the same.

I also added an axis_mm.ini config for metric units for testing. 

Jianken has a servo driver they recommend (pre-program) for this type of spindle. Just ask them for it and if they can offer it with an EtherCAT interface.

Subroutines 'may only be called after they are defined'.
linuxcnc.org/docs/html/gcode/o-code.html

For now, I think I'll test it this way.
Later, I can make a support from cut aluminum plates.

I've also ordered the plasma source.
Any other tips and tricks? 

 

 

statusword is 4647, means the bit4 (Voltage enabled) is 0.
therefore your opmode is not set to be 8.

try remove sdoconfig lines in xml. I'm not sure if that's the reason.

the both ways have the same statusword, but their execution timing may not be the same, I'm not sure ^_^

RPI 4B 4GB with Gmoccapy 2.9.4

First efforts at using G10 L2, and subroutines.

The attached code produces two arbitrary short curves.
Simulation only, steppers off.

1) Spent hours chasing down anomalous z value when this code executed.
  It seemed like G10 needed full population of values to avoid ambiguous DRO:
   - G10 L2 P1 x0 y0 z0 R0 to get sensible G54 datum point.
  Is this always the case ?
  Eventually I discovered persistent parameters that had incorrect values.
  Cleared them, and now all good.
  I will ensure in future that the start-up procedure includes a check on parameters.
   
2) My O100 subroutine must be located prior to what I call the "main".
  The main short x20 y20 move checks that the main executed before the subroutine.
  If I position O100 after main, the Gmoccapy editor refuses to save the file.
  This seems odd to me, but I cannot see my error.
  What is the correct way to use subroutines ?
  
John. 

@mBender

You can simply copy the directory and use it, then enter the updates. It's the same as a ‘do not overwrite’ switch.

If you send me your configuration and your edited files, I can see if there is a way to integrate this into the json-configuration.