Advanced Search

Hello! After fiddling with uccnc on my Stepcraft M700 for a while I thought linuxcnc does look good.

Unfortunately I have lots of troubles to get things up and running.

The uccnc config does not look overly complicated:

Or so I thought. Putting all this in pncconf does leave me with a non-working config, the axes just click when trying to move them. It seems that my machine has swapped dir and step pins but the firmware does not allow that. 

After some trial and mostly error I found this:

- If DB25 Pin 2 is left disconnected ("open"), the spindle starts immediately
- The Y-axis Step signal is on Pin 6, not sure about the Dir 
- The Z-axis Step signal is on Pin 8, not sure about the Dir

But I fail to make sense of it or get anything else to work. Did search through for forum for 7i92 or stepcraft machines, no luck so far. Wrong keywords maybe? Added my config, any help is appreciated!

 

The hal_gpio driver will only work with the bookworm images and not with the recent Trixie images due to incompatibility issues with libgpiod2 supplied with bookworm and libgpiod3 supplied with Trixie.
I raised this issue on GitHub and it is on the to do list.
So what it means is that for gpios builds use the images dated around Jan 2025 or earlier.

Preop is fine, preop E is not.
General debug tool is to look at the last 10-20 or so lines from "sudo dmesg". Ethercat master always says there what is the problem. Not always easy to decode, still it gives some clues.
Many many times the eeprom and firmware have come out of sync. Like added something with eeprom_generator and didn't upload eeprom.
Also, does the counter update in Twincat?
If you get stuck and it won't work, please push the updated code to github and I can have a look.

Y1 and Y2 are not reading any index for me. I better bring my oscilloscope next time to check what, if anything, is coming in those z and /z wires from the linear scales. Latching hal logic connected to [HMOT](CARD0).encoder.04.input-index and [HMOT](CARD0).encoder.05.input-index does not catch any pulse.

Y1 and Y2 also has reference sensors near the bottom, it'll never reach that position after he installs the punch and the die. I see the book calls them Max Down Left, and Max Down Right. So they are bottom limits. Maybe to avoid overextending the cylinders when no tooling is installed.

edit:
Finally got R1 axis (joint 6) homing to index.
When all else fails, read the instructions! Here is the magic that makes home-to-index work:
net    j6-index-enable        [HMOT](CARD0).encoder.00.index-enable
net    j6-index-enable        joint.6.index-enable

No RPI3 image, for RPI4 and RPI5 see here
www.linuxcnc.org/iso/

the ALM signal is just an OPTO output

The ALM is an OPTO output from the CL57T right? According to the manual section 6, it shows that I need to supply 5-24v DC to that circuit. Can I just use my 24v field power for this? Doesn't it matter that I have the driver setting S3 set to 5v (due to STEP/DIR 5v signals)? I don't want to fry this thing before I get going.

Note that you could parallel the ALM outputs to save inputs

This was the plan.Also, how do I use a spare STEP/DIR for ENA? Is that in my HAL settings? The manual mentions:

ENA must be ahead of DIR by at least 200ms

Hello everybody,
I do have a self build portal CNC machine with LinuxCNC and LPT running. Never really disappointed me, well maybe it could be a little less noisy (more microsteps). Now a WABECO F1210 found its way into my workshop. I would like to convert the whole system to LinuxCNC since I got the whole "chain" for this running. When I started LPT was the choice of the time. Now I read about MESA and stuff. ChatGPT brought me to the PicoBOB-DLX. This thing looks quite like a smart "simple" solution. Does someone has this running? Since I got some kids since my last build the "free" time did not get more... So I am really interested in a quite simple solution.
regards Han

PNP for mixed voltages...

But I don't think you really need mixed input voltages, as the Enable is an output
and the ALM signal is just an OPTO output

For 24V NPN proximity switches:

+24V --> 7I96S input common

NPN proximity switch 0utputs --> 7I96S isolated inputs

ALM+  --> 7i96S isolated input

COM- 24V COMMOM

Note that you could parallel the ALM outputs to save inputs

PCW wrote:

You can mix  7I96/7I96S input voltages if the inputs are wired in sinking mode,
that is if the input common (TB3 pin 12)  is wired to ground and the sources 
(switches, alarm signals etc) supply a positive voltage relative to this ground,
in this case, the input will respond to positive voltages between 5V and 36V

so
Home/limit switch:

+24V  --> SWITCH COM  --> SWITCH NC --> 7I96S isolated input

 5V alarm signal:

 +5V alarm signal --> 7I96S isolated input

7I96/7i96S input common --> 24V GND and 5V GND

Note the the enable signal is a 7I96S output so can either be driven from a free step/dir output
or an isolated output
 

Yeah I realized what I said about ENA after I posted. Thanks for the clarification. So, should my prox sensors be NPN or PNP in this case? Crossing fingers for NPN.

You can mix  7I96/7I96S input voltages if the inputs are wired in sinking mode,
that is if the input common (TB3 pin 12)  is wired to ground and the sources 
(switches, alarm signals etc) supply a positive voltage relative to this ground,
in this case, the input will respond to positive voltages between 5V and 36V

so
Home/limit switch:

+24V  --> SWITCH COM  --> SWITCH NC --> 7I96S isolated input

 5V alarm signal:

 +5V alarm signal --> 7I96S isolated input

7I96/7i96S input common --> 24V GND and 5V GND

Note the the enable signal is a 7I96S output so can either be driven from a free step/dir output
or an isolated output
 

PCW wrote:

I would use all 5V signals for the step motor drive interface 
and possibly 24V for limit switches .
 

How does one go about mixing 24v limit switch inputs with the 5v ENA/ALM inputs?

what is kernel and what is compiler, how use or how install and use. what is repos and scratch. not understand code nerd super engineer language. (translator not help this)

but were is rpi 3 and 4 linuxcnc os.

It appears to work fine. Any further suggestions would be appreciated.

Full HAL file attached.
 

And here is the Snippet:
# --- SPINDLE PWM CONFIGURATION ---

# Load and Add Functions
# (Ensure loadrt pwmgen and loadrt scale are at the top of your file)
addf pwmgen.make-pulses base-thread
addf pwmgen.update servo-thread

# 3. Connect Spindle Enable
# This turns the PWM generator on only when the spindle is active and needs to be eliminated.
# (removed) net spindle-on spindle.0.on => pwmgen.0.enable

#Inverted Voltage Corrections
# First you need to add the next two lines above
# loadrt not count=1
# addf not.0 servo-thread

# set the Frequency
# first we need to set a frequency for the PWM in order for it to smooth the output
setp pwmgen.0.pwm-freq 100.0

# Calibration
# When using a 0-10v output signal, you may need to invert the PWM output
# so that the 5 axis breakout board will output 0V at spindle off,and
# 10v when the spindle is full on. The Negative sign before the 60000 usually does this.
# unfortunately the resting (spindle off) voltage is 10.0V - bad -very bad!
# rather we leave everything alone and creat a not.0 variable that reverses the values
setp pwmgen.0.scale 60000
setp pwmgen.0.offset 0.0
setp pwmgen.0.enable 1

# Connect Spindle Speed Command
# This signal carries the RPM requested by G-code (S command)
net spindle-speed-cmd spindle.0.speed-out => pwmgen.0.value

# Prepare to invert it
net spindle-pwm-raw pwmgen.0.pwm => not.0.in

# Connect and invert the PWM Output to Physical GPIO
# This is the "Writer" for GPIO18
net spindle-pwm-final not.0.out => hal_gpio.GPIO18-out

#End Spindle Code