Advanced Search

If i leave G91 and g90
It gives me an error:
Bug: Reached convert_stop() from M99 as subprogram return

Btw I'm on Debian 10 with linuxcnc 2.8.4

After some back and forward i have an analog output. the scale gain is suited for 18k rpm. The rpm limiting needed to be added otherwise an incorrect value, say M3S30000 could be input with 0v output on the pin. Unfortunately this doesn't limit the gui rpm.

# Load required components
loadrt limit2 count=1
loadrt scale count=1
loadrt conv_float_u32 count=1

# Add functions to servo-thread
addf limit2.0 servo-thread
addf scale.0 servo-thread
addf conv-float-u32.0 servo-thread

# Configure RPM limit range
setp limit2.0.min 0
setp limit2.0.max 18000

# Connect motion RPM output to limit2 input
net gui-rpm spindle.0.speed-out => limit2.0.in

# Output from limiter feeds scale
net clamped-rpm limit2.0.out => scale.0.in

# Scale: gain = 1 / 18000 = 0.22
setp scale.0.gain 0.22
setp scale.0.offset 0

# Connect to float→U32 converter
net scaled-float scale.0.out => conv-float-u32.0.in
net final-u32 conv-float-u32.0.out => lcec.0.0.Analog-Output-0

Its a bit tricky to setup a PID speed control with a unipolar analog control
and FWD/REV signals.

I would first try eliminating the feedback by running the spindle open loop
you can do this by setting all PID terms to 0 except FF0 which should be  set to 1.0

If you need feedback (this is rare unless the spindle speed control is very poor)
The structure needs to be like this:

PID command = signed commanded spindle speed in RPM
PID feedback = encoder RPM
PID output --> ABS --> analog out
ABS sign (of PID output) determines the FWD/REV pins

Sorry, I guess the images didn't upload. "hm2_5i24.0.7i83.0.2.analogout0 <= spindle-speed-DAC" reads in forward and reverse. In reverse it shows a reading of 8.25 and stays there, doesn't move when trying to adjust the speed.. In forward it stays around 2.69 with some variation in the next 5 digits. It moves up and down with speed adjustment.

www.welectron.com in Germany has only 7i96S in stock right now
www.eusurplus.com in Portugal has all mentioned above in stock
eusurplus.com/index.php?route=product/product&product_id=64
or
eusurplus.com/index.php?route=product/product&product_id=159
eusurplus.com/index.php?route=product/pr...t_id=105&search=7i85

7I95T looks good. But I couldn't find a cheap German oder european source. There's a German shop that has none in stock. Just contacted them to see when it'll be re-stocked

Will this be a good choice for a hobby machine?
de.aliexpress.com/item/1005005087476488.html

gnipsel.com/linuxcnc/flexgui/index.html

Is the Phoenix the 600x600x600 one with cnc machined parts?
If yes, probably will not be released, i recall they mentioned it was made for a client so they might not approve.
Adding my view here, there are 3D models for Voron with 9 or 10mm wide belts, i built one with those at a 460x460x1200mm size, so no need to wait, just print the gantry and toolhead parts for wide belt and size it up as you see fit.
Resizing a Voron is easy, use their wizard on the website, it spits a complete list of parts with actual sizes.

Post to change the title back to Remora - NVEM / EC300 / EC500.

A 7I95T combines the functions, or a 7I96S+ a 7I85, or 7I85S can do it in two cards

Sure, but I would get rid of the LPT-port requirement and could use any computer, not only this specific one (if broken one day...).

Which Mesa card would be your go-to in 2025?

And is there a card combining outputs (for the stepper drivers) as well as inputs for linear magnetic encoders or would this be a combination of two cards?

N25 G0 X0. Y0. Z25.
N26 M99
O3
N27 G91 G0 Z-23.
N28 G1 Z-12. F300

When i Look at this lines, there is incremental rapid from Z25 to Z2 and feed move incremental to -10
The moves with G3 are all incremental G91, so a pitch of 1, with lead in and out exactly to Z0
Line 44 switch back to abs G90

Sounds like it did not get an Arc OK signal.  

The criteria is different depending what Mode you are using:

linuxcnc.org/docs/devel/html/plasma/qtplasmac.html#plasma:modes

Hi .
I'm trying to output an operation with thread milling but the gcode output for z position is static.
Code:
(THM_DRILL.NGC)
( 2-AUG-2025)
N1 
N2 #12=1
N3 #13=0
N4 
N5 #14=1
N6 #15=0

N7  G21 G49 G0 G54 ( ABS mm clr_Z_offset rapid_motion )

(MSG, Óñòàíîâè T1.
N8 M9 (turn all coolant off)

o106 repeat [#14]
o105 repeat [#12]

S3500 M3
G0 X0. Y0. Z25.
N9 M98 P2

N10 g0 x0 y0
N11 g92 x-#13
N12 o105 endrepeat
N13 g92.1
N14 g90
N15 g92 y-#15
N16 o106 endrepeat
N17 g92.1
N18 g90
N19 M30
O2
N20 (

---

)
N21 (THM-DRILL - THREAD)
N22 (

---

)
N23 G0 X0. Y0. Z25.
N24 M98 P3
N25 G0 X0. Y0. Z25.
N26 M99
O3
N27 G91 G0 Z-23.
N28 G1 Z-12. F300
N29 G41 G1 X7.379 Y-1.516 F800
N30 G3 X0.286 Y8.692 Z0.12 R6.15
N31    Z1. I-7.665 J-7.176
N32    Z1. I-7.665 J-7.176
N33    Z1. I-7.665 J-7.176
N34    Z1. I-7.665 J-7.176
N35    Z1. I-7.665 J-7.176
N36    Z1. I-7.665 J-7.176
N37    Z1. I-7.665 J-7.176
N38    Z1. I-7.665 J-7.176
N39    Z1. I-7.665 J-7.176
N40    X2.835 Y-7.176 Z0.88 R-10.5
N41    X-6.15 Y6.15 Z0.12 R6.15
N42 G40 G1 X-4.35 Y-6.15
N43 G0 Z24.88
N44 G90 M99

The Z axis stays at 1 mm.
I did discover if you open the machine editor of youre machine and go to Working Style => General and you put Software Transform to YES , the gcode output gives the corect z output:
Code:
N16 (

---

)
N17 (THM-DRILL - THREAD)
N18 (

---

)
N19    X0. Y0. Z25.
N20    Z2.
N21 
N22 G1 Z-10. F300
N23 G41 
N24 G1 X7.379 Y-1.516 F800
N25 G3 X7.665 Y7.176 Z-9.88 R6.15
N26    Z-8.88 I-7.665 J-7.176
N27    Z-7.88 I-7.665 J-7.176
N28    Z-6.88 I-7.665 J-7.176
N29    Z-5.88 I-7.665 J-7.176
N30    Z-4.88 I-7.665 J-7.176
N31    Z-3.88 I-7.665 J-7.176
N32    Z-2.88 I-7.665 J-7.176
N33    Z-1.88 I-7.665 J-7.176
N34    Z-0.88 I-7.665 J-7.176
N35    X10.5 Y0. Z0. R-10.5
N36    X4.35 Y6.15 Z0.12 R6.15
N37 G40 
N38 G1 X0. Y0.
N39 
N40 G0 Z25.
The problem is there is no more subroutines, the o call does not exist anymore so is M98 , M99.
The post processor is from here: github.com/G-S-E/SolidCAM-postprocessor-for-LinuxCNC
Thanks

Thats what i got so far