Advanced Search

Done!
Thanks.

Spindle mounted at the right side of the X axis, and all that mounted to the Z axis (see HiWin HGH25 left)

How the tool should be positioned during cutting, only the see-thru thin sheet over the brown (orange?) part must be cut, so about 0.5-1mm deep

C axis (3D printed round part with teeth for belt) is mounted above the Y axis and moves with it.

Full view of the important parts:

So far i have tried and tested over 10 configs, and over 40 scripts generated by Gemini, so this has become a late evening sport and an exercise in futility, although there were some moments when it seemed it is working, just to fail miserably.
None of those had TCP, just variations of axis/joints and different planes of works space.
Last two configs have TCP, and the subsequent failure to obey the joint/axis velocities and accelerations, the parallel port one a lot, the Mesa one less often so at least i can test a bit with TCP on and feed/speed lowered beyond useful.
And that brings us to:
Aciera, thank you for helping with this, i did not reply last night as i did not have much time to test, so gave it another go tonight (well yesterday, i see now it is 2:04AM here), so here are some of the results, probably in a non-coherent manner:
-in sim, it starts sine but after a short while it rolls/winds the C axis a lot as can be seen on the preview by the tool making a full rotation around the cut line, but in the same spot on the cut line, then continues nicely for most of the cut.
-on a machine, after fixing the limits and G53 0, it does the same thing, but here LinuxCNC errors out and stops the machine, even at very, very low feed rate and low max velocity, usually C axis at the beginning and Y axis a bit later. I am sure i read somewhere here about TCP causing this but could not find it, so i will try again today/tomorrow.
-the thing with C axis going from approx. 110 to -280 in about 10 lines of gcode does not seem OK, and i could not find what is causing that, it happens on several converted/filtered files, not just one, always after several lines of OK code.
-
Again, thank you, will report back whenever i can.

That's great!


It does need:

setp   hm2_7i97.0.pwmgen.00.offset-mode 1

and not required but an added feature would be:

setp   hm2_7i97.0.pwmgen.00.dither 1


for all 7I97/7I97T PWMgens


Also a better default PWM frequency would be 48 KHz



 

Was this same question asked before on this forum?
-
P.S.
Still did not finish my build of a 4 axis foam cutter, so not yet at this point.

hi I have a 4 axis hot wire foam cutter. i move each axis separate, but I'm wondering if the text below makes sense for, essentially moving U and V with X and Y commands. Thanks

The Fix:
How to make UV emulate XY while keeping their own step calibrations.
To force U and V to mirror X and Y while preserving their independent step calibrations, you must link the signals before the math calculates the motor steps.
In LinuxCNC, you do this by linking the motion controller outputs to both sets of joints, rather than linking the step generators directly.
Open your machine's main .hal file and look for the lines connecting motion.coordinate-system to your joints. Change them to route like this
:hal# Force Joint 2 (U) to listen to the X-axis motion command
net x-output motion.coordinate-system.0.pos-cmd => joint.0.motor-pos-cmd joint.2.motor-pos-cmd

# Force Joint 3 (V) to listen to the Y-axis motion command
net y-output motion.coordinate-system.1.pos-cmd => joint.1.motor-pos-cmd joint.3.motor-pos-cmd
Use code with caution.Why this fixes the calibration issue:
The Signal: LinuxCNC outputs a raw request in millimeters (e.g., "Move X to 50.0mm").The Split: The HAL lines above take that 50.0mm request and send it to both Joint 0 (X) and Joint 2 (U) simultaneously.The Calibration: Joint 0 looks at its own SCALE setting in the .ini file to convert 50.0mm into steps for the X motor. Joint 2 looks at its own independent SCALE setting in the .ini file to convert 50.0mm into steps for the U motor.

I am currently trying to install this and it doesn't validate. When making a bootmedium RUFUS even said, that 1.5GB of the image are missing. Could you please check?

It looks like the 2.9.10 image is half the size of the previous.

The idea/plan is to trace an object (existing part) or a drawing and capture that as a DXF. I don’t have a video of it running but I do have an example from a test run. The image file is what I traced with my machine & the DXF is the result of that trace, there are some issues right now as you can see in the DXF.. The app got "lost" on the smaller radius corner and circle and just started moving back & forth, but the bigger radius and straight line test weren't to bad. I haven’t tried to trace the bottom shape yet, the app isnt ready for that one yet

Will this eventually allow line following while cutting or you are using it to 'trace' an object or drawing? then use that data to make a DXF which then could be used to run the cutter?

Do you have any video of a test?

I remember at my work we had a line following torch and a always thought that would be a great utility to add to linuxcnc's plasma cutter.

I want to flash the rt1052 bin onto a DigitalDream EC300 ethernet controller board. The controller board has an MIMXRT1052 chip. I'm using the STM32 ST-LINK utility running on Windows 11. I've connected the pc to the board via an ST-Link V2 emulator. When I try to connect to the target in the utility, I get the error "Cannot connect to the target!". I've tried the Connect Under Reset and Hot Plug modes to no avail. The EC300 board is powered externally by a 24Vdc supply. I've connected the board to the ST-Link V2 using 3 pins as marked in the picture. What am I doing wrong?

I added the 7i97t code to pncconf in master now.

Onkelmat wrote:

Pnconfig wizard doesn´t work for it. (....)  that config with linuxcnc it doesn´t even find the card and exits with an error that the network isn´t reachable. 
 

I tried pncconf once and only wasted time... But always when I found config similar to my needs I can change those as I want. 
This is Windows Philosophies contra Linux Philosophies...
In Linux you can do anything in terminal. You must know some magic words, but always you can find solution.
In Windows you use mouse and click on icons. You can nothing when no icons for click...
BTW. Read carefully error messages. Not always, but mostly they gives solution.

i believe all amps start with 0x92, then 0x7a, then you can continuously poll it with 0xa2. the answer to 0xa2 is straight-forward:
int16_t abs1 = rxbuf[5] + (rxbuf[6]<<8);
uint32_t cyc1 = rxbuf[2] + (rxbuf[3]<<8) + (rxbuf[4]<<16);
abs1 is full rotations counter, cyc if within-rotation position. afair, it is left-justified to 20 bits, so with your OBA18 encoder 2 LSBs may be always 0.
the last byte of a frame is a xor checksum, not a CRC

Good! I need it too, I just ordered one from Europe to replace my DC servos with Sanyo RS2 ac

There is also:
spindle.N.inhibit - (bit, in) When this bit is TRUE, the spindle speed is set to 0.

Ok if you are using Master branch, I just pushed work to return to the original mode for most things I could find, probing, MDI convince buttons, etc.

Please check it out and let me know what you think or what I missed.

Chris