Advanced Search

The tailstock on the machine works by movving the z axis so it is aligned with the tailstock then a pin gets hydraullically pressed into the tailstock so when you now move the z axis it brings along the tailstock so you can move it to the desired position. Does anybody have a idea how i can make this. I tried to use a m100-199. But then i saw you canttt use subroutines. Otherwise ittt would be perfect since you can use its related parameters to set where you want to place the tailstock. Is it possible to use parameters with unallocatted mcode. So you could maybe do a m20p200?
 

PCW wrote:

I would say a couple hundred KHz so pretty much any encoder interface could work.
Also better if its a square wave so it asks the least of the frequency/period counter.
An offset frequency is good also so you don't lose bandwidth near 0 frequency.

Hi PCW

   If in your opinion square signal 20kHz = 70kHz for 0-10mm distance ( 0.05kHz / 0.01mm resolution ) is good idea ?
And 10kHz for colision.

rebelx wrote:

Thank you so much! I will certainly give it a try. I have been monitoring your github page for quite a while in anticipation of your great work, so this is big news!

As for Clear Linux, I believe they finally got it working on their distro, but made it part their maker-cnc bundle, which includes programs like inkscape and OpenSCAD, which are great apps but should not run on a machine controller. They do not seem to fully understand the LinuxCNC ecosystem, so I have already lowered my expectation (or should I say hopes?).
 

Be aware that my image does not work with axis for some reason due to this bug:

github.com/linuxcnc/linuxcnc/issues/3205

I recommend FlexGUI or gmoccapy until qtpyvcp switches over to pyside6. All GUIs that depend on Qt5 such as QtDragon do not work as Gentoo is dropping support for Qt5. There's been progress with qtpyvcp and pyside6 but is not stable yet.

If you use a run-in-place environment instead of using the gentoo package, axis does in fact work.

Hi cmorley,

I am designing the QT interface from scratch but it will take some time to be able to use my machine in the meantime i wanted to load this vertical qtdragon_hd ui but it didnt work linuxcnc gives an error and close,

i have this folder ../linuxcnc/config/sim.qtdragon_hd.qtdragon_hd_xyz/qtvcp

i made a copy of qtdragon_hd with qtvcp copy command to this folder, so its like this now:

../linuxcnc/config/sim.qtdragon_hd.qtdragon_hd_xyz/qtvcp/screens/qtdragon_hd/

and there is the .ui file etc in there,

i copied the original one to somewhere else and put the vertical ui file there and changed the name to qtdragon_hd.ui

so when i select qtdragon_hd_xyz or _xyza during linuxcnc launch under my configurations it should load the vertical ui right? am i wrong?

The image below is where I'm getting the idea to use Stepgen to convert my fanuc encoder data to AB quadrature for outputting to the drives, but I could use some insight as to whether or not it will function or if I'm missing some important details.. Does anyone know?  



Within the HAL, is there a possibility of connecting the fanuc encoder data position output to the 'position in' port of Stepgen as it runs in quadrature output mode, to then output the quadrature pulses to my AB Ultra3000 drives to provide the position feedback to them? Or is there some major flaw in this concept.. 

Thanks

This machine had a Burny controller so unfortunately no hal scope What I had was my portable scope, that's it.

If you compare my plot to yours, you have 50V/div if I look at it correctly. You can see my capture tops at around 400-450V so that would be already out of your display range? Maybe it is roughly the same?

PCW wrote:

If you measure from the work to signal ground with a GHz capable scope at say 1ns/per division,
You will see fractional ns spikes of significant voltage on the work (or low end of divider)  
( arc starting is instantaneous an even if the current ramp is controlled, the plasma wiring
has significant capacitance)
 

Ok, yeah that can make sense. Its something I can't test so I will take you word for it. My original point was that I thought that having 500V OVP on the divided signal was an exaggeration.

yeah sorry about the typo in the title. should have been 2.10

checked that motion.is-all-homed is TRUE , then had one last prod of the buttons, and the axis does jog on the wheel.
Only if both my plus and minus buttons are pressed together.
With both pressed the scale is .0001

What I really want to do is with those buttons is to duplicate what the rapid plus and minus do on the Axis gui

my1987toyota wrote:

Take care rodw. that kind of weather is no joke. hopefully damage will be minor. Prep for the worst, hope for the best.

+1

Shit mate didn't know it was that bad up there.

We Gunna have to send truck & dogs and bobcats up again to help with the cleanup ?

You could always choose not to run them.
Or you could run them when not machining a part.
Why is it when I hear the word "maker" I envisage some hipster with soft hands, a man bun and a manicured beard ?

I know, cos I'm a cranky old bastard who's one step away from yelling at clouds, I've already got the yelling at tv and birds covered.

Thank you so much! I will certainly give it a try. I have been monitoring your github page for quite a while in anticipation of your great work, so this is big news!

As for Clear Linux, I believe they finally got it working on their distro, but made it part their maker-cnc bundle, which includes programs like inkscape and OpenSCAD, which are great apps but should not run on a machine controller. They do not seem to fully understand the LinuxCNC ecosystem, so I have already lowered my expectation (or should I say hopes?).