Advanced Search

tommylight wrote:

I have no power for the last 3 days so using the phone makes it dificult to quote stuff with this editor, but in short:
-i have PlasmaC in daily use in over 10 machines, users got used to occasional stops and are OK with it, luckily rareley happens, i think it is 2.8.4 version.

Tommy, the last commit to the PlasmaC component for 2.8 (that was not just a version rev) was December 15, 2020. I'm not sure why you expect anything from that branch that may be a bug to get fixed. I can tell you for sure that I have 0 interest in squishing bugs in 2.8.

Certainly many improvements have been made to PlasmaC in 2.9. That said, even from a Qt/PlasmaC perspective, we haven't pushed any changes to 2.9 since ~August 21, 2024. But it's worth noting that component hasn't seen a ton of changes in 2.10.

-PlasmaC has 4 or 5 ways of disabling the THC, most are useless, except VAD, and void lock sometimes. Right now i can not think of any other valid use case.

From the 2.10 component code:
Main GUI disable
Disable(and enable) via G-Code
Velocity Anti Dive (sometimes called Corner Lock)
Void Lock
Mesh Mode/Ignore Arc OK

Personally don't see any of those as useless, but you also aren't required to use any of it.

I will tell you that I am the reason for Mesh Mode/Ignore Arc OK, because it would be impossible to cut expanded metal otherwise, something my Hypertherm is very capable of, and I do use this feature from time to time.

BTW default should be at 60-70%, not at 95%.

For QtPlasmaC, Cornerlock/VAD is defaulted to 90%. It's on the PARAMETERS tab as "VAD Threshold", luckily you change it once, click "SAVE" and it stays as your new default . It's literally impossible to make a default that works for everyone. If I change it to 65%, there will be other users telling me it should be 90%. That's why it's tunable.

-what should i monitor for the issue where the torch keeps going down till float switch triggers during cut

If you are talking 2.9 or 2.10, then I would say without knowing what is causing it, I would just be guessing. But, I would want to know, is the voltage staying constant, and the plasmac component is commanding the THC down anyways? Is my voltage erratic? Etc.

As a start, I'd plot plasmac.arc-voltage-in with plasmac.move-down, plasmac.move-up, and plasmac.current-velocity. If you're telling the component what voltage you want it to shoot for, then I'd add plasmac.cut-volts. If you're using auto-volts then I'd want plasmac.target-volts.

happens often and ruins the nozzle. More often in QtPlasmaC but that might be just luck

Phill added plasmac.low-cut-volts as an attempt to prevent this. You could try setting that to prevent it form diving into the material.

-what should i monitor for the issue where the machine moves to a cut position and just stays there? This happens sometimes even when doing a manual cut where pressing F9 shows the machine running on screen but nothing moves. DRO does not change, machine state does. Again this happens a lot on QtPlasmaC, do not recall this happening on PlasmaC so might be a QT thing.

Again, I'd want to know if motion.eoffset-limited goes true when this occurs. I am also curious what state it's stuck in, so I would want you to turn on plasmac.debug-print before starting any cutting, etc. and copy and paste the terminal output from when it hung up.

-i think i have an old config for QtPlasmaC on a laptop for my machine, i can upload that if you think it might help.
-i absolutely have no issues with intelectual property, i do and will post everything i do personally, as can be seen from many posts here, i am strictly Open Source everything, and i did reject any job requiring me to sign an NDA
Ooops, not very short

If you have it, I'd look at it. Sure.

Thanks, @Surmetall. I took a closer look at your solution and it looks very clean.The main issue on my side is that it appears MAHO did not use the same motor construction as Deckel.



I’ve been trying to find maintenance or service manuals to see whether the front bearing housing of the original motors could be reused, but so far I haven’t had any luck.

Another complication is the very limited space available for the Z-axis, which makes motor placement quite challenging. I’m planning to make a short video soon to better illustrate the space constraints and the problem I’m running into.

Happy New Year!

github.com/grandixximo/linuxcnc

I've committed a working multi axis s-curve optional trajectory planner in this fork, happy to get some feedback from the community, it is based on master.

Here is my s-curve, go wild...

github.com/grandixximo/linuxcnc

I have not build it yet, but it should work, test and let me know,

happy new year!

Yes, if you have good latency it does not really matter how close you are to the sync0shift window, as long as you never jump ahead and back, I also had to force the issue to come up and induce latency in my system by opening graphical intensive applications. But I believe my fix is a real game changer for running ethercat on something that is not super latency optimized, and not see the grinding issue, I hope the community will receive it well, I'm here for any bug fixes or whatnot...

github.com/grandixximo/linuxcnc

I have not build it yet, but it should work, test and let me know

Edit:
There were build errors, should all be fixed, built on debian 13 no issues, have a simple xyz test configuration, it works with more joints/axis no problem, xyz just quick and simple.

Quoting from the Xfce tour:
"Screen Locking
There is no dedicated "security" tab anymore. Lock screen management was massively simplified and "Light Locker" was dropped. "lock-on-sleep" is now synchronized with xfce4-session and xfce4-screensaver.

Screen locking settings are now only handled by xfce4-screensaver to avoid conflicts. A button to open xfce4-screensaver-preferences was added"

That gave me the idea "Power Manager" does nothing anymore. So I installed xfce4-screensaver and set the screensaver to blank and lock after one minute. I watched it for 6 minutes and it stayed lit and didn't lock.

Next I set "Power Manager" from "never" to blank and lock after 1 minute. It blanked and locked after 1 minute. I set it back to 'never'.

We'll see how it behaves over the next couple days, but it appears installing xfce4-screensaver allowed my power manager settings to work but the screensaver settings didn't work. Go figure.

I do get the idea xfce4 is experiencing a major rewrite and expect over time there will be updates that make it behave.

Edit: It appears this is the associated bug report:
bugs.debian.org/cgi-bin/bugreport.cgi?bug=1034607

Ok it is officially January 1 2026 in my area and fusion 360 is still working on my Windows 10 machine.
Will it stay that way through out 2026? Don't know. However given the bad press of
Windows 11 Autodesk would do well to let fusion 360 run on Win 10 for a while longer.
So we will see.
With that said Have a Happy New year every one.

Disabled screen lock in power manager, also sleep = never. This is on Debian Trixie  with XFCE 4.20 on x11 without LinuxCNC. I don't touch the mouse and keyboard for 5 minutes and it blanks and locks. Totally locks me out if I was accessing it via vnc, it has to be locally unlocked before vnc can reconnect to the display but ssh still works. 

I don't have screensaver installed at all. I installed screensaver on a friend's Trixie machine and screensaver's sleep settings seemed to work as advertised. My previous Bookworm install didn't have screensaver either but screen lock settings worked. I read something on Debian forums about more people with similar problems. I think it was here:
forums.debian.net/viewtopic.php?p=814969
xfce.org/about/tour

Enabled user autologin in /etc/lightdm/lightdm.conf so it logs me right to the desktop upon boot. That didn't seem to change anything with the screen blanking problem.

I can put up with it on the workstation for now, it would be a real problem on a cnc machine.

Even so, XFCE is my favorite desktop. It is similar to what Gnome was the first time I used Linux. Some years ago Gnome changed and completely lost my interest.

Thllll wrote:

Mr.Rodw big fan of your work! I wouldnt disagree with that however if there is solution to continue to use ER16 tooling (I should have mentioned that in the og post) then I would rather do that then invest in a TTS spindle and TTS collets. I know there is a solution to this

Thanks. I think the framework is there in qtdragon aleady but be prepared to read the docs and play. Conceptually not that hard. probe towards surface, then probe away until contact breaks then set Z0. You can also do it manually until you can roll a cylinder (eg drill or milling cutter) until it just fits under than set Z10 (if it is a 10mm dia cylinder).

Sounds like a good way end the day / year to me rodw. I am about to break the seal on a bottle of
a Single Malt Irish Whiskey.

Forgot to mention ( in case of incoherency) that I am working my way through the last of a bottle of Canadian Whiskey before I fly home tomorrow 

Well they could charge your credit card like they did for me! Took a bit to get that reversed...
Also cancelled my adobe subscription.

Unfortunately, I have to boot Windows to run the Anycubic Slicer for my 3d printer so I am not shuffling gcode and USB drives.
There was a time I knew Windows backwards, now I am a bit bamboozled by it!

Same thing happening to me, nearly the same setup too, did you ever find a solution?