ESCape key issues in Pre1-4587 build ???

Mpfan should be close enough for Linux cnc.

andypugh wrote: Before we go off on weird tangents, is anyone else running this version of LinuxCNC and can check that ESC only stops the program and does not turn the machine off?

Sorry but i am using a version with reverse run and external offsets testing the PhillC's configuration for plasma and THC. It just says Pre1 !!! Is there any other way i can find out what build it is?

I just tried an experiment with an up-to-date build of LinuxCNC and can not reproduce the problem.

So, it seems that it might be specific to your setup.

Can you try running one of the sim configs (I used sim/axis/axis_mm) and see if you also see the problem there?

LinuxCNC emulator works perfectly. Does not knock machine off.

It is not a Linuxcnc Issue or build issue.
It is an induced electrical interference to the parallel port that will do that to a computer.
Can you disconnect the spindle and run the machine ?
There is another possibility, but that happens rarely. A failing power supply can do that when stopping the motors at once, from the back emf.

see this post
[strike]I can confirm that similar behaviour exists on configs run with my experimental plasmac component .
It is identical behaviour using either Axis, GladeVCP, Gmoccapy or QtVCP.
Up until a M3 command, ESC, F1 & F2 behave normally.
As soon as a M3 is issued then none of the keys work.
After issuing a STOP from a button then all the keys are active.
If I PAUSE the program then all the keys are active.
If I comment out the M3 and run the program then everything behaves as it should.

This seems to point to something in the component stopping keypresses getting to the GUI.

Any ideas on what could cause this???

Cheers, Phill

Edit 1: The halshow window is while running a normal program[/strike]


[strike]Edit 2: To further confuse the issue, if I let a program run to the end the keys are still inactive, but... if I then wait a couple of minutes they then become active again.[/strike]

Spindle weren't running during the failure.

I had also NOT applied power to the Spindle Controller .. or pumps.. So no power there..

It would seem very odd to me, though possible.. though highly improbably.. that "induced emf" or whatever, wouldn't show up under normal, adverse running conditions.. which it has never, ever shut off on it's own.

So, for a singular event to be so consistent.. it is interesting. But, since I do have the tools to check for that particular type of failure, I will attempt to do so. I will be down for a few weeks, so I won't be able to do any serious testing for a while. Thanks for the guides....

phillc54 wrote: I can confirm that similar behaviour exists on configs run with my experimental plasmac component ..

Well, that's interesting!

Which LinuxCNC version?

andypugh wrote: Well, that's interesting!

Which LinuxCNC version?

see this post
[strike]It is up to date (28th Feb) master with reverse-run as well.
I will try later today with a vanilla master version.

Cheers, Phill.

Edit: I have just tried with 2.8.0~pre1.4625.g8f279b835 which I just upgraded to from the buildbot (stretch master-rtpreempt) and it still showing the above behaviour.
I probably should add that all this is being run on a VMware Fusion virtual machine.[/strike]

I am using the same version as Phill, but i can not reproduce that behaviour ! When i hit ESC everything works properly.