Ignore Limits option grayed out

Rods guide is quite detailed (that's good!), so it looks more complicated at first sight than it is. So don't hesitate

HansU wrote:

Rods guide is quite detailed (that's good!), so it looks more complicated than it is at first sight. So don't hesitate

Yeh, thats the problem. As soon as you document something like that step by step it seems to go on forever!

HansU post=270040 userid=27132

Geosec post=270021 userid=33136Ticking the Ignore Limits box allows me to turn the machine back on and move off the limit switch, but the check box is not automatically cleared

However this should be fixed in 2.9. If you find the time and try the 2.9 version I would appreciate your feedback.

Well, I just installed Debian 12 - Bookworm and LinuxCNC 2.9.

I'm not having any trouble with the checkbox clearing automatically.

I am still experiencing the anomaly where the limit switch is ignored when the checkbox is unchecked AND the limit switch is inside the soft limits. I moved the limit switch inside the soft limits just to check this. The initial activation of the limit switch operates properly, shutting off the machine. Checking the ignore limits checkbox allows the machine to be turned back on and the axis moved off the limit switch. The checkbox also clears. Now, with the checkbox unchecked, moving back onto the limit switch has no effect and the axis can be driven into the hard-stop and crashed. Though I didn't actually test crashing it.
I know it's not normal for the limit switch to be inside the soft limits, but I feel that a limit switch should not be ignored unless the ignore checkbox is checked. I'd feel that hard limit switches should have priority over soft limits. But that's just me... YMMV.

Limits should always trigger, no matter what.
Probably something not right in the config.
How many switches per axis/joint?
how are they wired and to what board?

tommylight wrote:

Limits should always trigger, no matter what.
Probably something not right in the config.
How many switches per axis/joint?
how are they wired and to what board?
 

It's a typical 3 axis mill (G0704). Each axis has 3 switches, minimum and maximum limits and a home switch. The "switches" are PNP, normally closed inductive proximity sensors. They are wired to a Mesa 7i76E FPGA card, each to a separate input terminal. They are powered by the same 24VDC Meanwell power supply that powers the Mesa card. I set them up using PncConf Wizard selecting the appropriate selection(s), i.e. => Mesa Card 0 => 7I76 I/O (SS# 0) => [function] Limits => Main Axis => X Maximum Limit... and so on for the minimums and homes.  

I think an important thing to note here is that all of the limit switches operate as they should, as does gmoccapy clearing the ignore limits checkbox, if the limit switches are activated prior to homing.

The Home switches also work properly for homing sequences either individually or home-all.
I only see the problem after the mill is homed and only if the limit switch is inside the soft limits. I duplicated the same issue on two axis Y- and Z+.

Another clue that may be important... I tried to duplicate the problem on the Z+ limit, first by homing only the Z axis. With the Z axis homed and the X & Y not homed, the limit switch operated normally. Even when inside the soft limits. Once ALL axis were homed the problem appeared on the Z+ limit switch, also (with the limit switch inside the soft limits).

When the problem occurs (the limit switch is ignored when the "Ignore Limits" checkbox is unchecked) the Ignore Limits checkbox toggles from greyed-out to active when the limit switch is activated. So, the system is seeing the limit switch, it's just not doing anything about it.

The limit switch should NEVER be inside soft limits. Think of its as an Oh shit, the world is falling in pin. It needs to be outside the soft limits and should never ever be hit unless something has gone terribly wrong!