Limit Switch Troubles - HAL Debounce

This arc tolerance issue is fixed in the just-released 2.7.0~pre4.

Hi, does this work for limit switches too? eg all-limit? Would it be:

net debounce-all-limit debounce.0.0.in <= parport.0.pin-10-in
net all-home debounce.0.0.out

Does it have to go in the main .hal file or can you just put it in the custom.hal file?

Is there documentation were all the home-x etc are listed and defined?
Thanks.

It should work.

It can go in any HAL file, but ina secondary HAL file you might need to unlink some nets made in the primary HAL file.

There is no documentation for home-x etc, and never can be, as the signal names are entirely arbitrary and can be freely chosen. It is the HAL pin names that matter.

The HAL pins of all components are described here:
linuxcnc.org/docs/2.7/html/
Click on "expand man pages" and pay special attention to "motion" "iocontrol" and "halui"

It seems we may have this issue and I would like to investigate possibly increasing the voltage on our NC switches. As our breakout board inputs are 5V we would need to use something like opto couplers at the breakout board end. Is this what you did to increase voltage?
As we now think individual home switches may be the way to go since we have no other use for three of the five input pins would increasing the voltage for a single switch and homing on that axis be a suitable diagnostic test?

Although we still have joint error issues we did alter the wiring to put home and limit switches for individual axes on separate pins. After talking to the BOB designer we have installed 0.1 microfarad capacitors on those input pins which has eliminated some of the interference and made the homing action smooth and reasonably fast with no random stopping without error messages (as occurred before).
While I don't know if using HAL debounce will have any effect on our errors I will follow some of the recommendations in this thread and see if it helps.
Our current problems are occasional and apparently random J0 errors including when the machine is motionless plus more persistent J2 errors . The Y axis is free of errors itself although, of course, affected by the other axis errors.

The HAL debounce can be used to provide different levels of filtering for the same signal.
I typically use a 1-2mS filtered version for homing, and a 10mS version for limit switches.
The machine doesn’t travel _that_ far in 10mS compared to the inconvenience of nuisance tripping.

I would suggest removing the hardware filter that you added.