Advanced Search

Thanks for your discussion. A few minutes ago I got my simple concept breadboard running with adaptive feed so that I can throw a physical switch that is attached to a GPIO pin on the Rpi5 and the trajectory backs up. As is typical, once I got some of the basic ideas of the HAL programming (thanks A METRIC TON to swolebro and his excellent tutorials!), I has able to hack a HAL file together. I now need to think more about the system architecture which relates to your discussion. Am I missing something or does the linuxCNC documentation lack one key element that is critical for the newbie: there are precious few if any examples of HAL command syntax in the documentation...not the case for most languages. I think including a few examples per command would greatly lower the learning curve.

This is normally what motion.spindle-at-speed is used for.
LinuxCNC checks motion.spindle-at-speed  after the
spindle speed has changed in the Gcode and pauses
the first cutting move after the spindle speed change until
motion.spindle-at-speed is true. 

If you have a way to read the spindle speed, usually the
"near" hal component is used to generate the spindle-at-speed
signal. This can also be done if the spindle drive has a signal to report
that is is at speed.

Note that this will not work if you want to stop or alter the feed rate
if the spindle slows during operation, as motion.spindle-at-speed is ignored
after the first cutting move that follows a spindle speed change.

If you need to alter or stop the feed if the spindle speed drops during operation,
you would need to use adaptive feed (motion.adaptive-speed) and some hal math
or logic.