Parallel pin state on exit, stuck ON? Sometimes?
17 Sep 2013 04:04 #38905
by TroyO
Parallel pin state on exit, stuck ON? Sometimes? was created by TroyO
I am a newb to LunixCNC and cnc in general I have been working on getting a system set up with steppers and drivers and have run in to a snag I just don't understand.
I am using a Vactrol device hooked to a DC motor controller (KBIC 125) to control spindle speed. So far, so good... I can indeed control the spindle speed. I also have a relay in place to control the AC power to the controller. Again, this seems to work well.
So, that's the machine setup but here's the rub.... when I exit LinuxCNC it *sometimes* leaves the parallel port pins HIGH for the relay and the PWM when I exit but haven't otherwise E-stopped or shut the spindle down first. It seems to do this more when the spindle speed (PDM Mode) is set over halfway up. Probably because it's ON more than OFF at that setting. I verified that it stops switching the output and it's just stuck ON.
Obviously I don't want the machine running rampant and replicating itself like Skynet! It's also probably not a great idea to leave a spindle running full blast. While I'm sure exiting without having first "shut down" the CNC operations also isn't a great idea it just seems like it should "Fail off" if anything.
LinuxCNC 2.5.1 by the way.
Is there a way to set the "Exit state" of the ports/pins or should that behavior not be happening at all?
I am using a Vactrol device hooked to a DC motor controller (KBIC 125) to control spindle speed. So far, so good... I can indeed control the spindle speed. I also have a relay in place to control the AC power to the controller. Again, this seems to work well.
So, that's the machine setup but here's the rub.... when I exit LinuxCNC it *sometimes* leaves the parallel port pins HIGH for the relay and the PWM when I exit but haven't otherwise E-stopped or shut the spindle down first. It seems to do this more when the spindle speed (PDM Mode) is set over halfway up. Probably because it's ON more than OFF at that setting. I verified that it stops switching the output and it's just stuck ON.
Obviously I don't want the machine running rampant and replicating itself like Skynet! It's also probably not a great idea to leave a spindle running full blast. While I'm sure exiting without having first "shut down" the CNC operations also isn't a great idea it just seems like it should "Fail off" if anything.
LinuxCNC 2.5.1 by the way.
Is there a way to set the "Exit state" of the ports/pins or should that behavior not be happening at all?
Please Log in or Create an account to join the conversation.
17 Sep 2013 15:07 #38927
by ArcEye
Replied by ArcEye on topic Parallel pin state on exit, stuck ON? Sometimes?
Hi
When the parallel port driver exits, the port pins will be in an unknown state.
Most machines would either use a charge-pump circuit or an interlocked contactor, so that as soon controller exits, it fails safe.
With some revised wiring you should be able to do the same.
regards
When the parallel port driver exits, the port pins will be in an unknown state.
Most machines would either use a charge-pump circuit or an interlocked contactor, so that as soon controller exits, it fails safe.
With some revised wiring you should be able to do the same.
regards
The following user(s) said Thank You: TroyO
Please Log in or Create an account to join the conversation.
17 Sep 2013 21:46 #38936
by TroyO
Replied by TroyO on topic Parallel pin state on exit, stuck ON? Sometimes?
Yep, I think I can handle a charge pump on the relay driver so it will cut the mains as a "Fail safe" on the spindle. The steppers should be OK too, as long as they aren't getting a pulse train they would stop.
Just when I thought I had it nailed, LOL.....
Thanks for the info though, just knowing it's not a software/unexpected/unknown issue lets me know what path to chase for a fix.
Just when I thought I had it nailed, LOL.....
Thanks for the info though, just knowing it's not a software/unexpected/unknown issue lets me know what path to chase for a fix.
Please Log in or Create an account to join the conversation.
Time to create page: 0.060 seconds