Assuming I read your snippet of hal file properly, originally you had
estop-latch.0.ok-out
=> EstopOkOut
=> iocontrol.0.emc-enable-in
=> ppmc.0.dout.07.out
=> ppmc.0.dout.Estop.out
=> and2.0.in1
You disconnected EstopOutOk from iocontrol.0.emc-enable-in but that
signal is still driving the others: your ppmc outputs and the internal
'and' that is directed back to the estop latch.
Perhaps you wanted to remove the line giving EstopOkOut its value,
and then drive the signal to true:
#linksp EstopOkOut estop-latch.0.ok-out
sets EstopOkOut true
Anyway, my "way out there" suggestion is: Depending on your UI, the F1
key on the keyboard can activate an estop (this is true of all UIs
except I think touchy and maybe mini). You can find out whether it's
some UI source by setting the inifile debug level to 0x10 and running
linuxcnc in the terminal. If it's a UI estop, you'll see a line like
Issuing EMC_TASK_SET_STATE -- (+505,+24, +25, +1,)
(the last +1 indicates a request to change to ESTOP; +2 is ESTOP RESET,
+3 is OFF, and +4 is ON)
Things like amplifier faults and following errors are probably not what
is happening because (A) they take the machine to the "machine off"
state, not the "estop" state and (B) these show operator messages when
they occur (e.g., "joint 0 amplifier fault").
I did do some looking at the source while writing this e-mail to see if
it furnished any ideas, but I was looking at the development version
(2.7.0pre) which is quite different from 2.5.0. In particular, version
2.6.0 moved some estop handling from the program called "iocontrol" to
the program called "task", so the code I looked at is sure to not quite
be matching the code you're running. I don't see any estop related
bugfixes in our changelogs from 2.5.0 to 2.6.2, though.
(the moment you're struggling with a problem that is still likely to be
hardware is not the right moment to upgrade linuxcnc anyway)
Jeff
------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce.
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
