Re: Revised Tkbackplot

[Ray Henry <rehenry-at-up.net>]

At 08:18 PM 1/29/2000 -0500, Bill wrote:

>Don't have EMC hooked up to anything at this point, but tried out the
backplot
>program, and am unable to get it out of e-stop. Will actually it comes out of
>e-stop but the machine never turns on. Message reads 'hit reset again to turn
>machine on' or something close, but no matter how many times I hit that reset
>nothing happens. Once or twice I have got it to say 'waiting for home' but
that's
>it. Is this because it is not hooked up and looking for home switch
actuation?

I set the home switches so the control thinks they are closed.  I set the
overtravel so the machine thinks they are open.  Not having them connected
should not make a difference.  I've read about folk on the list who never
hook some of these things up.

You don't have a printer hooked up do you?

But I take it that emc comes up otherwise.  

When you get it to run with xemc, are the axis position numbers yellow or
red?   

>Works fine with Xemc, Yemc, and TkEmc. Tried to change from Bridgeport to
minimill
>with no luck (thats what I had to do with Xemc to get it to come out of
estop).

Are you able to home out with xemc?  Do the big numbers turn green?

>Using RH5.2, dec version of EMC, P-166 with 32meg. Also the mouse is a
little jerky,
>what setting to increase to fix that? From what I see I like though.
Thanks for the
>great work

I used 64 meg in the ini file.  Maybe you changed that already if so ignore
the next little bit.  

With 32 meg memory you change backplot.ini from:

; Base address for physical shared memory, e.g., for real-time motion.
; Note that if you change this, you may need to change OS parameters, e.g.,
; /etc/lilo.conf
; SHMEM_BASE_ADDRESS =	0x1F00000
SHMEM_BASE_ADDRESS =	0x3F00000

to:

SHMEM_BASE_ADDRESS =	0x1F00000
; SHMEM_BASE_ADDRESS =	0x3F00000


Backplot was set up to use the steppermod.o and minimill files.  I can't
believe that your chip speed would cause much of a problem.  I made the
display and the motion programs repeat often in order to get the resolution
of the plot up.  You could try doubling the value of these.

; Sections for display options
------------------------------------------------
[DISPLAY]

<snip>
; Cycle time, in seconds, that display controller will sleep between polls
CYCLE_TIME =		0.0500


[TASK]

<snip>
; Cycle time, in seconds, that task controller will sleep between polls
CYCLE_TIME =		0.0050

The last set of things to check is the polarity of the i/o stuff.  Each
axis has a number of polarities.  They include:

ENABLE_POLARITY =			0
MIN_LIMIT_SWITCH_POLARITY =		0
MAX_LIMIT_SWITCH_POLARITY =		0
HOME_SWITCH_POLARITY =		1
HOMING_POLARITY =			1
JOGGING_POLARITY =			1
FAULT_POLARITY =			1

This is the way that they run for me.  Min limit and max limit will turn
the axis numbers red if they are backwards.  I think that Enable and fault
can both hang up estop reset or machine on.

The last thing to check is down in the i/o section.

; digital in polarity, 0 is inverted, 1 is normal
; controller compares with polarity, equal means on, not equal means off

ESTOP_SENSE_POLARITY      =	0
LUBE_SENSE_POLARITY       =	1

Here estop sense is a signal that tells the computer that the machine has
gone into an external estop condition and that it should estop also.

I'd fiddle with these because parallel ports vary from machine to machine.
Changing from 0 to 1, one item at a time and see if they make a difference.  

Don't know about the mouse.  I've had problems with mine some also.

Get back to me cause I love a good fight. (read problem)

Ray