Re: [Emc-users] arcgenm18.py or lcnc's math or my brain has blown a fuse

[Gene Heskett]

On Friday 06 April 2018 18:19:15 John Figie wrote:

> G91 (incremental distances for arc)
>
> G91.1 (incremental distances for IJK in arcs, normal IOW)
>
> ( draw a semi-circle in climb mode, ccw from 90 to 270 degrees)
> ( this runs y .12", .06 up to .06 down )
> G3 X-0. Y-0.0600 I-0. J-0.0600
> G90 (back to absolute distances )
>
> I haven't programmed a g3 in a while. It looks to me like the end
> point and center are the same.
>
> G3 X-0. Y-0.0600 I-0. J-0.0300 ?
>
I don't think thats what I sent, but I'll check in the morning.

Thanks for looking at it, John.

> John Figie
>
> On Fri, Apr 6, 2018, 3:03 PM Gene Heskett  wrote:
> > Greetings;
> >
> > Trying to cut a short trench in the "tap hat" to put that m4 (7mm
> > diameter) screw head into.
> >
> > Machine is touched off just out of end of workpiece.
> >
> > Code, with G3 line copy/pasted from arcgenm18.py.
> > ===
> > g20g7
> > #<_pocket_width>=   0.27900 (about 7mm's)
> > #<_tool_dia>=   0.24900 (let dull tool flex .0005")
> > ( assumes mill is 1/4" thru, so circle diameter is .0300")
> > #<_circle_dia>  =   [#<_pocket_width> - #<_tool_dia>]
> > #<_wobble>  =   [#<_circle_dia> / .500] ( .015"
> > Y trades sign at
> > circle)
> > (debug,wobble=#<_wobble>)
> > ( x stop, draw semi circle to other side of wobble )
> > #<_x_minus> =   -0.300 (depth of circle
> > center from ends of arc)
> > #<_x_safe>  =   0.150 (out of workpiece)
> > #<_y_center>=   0.000 (for reference)
> > #<_z_top>   =   0.2800  (top surface of
> > jig) #<_z_final> =   -0.078 (depth into
> > jig to cut 2mm into brass)
> > #<_z_dec>   =   [0.358 /20] (establish
> > cut decrement)
> > (debug, zdec=#<_z_dec>)
> > s600
> > #<_z_tmp>   =   #<_z_top>
> > m6t10 (tool image small enough I can see past it in Z view)
> > m3
> >
> > o100 WHILE [#<_z_tmp> gt #<_z_final>] (do depth loop)
> >
> > #<_z_tmp>   =   [#<_z_tmp> - #<_z_dec>]
> > (decr depth till done)
> >
> > G90 (absolute distance mode for depth)
> >
> > (G1 F10 z#<_z_tmp>)
> >
> > G0 z0.3 (cut air)
> >
> > G1 F5 x#<_x_safe> ( put x out of workpiece)
> >
> > G1 F10 y#<_wobble> (now put Y at plus offset location)
> >
> > G1 F5 x#<_x_minus> (drive x to start of 180 degree arc)
> >
> > G91 (incremental distances for arc)
> >
> > G91.1 (incremental distances for IJK in arcs, normal IOW)
> >
> > ( draw a semi-circle in climb mode, ccw from 90 to 270 degrees)
> > ( this runs y .12", .06 up to .06 down )
> > G3 X-0. Y-0.0600 I-0. J-0.0600
> > G90 (back to absolute distances )
> >
> > G1 F5 y-#<_wobble> (should already be here)
> > (should be back to -wobble)
> >
> > G1 F5 x#<_x_safe> ( cut other edge of trench back out of workpiece)
> >
> > o100 ENDWHILE
> > m5
> > m2
> > %
> > ==
> > And nothing I adjust can keep me from huge G3 errors, or zero radius
> > errors.
> >
> > This is, or sure looks like, errors like I was getting when I
> > attempted to carve the pillows. There I asked for help but got
> > crickets, so I finally wound up digging the trenchs with a
> > SIN[#<_deg>] and a COS[#<_deg>] for the starting XZ position of a
> > cut nearly 4" long on the Y axis. Worked a treat but took a while
> > since I was only incrementing #<_deg> by 1 degree at a time.
> >
> > What error(s) are there in the above code?
> >
> > --
> > Cheers, Gene Heskett
> > --
> > "There are four boxes to be used in defense of liberty:
> >  soap, ballot, jury, and ammo. Please use in that order."
> > -Ed Howdershelt (Author)
> > Genes Web page 
> >
> >
> > 
> >-- Check out the vibrant tech community on one of the world's
> > most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > ___
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users
>
> --
> Check out the vibrant tech community on one of the world's
> most engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users



-- 
Cheers, Gene Heskett
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

--
Check out the 

Re: [Emc-users] arcgenm18.py or lcnc's math or my brain has blown a fuse

[John Figie]

G91 (incremental distances for arc)

G91.1 (incremental distances for IJK in arcs, normal IOW)

( draw a semi-circle in climb mode, ccw from 90 to 270 degrees)
( this runs y .12", .06 up to .06 down )
G3 X-0. Y-0.0600 I-0. J-0.0600
G90 (back to absolute distances )

I haven't programmed a g3 in a while. It looks to me like the end point and
center are the same.

G3 X-0. Y-0.0600 I-0. J-0.0300 ?

John Figie

On Fri, Apr 6, 2018, 3:03 PM Gene Heskett  wrote:

> Greetings;
>
> Trying to cut a short trench in the "tap hat" to put that m4 (7mm
> diameter) screw head into.
>
> Machine is touched off just out of end of workpiece.
>
> Code, with G3 line copy/pasted from arcgenm18.py.
> ===
> g20g7
> #<_pocket_width>=   0.27900 (about 7mm's)
> #<_tool_dia>=   0.24900 (let dull tool flex .0005")
> ( assumes mill is 1/4" thru, so circle diameter is .0300")
> #<_circle_dia>  =   [#<_pocket_width> - #<_tool_dia>]
> #<_wobble>  =   [#<_circle_dia> / .500] ( .015" Y
> trades sign at
> circle)
> (debug,wobble=#<_wobble>)
> ( x stop, draw semi circle to other side of wobble )
> #<_x_minus> =   -0.300 (depth of circle center
> from ends of arc)
> #<_x_safe>  =   0.150 (out of workpiece)
> #<_y_center>=   0.000 (for reference)
> #<_z_top>   =   0.2800  (top surface of jig)
> #<_z_final> =   -0.078 (depth into jig to cut
> 2mm into brass)
> #<_z_dec>   =   [0.358 /20] (establish cut
> decrement)
> (debug, zdec=#<_z_dec>)
> s600
> #<_z_tmp>   =   #<_z_top>
> m6t10 (tool image small enough I can see past it in Z view)
> m3
>
> o100 WHILE [#<_z_tmp> gt #<_z_final>] (do depth loop)
>
> #<_z_tmp>   =   [#<_z_tmp> - #<_z_dec>] (decr
> depth till done)
>
> G90 (absolute distance mode for depth)
>
> (G1 F10 z#<_z_tmp>)
>
> G0 z0.3 (cut air)
>
> G1 F5 x#<_x_safe> ( put x out of workpiece)
>
> G1 F10 y#<_wobble> (now put Y at plus offset location)
>
> G1 F5 x#<_x_minus> (drive x to start of 180 degree arc)
>
> G91 (incremental distances for arc)
>
> G91.1 (incremental distances for IJK in arcs, normal IOW)
>
> ( draw a semi-circle in climb mode, ccw from 90 to 270 degrees)
> ( this runs y .12", .06 up to .06 down )
> G3 X-0. Y-0.0600 I-0. J-0.0600
> G90 (back to absolute distances )
>
> G1 F5 y-#<_wobble> (should already be here)
> (should be back to -wobble)
>
> G1 F5 x#<_x_safe> ( cut other edge of trench back out of workpiece)
>
> o100 ENDWHILE
> m5
> m2
> %
> ==
> And nothing I adjust can keep me from huge G3 errors, or zero radius
> errors.
>
> This is, or sure looks like, errors like I was getting when I attempted
> to carve the pillows. There I asked for help but got crickets, so I
> finally wound up digging the trenchs with a SIN[#<_deg>] and a
> COS[#<_deg>] for the starting XZ position of a cut nearly 4" long on the
> Y axis. Worked a treat but took a while since I was only incrementing
> #<_deg> by 1 degree at a time.
>
> What error(s) are there in the above code?
>
> --
> Cheers, Gene Heskett
> --
> "There are four boxes to be used in defense of liberty:
>  soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author)
> Genes Web page 
>
>
> --
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] One servo system

[Marius Liebenberg]

Chris
I would like to run Linuxcnc but headless so the PocketBeagle will be 
great for that. I dont really need the network connection as the system 
will run a couple of simple commands that can be activated from buttons. 
My brief required a simple operators interface so it's buttons only.


-- Original Message --
From: "Chris Albertson" 
To: "Marius Liebenberg" ; "Enhanced Machine 
Controller (EMC)" 

Sent: 2018-04-05 18:54:35
Subject: Re: [Emc-users] One servo system


The comment to use psocketbeagle and MK if Linux is required is good.
I just looked up psocketbeagle and I don't see any network interface.
 Likely if you want a Linux based controller it is so you can access
it remotely. You might get a USB/Wifi dongle to work but it would
be better to buy a board with built-in networking if networking is
important.   But the pocket beagle could always use a serial line for
networking if all you need is a text window and a few file transfers.

If you don't need Linix then it can be much simpler (An Arduino would
work) but remote access is a very nice feature to have.  Nice enough
that I'd go with a card that has built in wifi.

On Thu, Apr 5, 2018 at 6:32 AM, Marius Liebenberg
 wrote:

 Hi Alexander
 I decided that this is the way to go. I did see that there is some 
talk that

 not all kernels are supported on the board.
 Is there a place where I can go look to see how to get this board 
setup with

 MachineKit please?

 -- Original Message --
 From: "Alexander Rössler" 
 To: "Marius Liebenberg" 
 Sent: 2018-04-04 13:37:56
 Subject: Re: [Emc-users] One servo system

 I think a simple real-time component should also do the trick in 
case
 the ladder logic component (it's also RT right?) does not work for 
your

 application.

 The PocketBeagle also has a PRU like the BeagleBone, so step 
generation

 or encoder feedback should be fine.

 --
 Alexander

 Marius Liebenberg writes:

  I just looked at the PocketBeagle and it seems to be a good choice 
for

  servo or stepper.
  It will have to be headless so I would imagine that I will be 
using

  ladder to control the system with.

  The track length is 6m, a button is pressed and I accelerate to 
8m/s and

  stop at the other end waiting for a command to do the same in the
  opposite direction. Machine must home in the one axis on demand as 
well.


  Would the ladder logic route be good enough or are there some 
other

  suggestions?


  -- Original Message --
  From: "Alexander Rössler" 
  To: "Enhanced Machine Controller (EMC)"
  
  Cc: "Marius Liebenberg" 
  Sent: 2018-04-04 09:24:43
  Subject: Re: [Emc-users] One servo system


 Machinekit + PocketBeagle is good combo if you are looking for
 something
 running Linux.

 Chris Albertson writes:

   Of course.   Write a PID based control loop and put it an 
something
   like an Arduino.  I've done this a few times the control the 
wheels

 of
   a mobile robot.

   The details depend on the interfaces to botkhthe motoring the 
higher

   level control


   As for higher level control...  Is it a human using a joystick 
or

 some
   kind of automation that does the same thing repetivity  or must 
the
   device interact with the environment some how.  Perhaps opening 
a for

   as a person approaches.

   Then I have to ask what kind of hardware driver you have.   
Some

   drivers do most of the work for you.  These might take a
   step/direction pulse, r have an analog input.  But other 
drivers are
   just an H-Bridge and raw encoder signals.This is the kind 
of
   serves I use.   I end u using a small micro controler (Arduino 
or

   ARM-M) and relaying on a PID library.

   In ant case most of your effort will be spent in the interfaces 
not
   the actual PID control loop as that is just a few calls to a 
PID

   library

   On Wed, Mar 28, 2018 at 8:15 AM, Marius Liebenberg
    wrote:


   Hi all
   I have to put together a system that will use one servo to 
drive a

 small
   platform over a distance of 6m at a speed of about 25km/h or 
8m/s or

 there
   about.

   I feel that using a full computer and mesa card for just one 
axis

 that does
   not need a display sounds a bit excessive. The question is 
what are

 the
   other options if any.



   -
   Regards / Groete

   Marius D. Liebenberg
   +27 82 698 3251
   +27 12 743 6064


 
--
   Check out the vibrant tech community on one of the world's 
most

   engaging tech sites, Slashdot.org! http://sdm.link/slashdot
   ___
   Emc-users mailing list
   Emc-users@lists.sourceforge.net