Re: [Emc-users] Slaved axis

[John Dammeyer]

I'm still pondering over replacing my ELS project with a Beaglebone running
MachineKit so it can run G-Code but use the keypad and display from my
original ELS and duplicate the manual style of operation when needed.  The
LCD display might well be replaced with a larger HDMI type Cell Phone size
display driven by the Beagle.  

What you see in this 11 year old video is automatic tapering.  But as I
described in the original posting of this thread the MPG knob and jog
buttons have the ELS cross slide and carriage synchronized.

https://youtu.be/RRaVIBhLgF0

So that's why I want it.  Whether it's a good idea or not isn't relevant.  I
could replace the PIC18F with a small module containing a PIC32 which has
Ethernet and USB.  But why reinvent a wheel when MachineKit does 90% of what
I want.

Thanks
John


> -Original Message-
> From: Gene Heskett [mailto:ghesk...@shentel.net]
> Sent: July-28-18 8:39 PM
> To: emc-users@lists.sourceforge.net
> Subject: Re: [Emc-users] Slaved axis
> 
> On Saturday 28 July 2018 20:13:20 John Dammeyer wrote:
> 
> > Quick question.  For a mill to move the tool bit from Point A to Point
> > B the software parsing the G Code Xn.nnnYm.mmm determines how fast to
> > accelerate and move each axis so a straight line is drawn between the
> > starting and new end points.  Similarly, if an arc is programed the
> > X,Y motion is including acceleration and deceleration to provide the
> > timed step pulses to move the tool in that arc motion.
> >
> > So with a lathe, the Z is the carriage, X is the cross slide. Again a
> > movement of Z with the idea of forming a taper means the Z and X are
> > specified and the system takes care of accelerating and moving each
> > axis the correct amount and speed so that a taper is cut.  For each
> > pass you'd move the X in for the next depth of cut and then specify a
> > relative move again. Spring passes without changing the X starting
> > point.
> >
> > What happens if you want to simulate something like my SouthBend with
> > the taper attachment where turning the carriage handwheel also moves
> > the cross slide along the taper.
> >
> > With MachineKit or LinuxCNC, the JOG keys or MPG moves the carriage.
> > Can the cross slide be programmed track the taper?  Is there a way to
> > 'link' the jog buttons so that they track with a mathematical formula?
> >  One could manually turn a ball end or some sort of other parabolic
> > curve for example.
> >
> > Thanks
> > John Dammeyer
> >
> Not too likely for doing it by hand John.  But in gcode, its a piece of
> cake to do to micron accuracy, whether its a straight line from A to B
> or even an arc, nurbs even if in the mood to cut a lot of air until
> you've got the pattern you want. See the G5 stuff for self calculating
> curves.
> 
> About the only way, and likely NOT repeatable, to do those sorts of
> curves is to find a big, long 4 way analog joystick. And figure out how
> to get its output into LCNC.
> 
> I have tried the usual usb "gamepad" like the Saitek, and have moved the
> little mill with it, but when zero to full speed is a 3/16" button
> motion, its simply uncontrollable.  With 3 or 4" extensions on the
> buttons and lots of practice, you might get used to it eventually, but
> broken tooling would still be a major expense while you were teaching
> your muscles to do whats pictured in your mind.
> >
> >
> >
> >
> > --
> > 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 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] Slaved axis

[Gene Heskett]

On Saturday 28 July 2018 20:13:20 John Dammeyer wrote:

> Quick question.  For a mill to move the tool bit from Point A to Point
> B the software parsing the G Code Xn.nnnYm.mmm determines how fast to
> accelerate and move each axis so a straight line is drawn between the
> starting and new end points.  Similarly, if an arc is programed the
> X,Y motion is including acceleration and deceleration to provide the
> timed step pulses to move the tool in that arc motion.
>
> So with a lathe, the Z is the carriage, X is the cross slide. Again a
> movement of Z with the idea of forming a taper means the Z and X are
> specified and the system takes care of accelerating and moving each
> axis the correct amount and speed so that a taper is cut.  For each
> pass you'd move the X in for the next depth of cut and then specify a
> relative move again. Spring passes without changing the X starting
> point.
>
> What happens if you want to simulate something like my SouthBend with
> the taper attachment where turning the carriage handwheel also moves
> the cross slide along the taper.
>
> With MachineKit or LinuxCNC, the JOG keys or MPG moves the carriage.
> Can the cross slide be programmed track the taper?  Is there a way to
> 'link' the jog buttons so that they track with a mathematical formula?
>  One could manually turn a ball end or some sort of other parabolic
> curve for example.
>
> Thanks
> John Dammeyer
>
Not too likely for doing it by hand John.  But in gcode, its a piece of 
cake to do to micron accuracy, whether its a straight line from A to B 
or even an arc, nurbs even if in the mood to cut a lot of air until 
you've got the pattern you want. See the G5 stuff for self calculating 
curves.

About the only way, and likely NOT repeatable, to do those sorts of 
curves is to find a big, long 4 way analog joystick. And figure out how 
to get its output into LCNC. 

I have tried the usual usb "gamepad" like the Saitek, and have moved the 
little mill with it, but when zero to full speed is a 3/16" button 
motion, its simply uncontrollable.  With 3 or 4" extensions on the 
buttons and lots of practice, you might get used to it eventually, but 
broken tooling would still be a major expense while you were teaching 
your muscles to do whats pictured in your mind.
>
>
>
>
> --
> 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 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] Slaved axis

[jeremy youngs]

On Sat, Jul 28, 2018, 7:32 PM andy pugh  wrote:

> On 29 July 2018 at 01:13, John Dammeyer  wrote:
>
> > With MachineKit or LinuxCNC, the JOG keys or MPG moves the carriage. Can
> the
> > cross slide be programmed track the taper?  Is there a way to 'link' the
> jog
> > buttons so that they track with a mathematical formula?
>
> Zero an indicator , then travel a known distance in z to figure taper,
> then g18 is your Friend.
--
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] Locked rotor

[jeremy youngs]

>
> One standard is A+ B- C- so I would suggest trying that.
> Tried that, no dice.


I believe these floating leads have something to do with it.
The mclr8 wire if touched will completely stop all function. Sometimes it's
smooth others it's not . Brand New encoder, factory assembled cable.

Thoughts on using the 7i90 and halscope to look at the waveforms???

And thoughts on termination for these unused leads? The mclr8 won't
terminate to ground or it stops the whole works ( of course no waveform
analysis yet )
And the u,v,w + floating on air don't seem to be a solid idea either but
there are no positions to terminate them on the drive.
I'm thinking it could be this encoder, I will try tomorrow to get
waveforms.
--
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] Slaved axis

[andy pugh]

On 29 July 2018 at 01:13, John Dammeyer  wrote:

> With MachineKit or LinuxCNC, the JOG keys or MPG moves the carriage. Can the
> cross slide be programmed track the taper?  Is there a way to 'link' the jog
> buttons so that they track with a mathematical formula?

All things are possible, but why?

With the external offsets branch you could use a lincurve linked to
the Z position to offset the X.

But why wouldn't you just do a coordinated G-code move?

-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916

--
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

[Emc-users] Slaved axis

[John Dammeyer]

Quick question.  For a mill to move the tool bit from Point A to Point B the
software parsing the G Code Xn.nnnYm.mmm determines how fast to accelerate
and move each axis so a straight line is drawn between the starting and new
end points.  Similarly, if an arc is programed the X,Y motion is including
acceleration and deceleration to provide the timed step pulses to move the
tool in that arc motion.

So with a lathe, the Z is the carriage, X is the cross slide. Again a
movement of Z with the idea of forming a taper means the Z and X are
specified and the system takes care of accelerating and moving each axis the
correct amount and speed so that a taper is cut.  For each pass you'd move
the X in for the next depth of cut and then specify a relative move again.
Spring passes without changing the X starting point.

What happens if you want to simulate something like my SouthBend with the
taper attachment where turning the carriage handwheel also moves the cross
slide along the taper.  

With MachineKit or LinuxCNC, the JOG keys or MPG moves the carriage. Can the
cross slide be programmed track the taper?  Is there a way to 'link' the jog
buttons so that they track with a mathematical formula?  One could manually
turn a ball end or some sort of other parabolic curve for example.

Thanks
John Dammeyer






--
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] hexapod

[andy pugh]

On 28 July 2018 at 19:30, Chris Albertson  wrote:
> Then it is not a "hexapod"  The word means literally "six legs"   What he
> has is a "Steward Platform".

Both are known as "Hexapods"

https://en.wikipedia.org/wiki/Hexapod


-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916

--
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] Locked rotor

[andy pugh]

On 28 July 2018 at 20:20, jeremy youngs  wrote:
> Also , cui says setting alignment will always align with u , so I presume
> the logic is standard, a,b,c is to u,v,w so I'm presuming that the a should
> be + and the b - for this alignment

One standard is A+ B- C- so I would suggest trying that.

-- 
atp
"A motorcycle is a bicycle with a pandemonium attachment and is
designed for the especial use of mechanical geniuses, daredevils and
lunatics."
— George Fitch, Atlanta Constitution Newspaper, 1916

--
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

[Emc-users] Locked rotor

[jeremy youngs]

Also , cui says setting alignment will always align with u , so I presume
the logic is standard, a,b,c is to u,v,w so I'm presuming that the a should
be + and the b - for this alignment , it's a 4 pole motor so it has 2
mechanical lock points 180 degrees apart , should it matter which one of
these mechanical points it's aligned with?
--
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

[Emc-users] Locked rotor for alignment

[jeremy youngs]

Kollmorgen eb 404
Does anyone know what phases to energize for locked rotor alignment of this
motor? I finally have my cui encoder and programming cables, I have aligned
with phase a + phase b- and set encoder. I can make it turn but it is rough
I don't think I have the right phase aligned. I have walked a,b,c wires to
no avail.
Also cui gives +u,v,w, as well as -u,v,w output , the amc b3040 drive only
has input for - and literature says it should be - so o have the + pins
floating I'm kinda amiff as to what to do with those. Thank you
--
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] hexapod

[Chris Albertson]

Then it is not a "hexapod"  The word means literally "six legs"   What he
has is a "Steward Platform". These are used as shown for mills but also for
airplanes simulators and now at amusement parts for those simulated rides

Steward Platforms have six motors and are 6DOF.  Hexapods hav 6 legs and
each legs and have 2, 3 or 4 degrees of freedom.

You can imagine why I was nterreted when I thought some one might
rebuilding a 24 axis machine kit based controller and was able to fit it
inside a robot.  typically peole ding this use much smaller controllers but
with poorer result, see video below.

Here is a video of about the simplest hexapod made with model airplanes
serves.   ITuses an 8-bit Arduino for control.   It needs a more powerfull
controller that 8-bit arduino is not able to run a fast enough control loop.
https://youtu.be/ojue8NCrs9g?t=1m13s

On Sat, Jul 28, 2018 at 1:24 AM Sven Wesley  wrote:

> I am pretty sure Dave's hexapod is a six axis milling setup. Not a walking
> spider robot.
> Like this one:
> https://www.youtube.com/watch?v=G_UmhUjZhNo
> https://www.youtube.com/watch?v=nebJ59TcYlQ
>
> On Sat, Jul 28, 2018 at 4:54 AM Chris Albertson  >
> wrote:
>
> > What kind of hexapod are you building?  I've have a long interest in them
> > but, so far have only built single legs as prototypes.   All of then had
> > short lifetimes.
> >
> > What what kind of motors are you using and how many per leg?  Finding
> good
> > motors at a reasonable price is important.  With only 3 per leg you need
> 18
> > motor and an 18-axis controller.  With 4 per leg you have a 24-axis
> system
> > and even if you only spend $100 per axis it costs a bit.
> >
> > The inverse kinematics is not hard to work out.  Just a long and repeated
> > application of the plain geometry and so trig.
> >
> >
> >
> > On Fri, Jul 27, 2018 at 7:23 PM dave engvall 
> wrote:
> >
> > > Oh damn, I just sent my reply to Andy not to the list so here goes
> again.
> > >
> > > I just had Ray Henry on the phone and picked his brains. He says tk was
> > > written for the hexapod so one needs to find the hexapod kinematics.
> > >  From there is should be easy or so he says. I've never tried it so all
> > > I can do is relay information.  Sorry I can't be more help.
> > >
> > > An "old" wiki may shed more light.
> > >
> > > Dave
> > >
> > >
> > > On 07/27/2018 02:59 PM, andy pugh wrote:
> > >  > On 27 July 2018 at 22:01, yomin estiven jaramillo munera
> > >  >  wrote:
> > >  >> hi to everybody, my application is a hexapod and i decided to
> change
> > > my GUI
> > >  >> from axis to tklinuxcnc, my problem is that in the axis GUI all my
> > axis
> > >  >> move without problems but in tklinuxcnc i have problems with the Y
> > axis
> > >  >> specifically , This is the error:
> > >  > Which version of LinuxCNC are you using?
> > >  >
> > >  > I am not sure if tklinuxCNC is entirely aware of non-identity
> > > kinematics.
> > >  >
> > >
> > >
> > >
> > >
> > >
> >
> --
> > > 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
> > >
> >
> >
> > --
> >
> > Chris Albertson
> > Redondo Beach, California
> >
> >
> --
> > 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
>


-- 

Chris Albertson
Redondo Beach, California
--
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] hexapod

[dave engvall]

On 07/27/2018 07:49 PM, Chris Albertson wrote:

What kind of hexapod are you building?  I've have a long interest in them
but, so far have only built single legs as prototypes.   All of then had
short lifetimes.

What what kind of motors are you using and how many per leg?  Finding good
motors at a reasonable price is important.  With only 3 per leg you need 18
motor and an 18-axis controller.  With 4 per leg you have a 24-axis system
and even if you only spend $100 per axis it costs a bit.

The inverse kinematics is not hard to work out.  Just a long and repeated
application of the plain geometry and so trig.



On Fri, Jul 27, 2018 at 7:23 PM dave engvall  wrote:


Oh damn, I just sent my reply to Andy not to the list so here goes again.

I just had Ray Henry on the phone and picked his brains. He says tk was
written for the hexapod so one needs to find the hexapod kinematics.
  From there is should be easy or so he says. I've never tried it so all
I can do is relay information.  Sorry I can't be more help.

An "old" wiki may shed more light.

Dave


On 07/27/2018 02:59 PM, andy pugh wrote:
  > On 27 July 2018 at 22:01, yomin estiven jaramillo munera
  >  wrote:
  >> hi to everybody, my application is a hexapod and i decided to change
my GUI
  >> from axis to tklinuxcnc, my problem is that in the axis GUI all my axis
  >> move without problems but in tklinuxcnc i have problems with the Y axis
  >> specifically , This is the error:
  > Which version of LinuxCNC are you using?
  >
  > I am not sure if tklinuxCNC is entirely aware of non-identity
kinematics.
  >


I'm back. Maybe this will help a bit.

http://parallelrobots.blogspot.com/2012/04/linuxcnc-hexapod-machine-tool.html

There is lots a material on the web but also many expired/dead links. 
;-) Google is sometimes your friend.


Dave





--
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] hexapod

[Sven Wesley]

I am pretty sure Dave's hexapod is a six axis milling setup. Not a walking
spider robot.
Like this one:
https://www.youtube.com/watch?v=G_UmhUjZhNo
https://www.youtube.com/watch?v=nebJ59TcYlQ

On Sat, Jul 28, 2018 at 4:54 AM Chris Albertson 
wrote:

> What kind of hexapod are you building?  I've have a long interest in them
> but, so far have only built single legs as prototypes.   All of then had
> short lifetimes.
>
> What what kind of motors are you using and how many per leg?  Finding good
> motors at a reasonable price is important.  With only 3 per leg you need 18
> motor and an 18-axis controller.  With 4 per leg you have a 24-axis system
> and even if you only spend $100 per axis it costs a bit.
>
> The inverse kinematics is not hard to work out.  Just a long and repeated
> application of the plain geometry and so trig.
>
>
>
> On Fri, Jul 27, 2018 at 7:23 PM dave engvall  wrote:
>
> > Oh damn, I just sent my reply to Andy not to the list so here goes again.
> >
> > I just had Ray Henry on the phone and picked his brains. He says tk was
> > written for the hexapod so one needs to find the hexapod kinematics.
> >  From there is should be easy or so he says. I've never tried it so all
> > I can do is relay information.  Sorry I can't be more help.
> >
> > An "old" wiki may shed more light.
> >
> > Dave
> >
> >
> > On 07/27/2018 02:59 PM, andy pugh wrote:
> >  > On 27 July 2018 at 22:01, yomin estiven jaramillo munera
> >  >  wrote:
> >  >> hi to everybody, my application is a hexapod and i decided to change
> > my GUI
> >  >> from axis to tklinuxcnc, my problem is that in the axis GUI all my
> axis
> >  >> move without problems but in tklinuxcnc i have problems with the Y
> axis
> >  >> specifically , This is the error:
> >  > Which version of LinuxCNC are you using?
> >  >
> >  > I am not sure if tklinuxCNC is entirely aware of non-identity
> > kinematics.
> >  >
> >
> >
> >
> >
> >
> --
> > 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
> >
>
>
> --
>
> Chris Albertson
> Redondo Beach, California
>
> --
> 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