Re: [Emc-users] West Coast Mills (USA)

[Kirk Wallace]

On 09/29/2016 01:11 PM, Jim Craig wrote:
... snip
> My machine uses NMTB30 tool holders so it would not be a direct fit. The
> concept is good. I might try to replicate the mechanism. However the
> wrist looks like it could be complicated to get that motion.
>
> Do you happen to have an exploded view drawing of the arm/wrist mechanism?
... snip

I have seen an exploded diagram of this arm assembly, but I don't recall 
where at the moment. I'll keep this in mind and pass it along if I find 
anything. Hmm, I think here:
http://www.wallacecompany.com/machine_shop//Shizuoka/atc_manual/bandit_changer-0064.png
http://www.wallacecompany.com/machine_shop//Shizuoka/atc_manual/bandit_changer-0065.png
http://www.wallacecompany.com/machine_shop//Shizuoka/atc_manual/bandit_changer-0066.png
http://www.wallacecompany.com/machine_shop//Shizuoka/atc_manual/

The arm slides in a vertical tube with a spiral slot that rotates the 
arm to or from the tool box or to the spindle as the arm slides up and 
down. The arm operates at fairly low pressure, so no one gets hurt if 
body parts get in the way, which is nice. The wrist has a bevel gear 
which is driven by a bevel gear on the vertical inner shaft which is 
fixed. When the arm rotates, wrist rotates as well.

I tend to think I would make a more robot like arm if I were making one 
from scratch. The current arm has to remove the old tool from the 
spindle and put it away before getting the new tool, then load it. Some 
changers have the new tool ready before making a change. With LinuxCNC 
on-board there is plenty of control resource available.

-- 
Kirk Wallace
http://www.wallacecompany.com/machine_shop/
http://www.wallacecompany.com/E45/

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] G33.1 overtravel revisited.

[Gene Heskett]

On Thursday 29 September 2016 17:26:26 John Thornton wrote:

> http://linuxcnc.org/docs/devel/html/
>
Thanks John. Skimming thru that, the most promising path I see might be 
in cobbling up something useing sampler, hal_sampler, streamer and 
hal_streamer, if the output could somehow be dumped, not to a file, but 
to a #<_named> memory variable, which could then be massaged in gcode 
and used as an -z[#<_z_arg> + #<_derived_offset>. in the G33.1 
invocation.  Has anyone actually done that, that someone here is aware 
of?

> On 9/29/2016 4:11 PM, Gene Heskett wrote:
> > On Thursday 29 September 2016 17:00:20 Gene Heskett wrote:
> >
> > This question may have the genesis of a clue in the Integrators
> > Manual, unforch the copy available on the web page, development
> > branch dated about 2 weeks ago, is not in English.  Is there any way
> > to fix that?
> >
> > Thanks.
> >
> > Cheers, Gene Heskett
>
> --
> ___
> 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 

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] G33.1 overtravel revisited.

[John Thornton]

http://linuxcnc.org/docs/devel/html/

On 9/29/2016 4:11 PM, Gene Heskett wrote:
> On Thursday 29 September 2016 17:00:20 Gene Heskett wrote:
>
> This question may have the genesis of a clue in the Integrators Manual,
> unforch the copy available on the web page, development branch dated
> about 2 weeks ago, is not in English.  Is there any way to fix that?
>
> Thanks.
>
> Cheers, Gene Heskett


--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] G33.1 overtravel revisited.

[Gene Heskett]

On Thursday 29 September 2016 17:00:20 Gene Heskett wrote:

This question may have the genesis of a clue in the Integrators Manual, 
unforch the copy available on the web page, development branch dated 
about 2 weeks ago, is not in English.  Is there any way to fix that?

Thanks.

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 

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] G33.1 overtravel revisited.

[Gene Heskett]

On Thursday 29 September 2016 12:15:01 Gene Heskett wrote:

> Greetings all;
>
> Sitting here waiting for the coffee to kick in, and lamenting the
> difficulty in doing math in hal.
>
> That bit of hal stuff I can watch with a halmeter needs some math to
> translate the encoder counts of the overshoot into actual distance it
> represents. Then somehow, dynamically as the g33.1 is running, diddle
> the z distance its using as the turn around point so the first one,
> cutting air, would establish the z offset for the next peck, doing
> this on a peck by peck basis.  A wrapper for the G33.1 could do that
> before starting the next peck. I think that overshoot value is still
> valid at the point where its back to doing its incrementing of the
> depth it hands g33.1 for the next peck.
>
> So, how can I store the output of a hal module to memory?,
>
> Such that the gcode can access it #4990 style and use it to adjust the
> z point its working with. IOW I need to be able to make a memory
> location out of something one would normally be looked at with a
> halmeter.  With that, I could do the math in gcode to make a g33.1
> peck cycle wrapper self correct for the overshoot.
>
> Here is what I have now: just incomplete enough to be confusing I'm
> sure. 
> # Now, lets see if we can track the G33.1 turnaround overtravel
> # first, make a sum2 a sub2
> setpsum2.ovrtrvl.gain0  -1.
>
> # then net the sample holds but sample needs s32, not float so use
> count net ovrtrvl1hm2_5i25.0.encoder.00.count
> sample.dirchg.in sample.spndlchg.in net spindle-ccw => 
> sample.dirchg.hold
> net spindle-reverse abs.encdir.is-negative 
> sample.spndlchg.hold
>
> # and calc the overtravel from S32ś
> net ovrtrvl2sample.dirchg.out  
> S32_float_cmd.in net ovrtrvl3S32_float_cmd.out
>   sum2.ovrtrvl.in0 net ovrtrvl4   
> sample.spndlchg.out S32_float_spndl.in net ovrtrvl5   
> S32_float_spndl.out sum2.ovrtrvl.in1
>
> # the hal pin sample.ovrtrvl.out should be shown with a hal__meter
> ==
> And its this last sample.ovrtrvl.out that I want to put in a memory
> location.
>
>
> Cheers, Gene Heskett

And I was running into a get out of here deadline to get the toy over to 
toy for some recall work, and forgot to say thank you for any ideas that 
could be used.

Thank you.

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 

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Any Interest or Ideas for a Linuxcnc Fest 2016 ?

[Stuart Stevenson]

There is room for a camper.

On Sep 29, 2016 3:25 PM, "Ed"  wrote:

> On 09/28/2016 05:36 PM, Charles Steinkuehler wrote:
> > On 8/22/2016 11:55 AM, Sebastian Kuzminsky wrote:
> >> On 08/22/2016 10:35 AM, Ed wrote:
> >>> I haven't been paying attention to this. Is this the correct date?
> >> The plan is:
> >>
> >> Monday October 17 through Sunday October 23
> >> Helix Machine, 3434 West Harry, Wichita, Kansas
> >>
> >> It's going to be a rocking nerdy good time.
> > Sounds great!  Is there any sort of tentative schedule?
> >
> > I'm trying to decide whether to come down Fri/Sat or Sat/Sun.
> >
> Is there a place to park a 24 foot camper on the grounds? If not is
> there a good place to camp close by?
>
> We would be there Thursday late or Friday early till Sunday afternoon.
>
> It looks like there are two of us coming from SW Wisconsin. Is there any
> others from the area going?
>
> I have a couple of projects that I will need some instruction on.
>
> One is an Fanuc AC servo setup with a Pico setup. The other is to setup
> a follower system; encoder on a spindle and a servo drive at a specific
> ratio. EG; hobbing on a horizontal mill.
>
> Ed.
>
>
> 
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Any Interest or Ideas for a Linuxcnc Fest 2016 ?

[Ed]

On 09/28/2016 05:36 PM, Charles Steinkuehler wrote:
> On 8/22/2016 11:55 AM, Sebastian Kuzminsky wrote:
>> On 08/22/2016 10:35 AM, Ed wrote:
>>> I haven't been paying attention to this. Is this the correct date?
>> The plan is:
>>
>> Monday October 17 through Sunday October 23
>> Helix Machine, 3434 West Harry, Wichita, Kansas
>>
>> It's going to be a rocking nerdy good time.
> Sounds great!  Is there any sort of tentative schedule?
>
> I'm trying to decide whether to come down Fri/Sat or Sat/Sun.
>
Is there a place to park a 24 foot camper on the grounds? If not is 
there a good place to camp close by?

We would be there Thursday late or Friday early till Sunday afternoon.

It looks like there are two of us coming from SW Wisconsin. Is there any 
others from the area going?

I have a couple of projects that I will need some instruction on.

One is an Fanuc AC servo setup with a Pico setup. The other is to setup 
a follower system; encoder on a spindle and a servo drive at a specific 
ratio. EG; hobbing on a horizontal mill.

Ed.


--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] West Coast Mills (USA)

[Jim Craig]

On 9/29/2016 2:37 PM, Kirk Wallace wrote:
> On 09/29/2016 12:00 PM, Jim Craig wrote:
>> On 9/29/2016 1:02 PM, Kirk Wallace wrote:
>>> Just in case someone might be interested:
>>> http://www.ebay.com/itm/121734977870
>>>
>>> My slightly smaller ST-N as reference:
>>> http://www.wallacecompany.com/machine_shop/Shizuoka/
>>>
>> That tool changer setup looks perfect for my machine. Do you have a
>> video of it in action? Any details on the tool changer arm?
> http://www.wallacecompany.com/machine_shop/Shizuoka/tool_changer_video.html
>
> The Shizuoka uses NMTB40 tool holders. If your machine uses NMTB40 the
> changer might work, otherwise the claw and tool pockets will be a
> problem (NMTB = single flange with boss on end of taper, CAT = dual
> flange, no boss). The arm is rotated by a pneumatic cylinder which makes
> the arm raise and lower, as well as rotate the wrist. The tool comes out
> of the spindle with the tool pointing down. The wrist turns the tool
> holder over, tool up, as the arm moves towards the carousel. The changer
> functions are controlled by relay contacts so are pretty easy to hook up.
>
>
My machine uses NMTB30 tool holders so it would not be a direct fit. The 
concept is good. I might try to replicate the mechanism. However the 
wrist looks like it could be complicated to get that motion.

Do you happen to have an exploded view drawing of the arm/wrist mechanism?

Thanks again,

Jim


--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] West Coast Mills (USA)

[Kirk Wallace]

On 09/29/2016 12:00 PM, Jim Craig wrote:
> On 9/29/2016 1:02 PM, Kirk Wallace wrote:
>> Just in case someone might be interested:
>> http://www.ebay.com/itm/121734977870
>>
>> My slightly smaller ST-N as reference:
>> http://www.wallacecompany.com/machine_shop/Shizuoka/
>>
> That tool changer setup looks perfect for my machine. Do you have a
> video of it in action? Any details on the tool changer arm?

http://www.wallacecompany.com/machine_shop/Shizuoka/tool_changer_video.html

The Shizuoka uses NMTB40 tool holders. If your machine uses NMTB40 the 
changer might work, otherwise the claw and tool pockets will be a 
problem (NMTB = single flange with boss on end of taper, CAT = dual 
flange, no boss). The arm is rotated by a pneumatic cylinder which makes 
the arm raise and lower, as well as rotate the wrist. The tool comes out 
of the spindle with the tool pointing down. The wrist turns the tool 
holder over, tool up, as the arm moves towards the carousel. The changer 
functions are controlled by relay contacts so are pretty easy to hook up.


-- 
Kirk Wallace
http://www.wallacecompany.com/machine_shop/
http://www.wallacecompany.com/E45/

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Function of MAX_JERK?

[sam sokolik]

There is a branch of linuxcnc that one of the developers was working on 
(micges/master/limited_jerk) based on work by Yi-Shin Li.  This used the 
1 segment look ahead planner (pre 2.7) and added jerk limited 
calculations.  It was never finished and still has issues with acc/vel 
violations.

This branch did use the MAX_JERK ini setting.

sam

On 9/29/2016 2:00 PM, Jeff Epler wrote:
> MAX_JERK is not used by LinuxCNC (up to and including the development branch)
>
> Jeff
>
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
>

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Function of MAX_JERK?

[Jeff Epler]

MAX_JERK is not used by LinuxCNC (up to and including the development branch)

Jeff

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] West Coast Mills (USA)

[Jim Craig]

On 9/29/2016 1:02 PM, Kirk Wallace wrote:
> Just in case someone might be interested:
> http://www.ebay.com/itm/121734977870
>
> My slightly smaller ST-N as reference:
> http://www.wallacecompany.com/machine_shop/Shizuoka/
>
That tool changer setup looks perfect for my machine. Do you have a 
video of it in action? Any details on the tool changer arm?

Thanks,

Jim


--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Function of MAX_JERK?

[dannym]

I was doing tuning for my 3D carving profile and started experimenting with the 
per-axis MAX_JERK in the .ini file [AXIS_x] sections.  LinuxCNC 2.7.4.

The value I had initially was 2000, I gave it 20 and 2.

It made no change at all to the runtime, nor did the machine look to jerk 
around any more or less.  

It's a 3D carving file with a LOT of small vectors and takes about 4x longer to 
run than the estimator says.  So acceleration and jerk should have prominent 
effects on runtime.

Is this "MAX_JERK" functionality hooked up to anything?  I didn't see it under 
the man page online, but the url does have "2.6" in it.

Thanks,
Danny


--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] West Coast Mills (USA)

[Kirk Wallace]

Just in case someone might be interested:
http://www.ebay.com/itm/121734977870

My slightly smaller ST-N as reference:
http://www.wallacecompany.com/machine_shop/Shizuoka/

-- 
Kirk Wallace
http://www.wallacecompany.com/machine_shop/
http://www.wallacecompany.com/E45/

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] G33.1 overtravel revisited.

[Gene Heskett]

Greetings all;

Sitting here waiting for the coffee to kick in, and lamenting the 
difficulty in doing math in hal.

That bit of hal stuff I can watch with a halmeter needs some math to 
translate the encoder counts of the overshoot into actual distance it 
represents. Then somehow, dynamically as the g33.1 is running, diddle 
the z distance its using as the turn around point so the first one, 
cutting air, would establish the z offset for the next peck, doing this 
on a peck by peck basis.  A wrapper for the G33.1 could do that before 
starting the next peck. I think that overshoot value is still valid at 
the point where its back to doing its incrementing of the depth it hands 
g33.1 for the next peck.

So, how can I store the output of a hal module to memory?,

Such that the gcode can access it #4990 style and use it to adjust the 
z point its working with. IOW I need to be able to make a memory 
location out of something one would normally be looked at with a 
halmeter.  With that, I could do the math in gcode to make a g33.1 peck 
cycle wrapper self correct for the overshoot.

Here is what I have now: just incomplete enough to be confusing I'm sure.

# Now, lets see if we can track the G33.1 turnaround overtravel
# first, make a sum2 a sub2
setpsum2.ovrtrvl.gain0  -1.

# then net the sample holds but sample needs s32, not float so use count
net ovrtrvl1hm2_5i25.0.encoder.00.count 
sample.dirchg.in sample.spndlchg.in
net spindle-ccw =>  sample.dirchg.hold
net spindle-reverse abs.encdir.is-negative  
sample.spndlchg.hold

# and calc the overtravel from S32ś
net ovrtrvl2sample.dirchg.out   S32_float_cmd.in
net ovrtrvl3S32_float_cmd.out   sum2.ovrtrvl.in0
net ovrtrvl4sample.spndlchg.out 
S32_float_spndl.in
net ovrtrvl5S32_float_spndl.out sum2.ovrtrvl.in1

# the hal pin sample.ovrtrvl.out should be shown with a hal__meter
==
And its this last sample.ovrtrvl.out that I want to put in a memory 
location.


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 

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[Florian Rist]

Hi
I just tried to link all three axis:

net emcmot.00.pos-cmd   axis.0.motor-pos-cmd=> pid.0.command
net emcmot.00.pos-cmd => pid.1.command
net emcmot.00.pos-cmd => pid.2.command

That seams to work. I can't home any more, but for now that's good 
enough.

Thanks for your super quick support. :-)

See you
Florian



--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[John Kasunich]

On Thu, Sep 29, 2016, at 11:15 AM, John Kasunich wrote:
> 
> 
> On Thu, Sep 29, 2016, at 10:48 AM, Florian Rist wrote:

> > Could I do something with the axis jog pins:
> > 
> >   axis.0.jog-vel-mode
> >   axis.0.jog-enable
> >   axis.0.jog-scale
> 
> Those are for using a jogwheel.  See
> http://linuxcnc.org/docs/2.7/html/config/core-components.html
> 
> You could use a stepgen in quadrature mode to fake the jogwheel
> counts, and use two buttons to make that stepgen go up or down.
> If you send the "jogwheel" counts to all axes, set the axis.N.jog-scale
> the same for all axes, and turn on axis.N.jog-enable for all axes,
> they WILL move the same distance.

Actually you don't even have to mess around with the stepgen.
All you need is an S32 signal that can be increment or decremented.

The updown component can generate such a signal.  It requires
pulses on its up or down input pins to count.

To make pulses from a pushbutton you could use a pair of xor2
components.  (One for up, one for down).  Connect the button to
one of the XOR inputs.  Connect the XOR output to its other input
(feedback).  When the button signal is "1", the output will toggle
every servo period - meaning 500 counts per second if using a 1mS
servo thread.  When the button signal is "0" the output will sit still.

If you know that the updown block will be counting up or down
at 500Hz, you can set axis.N.jog-scale to get whatever speed you
want at the motor.  For example, a scale of 0.01" per tick times
500 ticks per second would be 5 inches per second.





-- 
  John Kasunich
  jmkasun...@fastmail.fm

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[John Kasunich]

On Thu, Sep 29, 2016, at 10:48 AM, Florian Rist wrote:
> Hi John
> 
> > Do you need/intend to use g-code at all for this installation?
> 
> Yes.

In that case you want the LinuxCNC motion planner.  It will
handle all the jogging, homing, etc, for you.

> 
> And I'm now just trying to modify it quickly to do a manual 
> demonstration.
> 
> Could I do something with the axis jog pins:
> 
>   axis.0.jog-vel-mode
>   axis.0.jog-enable
>   axis.0.jog-scale

Those are for using a jogwheel.  See
http://linuxcnc.org/docs/2.7/html/config/core-components.html

I just read that using halui didn't work because the axes don't 
track perfectly (not really a surprise, since halui is a non-realtime
component).

You could use a stepgen in quadrature mode to fake the jogwheel
counts, and use two buttons to make that stepgen go up or down.
If you send the "jogwheel" counts to all axes, set the axis.N.jog-scale
the same for all axes, and turn on axis.N.jog-enable for all axes,
they WILL move the same distance.

 
> > I have a few custom HAL components that I could share if useful.
> > For example, one lets you switch from manually controlling each
> > motor (for "homing") to coordinated control.
>
> Oh, sounds very interesting.

As long as you are using LinuxCNC's motion controller and g-code
you don't need my components.


-- 
  John Kasunich
  jmkasun...@fastmail.fm

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[Florian Rist]

Hi Andy,
thanks for the suggestion:

> Connect the buttons to the halui.jog.N.minus  and halui.jog.N.plus pins

That is indeed very easy. I just tried this. It kind of works. 
Unfortunately the axis do not move fully synchronized. X starts first, a 
fraction of second later y start to turn and y is last. On multiple jogs 
this results in the three position slowly drifting apart.

I guess I could link all three motor position commands, but than I cant 
home individually any more?

See you
Florina

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[Florian Rist]

Hi John

> Do you need/intend to use g-code at all for this installation?

Yes.

And I'm now just trying to modify it quickly to do a manual 
demonstration.

Could I do something with the axis jog pins:

  axis.0.jog-vel-mode
  axis.0.jog-enable
  axis.0.jog-scale


> I have a few custom HAL components that I could share if useful.
> For example, one lets you switch from manually controlling each
> motor (for "homing") to coordinated control.

Oh, sounds very interesting.

See you
Florian

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[andy pugh]

On 29 September 2016 at 14:58, Florian Rist  wrote:

> Whats the best way to do this

Connect the buttons to the halui.jog.N.minus  and halui.jog.N.plus pins

http://linuxcnc.org/docs/2.7/html/man/man1/halui.1.html


-- 
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

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Any Interest or Ideas for a Linuxcnc Fest 2016 ?

[Gene Heskett]

On Thursday 29 September 2016 08:34:53 Jim Craig wrote:

> On 9/28/2016 9:35 PM, andy pugh wrote:
> > On 29 September 2016 at 02:32, Jeff Epler  
wrote:
> >> Besides the social aspects, some of my goals are to
> >>   - take a look at some of the things in our github tracker that
> >> are languishing
> >
> > One thing that I find myself coming back to, and is looking
> > increasingly like the real answer to "jog while paused" is extending
> > the 'position.txt" file, or a close relative, to  include "last
> > G-code line completed' and a collection of states such as tool and
> > spindle speed.
> >
> > Some users have said that mains power is unreliable where they are.
> > If the power goes off mid-job, then they want a pop-up next restart
> > saying so.
> >
> > Is is possible that we have over-complicated the whole
> > jog-while-paused thing? Could it really be as easy as "re-start
> > unfinished job" ?
>
> On a similar note. I have wondered why we could not have another
> option for run from line "X" where the interpreter actually reads all
> of the previous lines but does not act on them just to make sure all
> of the persistent codes are set correctly for the program to run from
> the line without having to worry about the units, coordinate system,
> spindle speed, feedrate, etc.
>
> Just a thought as sometimes restarting a program partially thorough
> can be a hair raising experience.
>
> Jim
>
+1 at least, I got bit the 2nd time I tried that, wrecked a $15 tool and 
$20+ bucks worth of mahogany. So I take my chances & restart from the 
top.
>
> --
> ___
> 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 

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] snagged touch probe on eBay - anybody know how to interface ?

[Gene Heskett]

On Thursday 29 September 2016 07:50:23 Sarah Armstrong wrote:

> i'm looking at a renishaw MP10 IR probe
> iv'e posted some snipps of what i am , ( or better think i am )
> getting as an output
>
> https://www.dropbox.com/sh/kwjxef6k4ptbhhb/AABdCLSjzypH574XoXJv_it3a?d
>l=0
>
> it looks to me all it's doing is giving a constant mark space ratio ,
> and stopping on a probe hit  , surely it cant be that simple , or
> could it i must admit at this point , i have not dug anydeeper as yet
> , or really looked to see if i can find any data , but it's very
> obvious everything stops on a hit
>
> i was expecting some data or id and say battery condition , but just
> early days
> found i had a better signal using a 1838 ir that a qse159 but i may
> need some better signal conditioning at the moment
>
What I can see in the scope shots looks pretty clean. All you'd need is 
to run it thru a comparator to make a rail to rail digital signal out of 
that if its not already.

If there was a timebase setting visible in the scope shots, I missed it. 
That would determine how fast you should be moving to achieve the 
accuracy needed.

Detection of the hit should probably be done with a re-triggerable one 
shot (the cmos 7555 comes to mind, batteries last till the rapture) 
whose timeout period is set for about 5% longer than the period between 
pulses, so the one shot would time out quickly after the first missing 
pulse, using that as the hit indicator. I would, on the off chance it 
puts out a last pulse at the hit point, probably longer, independent of 
the pulse train timing which might only be used as a keep alive.  So the 
scope trace should be looked at very carefully at the point where it 
stops, with an eye toward detecting only that unique pulse as the hit 
pulse.

But I suspect you knew that already. :)

[...] 

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 

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[John Kasunich]

Do you need/intend to use g-code at all for this installation?

If not, it is probably a good candidate for a "hal-only" configuration.
Stepgens, limit2 or limit3 blocks, muxes, integrators, etc, can be used
to do just about anything you can imagine.  PyVCP can be used for
on-screen controls, and joysticks, buttons or other devices for external
controls.

As an example: http://www.teslaorchestra.com/claw-game.html

I have a few custom HAL components that I could share if useful.
For example, one lets you switch from manually controlling each 
motor (for "homing") to coordinated control.



On Thu, Sep 29, 2016, at 09:58 AM, Florian Rist wrote:
> Hi,
> just a simple question:
> 
> I need to add two momentary switches to a 3 axis machine (not a mill, 
> more like a art installation) that jog all axis (xyz). Press one button 
> all axis move up, press the other one all move done in sync.
> 
> Whats the best way to do this. The three axis are moved by servo motors 
> via a Mesa card and Mesa amplifiers.
> 
> I have plenty of unused IOs on the Mesa card.
> 

-- 
  John Kasunich
  jmkasun...@fastmail.fm

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Best way to jog all three axis using a single push button?

[Dave Caroline]

Maybe some kins so it is a real vector (I think that is what another
art installation did)

Dave Caroline

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Best way to jog all three axis using a single push button?

[Florian Rist]

Hi,
just a simple question:

I need to add two momentary switches to a 3 axis machine (not a mill, 
more like a art installation) that jog all axis (xyz). Press one button 
all axis move up, press the other one all move done in sync.

Whats the best way to do this. The three axis are moved by servo motors 
via a Mesa card and Mesa amplifiers.

I have plenty of unused IOs on the Mesa card.

See you
Flo

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Any Interest or Ideas for a Linuxcnc Fest 2016 ?

[Jim Craig]

On 9/28/2016 9:35 PM, andy pugh wrote:
> On 29 September 2016 at 02:32, Jeff Epler  wrote:
>
>> Besides the social aspects, some of my goals are to
>>   - take a look at some of the things in our github tracker that are
>> languishing
>>
> One thing that I find myself coming back to, and is looking increasingly
> like the real answer to "jog while paused" is extending the 'position.txt"
> file, or a close relative, to  include "last G-code line completed' and a
> collection of states such as tool and spindle speed.
>
> Some users have said that mains power is unreliable where they are. If the
> power goes off mid-job, then they want a pop-up next restart saying so.
>
> Is is possible that we have over-complicated the whole jog-while-paused
> thing? Could it really be as easy as "re-start unfinished job" ?
>
On a similar note. I have wondered why we could not have another option 
for run from line "X" where the interpreter actually reads all of the 
previous lines but does not act on them just to make sure all of the 
persistent codes are set correctly for the program to run from the line 
without having to worry about the units, coordinate system, spindle 
speed, feedrate, etc.

Just a thought as sometimes restarting a program partially thorough can 
be a hair raising experience.

Jim


--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] snagged touch probe on eBay - anybody know how to interface ?

[Sarah Armstrong]

i'm looking at a renishaw MP10 IR probe
iv'e posted some snipps of what i am , ( or better think i am ) getting as
an output

https://www.dropbox.com/sh/kwjxef6k4ptbhhb/AABdCLSjzypH574XoXJv_it3a?dl=0

it looks to me all it's doing is giving a constant mark space ratio , and
stopping on a probe hit  , surely it cant be that simple , or could it
i must admit at this point , i have not dug anydeeper as yet , or really
looked to see if i can find any data , but it's very obvious everything
stops on a hit

i was expecting some data or id and say battery condition , but just early
days
found i had a better signal using a 1838 ir that a qse159 but i may need
some better signal conditioning at the moment




On 28 September 2016 at 02:01, Gene Heskett  wrote:

> On Tuesday 27 September 2016 13:19:49 Jon Elson wrote:
>
> > On 09/27/2016 10:36 AM, Gene Heskett wrote:
> > > On Tuesday 27 September 2016 09:29:47 Ken Strauss wrote:
> > >> The discussion has been regarding the optically coupled probe. Has
> > >> anyone had success with the inductively coupled ones? Is it
> > >> feasible to make your own IMM module?
> > >
> > > I'd have to say you could, but these machines are generators of a
> > > large amount of magnetic noise. So I'd say not magnetically coupled,
> > > but radio coupled just to get above the frequency of the machines
> > > own noises.
> >
> > Yes, it seems Blum and Renishaw have moved to RF
> > communication, and seem to be in the 800 MHz to GHz bands!
> > I guess the problem is no matter where you put the IR
> > transceiver, somebody would end up putting an obstruction in
> > the way, and then you break the probe tip. I did rig the
> > loss of signal watchdog to stop the machine, and it seems
> > like a really good idea.
> >
> > Jon
> >
> We are in what I've been known to call violent agreement on that point
> Jon. I recall now, about 2 years back, Dee  buzzed me on the intercom
> thats part of our basket of wireless phones here, in a total panic,
> wanted fresh batteries in the remote as she was missing one of her fav
> game shows.  On my arrival with a pocketfull of batteries, the first
> thing I noticed was she had set something in front of the tv's IR
> window.  I reached over and handed it back to her & told her to try it
> again.  Of course it worked, much to her amazement, it had never been
> grokked that if she couldn't see that window, she didn't have a remote.
> OTOH, she had cataracts removed from both eyes since, and its entirely
> possible she could not see that little black square in the bottom of the
> black face frame.  She sure can now.
> >
> >
> > --
> > ___
> > 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 
>
> 
> --
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>



-- 

The information contained in this message is confidential and is intended
for the addressee only. If you have received this message in error or there
are any problems please notify the originator immediately. The unauthorised
use, disclosure, copying or alteration of this message is strictly
forbidden. This mail and any attachments have been scanned for viruses
prior to leaving the RcTechnix network. RcTechnix will not be liable for
direct, special, indirect or consequential damages arising from alteration
of the contents of this message by a third party or as a result of any
virus being passed on.

RcTechnix reserves the right to monitor and record e-mail messages being
sent to and from this address for the purposes of investigating or
detecting any unauthorised use of its system and ensuring effective
operation.

(c) RcTechnix
--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users