Re: [Emc-users] "Black BeagleBoard: available at $45" (was LinuxCNC would be very nice on the UDOO board.)

[Jon Elson]

Jon Elson wrote:
> Ralph Stirling wrote:
>   
>> One show stopper I found with the RPi was the high probability
>> of SD card corruption during power cycles (power failure without
>> full OS shutdown first).  Is the BeagleBone Black immune to this
>> problem?  The quoted ten second shutdown is good, but I'd like
>> to have the flash image survive unforeseen power failures.  If
>> it handles those fine, then I'm ready to get excited about it.
>>   
>> 
> This is a problem that has been around for some time.  it is almost 
> certainly more
> complicated than at first glance.
Oh, one other data point.  I have been using SanDisk 4gb SD cards, with 
relatively
few problems.  I have a Dane-Elec disk that croaked after minimal usage, and
doesn't seem to be repaired by fsck.  A number of other users also 
report good
luck with official SanDisk and some other name-brand cards, and a lot
of problems with off-brand or cards that might be counterfeit, bought on
eBay or whatever.  I bought many of my SD cards at MicroCenter.

If the power is shut off while the SD card is in the middle of a 
wear-leveling
swap, I don't know how it will handle that.  There are methods of keeping
all data in a secure condition until the swap is complete, but some cheap
controllers may not have firmware that does this right.  If the 
wear-leveling
remap table gets corrupted, the card is generally scrap, there's no way to
reformat it.

Jon

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[Gene Heskett]

On Tuesday 23 April 2013 20:12:51 Steve Blackmore did opine:

> On Tue, 23 Apr 2013 11:57:24 -0400, you wrote:
> >What is the path tolerance thing, G64? In any event, it has no effect,
> >you can tell it to use a .5000" tolerance but you cannot get past the
> >brick wall that is the error checker for g2-3.
> 
> I don't think G64 was ever intended to override arc end position
> tolerances, but we ascertained that
> 
> "G64 - without P means to keep the best speed possible, no matter how
> far away from the programmed point you end up."
> 
> simply doesn't work anyway.
> 
> For arcs the manual says
> 
> "When programming arcs an error due to rounding can result from using a
> precision of less than 4 decimal places (0.) for inch and less than
> 3 decimal places (0.000) for millimetres."
> 
> A better approach to that would have been a user definable precision
> somewhere in the ini file.
> 
> If your machine doesn't have the resolution to do three or four decimal
> places it's a bit unreasonable to have to program them.
> 
> Steve Blackmore
> --

A thou I can do with careful homing, but a .0001" is trying to wet the rope 
way too high. :)

Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
My web page:  is up!
My views 

So far we've managed to avoid turning Perl into APL.  :-)
-- Larry Wall in <199702251904.laa28...@wall.org>
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
 law-abiding citizens.

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[Steve Blackmore]

On Tue, 23 Apr 2013 11:57:24 -0400, you wrote:

>What is the path tolerance thing, G64? In any event, it has no effect, you 
>can tell it to use a .5000" tolerance but you cannot get past the brick 
>wall that is the error checker for g2-3. 

I don't think G64 was ever intended to override arc end position
tolerances, but we ascertained that

"G64 - without P means to keep the best speed possible, no matter how
far away from the programmed point you end up." 

simply doesn't work anyway.

For arcs the manual says

"When programming arcs an error due to rounding can result from using a
precision of less than 4 decimal places (0.) for inch and less than
3 decimal places (0.000) for millimetres."

A better approach to that would have been a user definable precision
somewhere in the ini file. 

If your machine doesn't have the resolution to do three or four decimal
places it's a bit unreasonable to have to program them.

Steve Blackmore
--

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[Andrew]

2013/4/23 Michał Geszkiewicz
https://mail.google.com/mail/u/0/?view=cm&fs=1&tf=1&to=mic...@wp.pl>
>

> I think i fixed teleop jogging on ja3, please try it.
>
> Hi,
Thanks a lot!
I just did pull and make, there was a change in motmod. I'll try it
tomorrow on the machine.
FYI a warning appeared (or it was before, I can't say):

make[1]: Leaving directory `/usr/src/linux-headers-2.6.32-122-rtai'
/home/pkm/linuxcnc-ja3/src/emc/motion/control.c: In function
‘set_operating_mode’:
/home/pkm/linuxcnc-ja3/src/emc/motion/control.c:788: warning: unused
variable ‘joint_data’

Andrew
--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[John Kasunich]

On Tue, Apr 23, 2013, at 04:21 PM, Gene Heskett wrote:
> On Tuesday 23 April 2013 16:16:48 Stuart Stevenson did opine:
> 
> > Gene - The parabole function may not work but I see the hyperbole button
> > is working. :)
> 
> ROTFLMAO!  Thanks for the flowers Stuart!
> 
> But, really, does it HAVE to be that picky?  Seems to me that .0001" or 
> .005mm really ought to be close enough for the girls we go with. :)
>  

In my experience 0.0001 is fine.  When doing arcs I use 4 digits and
the only time LCNC has ever complained has been when I did
something like writing X1.2435 when I meant X1.2345.

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

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[Michał Geszkiewicz]

W dniu 23.04.2013 14:47, Andrew pisze:
> 2013/4/23 Yishin Li 
>
>> AFAIK, the world mode jogging had been fixed on araisrobo's ja3 branch.
>> This is the github repository:
>> https://github.com/araisrobo/linuxcnc/commits/feature/usb-rigid-tap
>> (We are working on "feature/usb-rigid-tap" branch.)
>> This is the guide to pull and build it:
>> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?ArtekBranch
>>
>> Thanks Yishin!
> It says 12.04, I'm still on 10.04. Does it matter?
> Also, what about S-curves? They might be very useful for a fast delta-robot.
>
> 2013/4/23 Michał Geszkiewicz 
>
>> I know where with world jogging is. I'll fix it on the evening.
>>
> Great, Michał, thanks!
> Please let me know.
Hi,

I think i fixed teleop jogging on ja3, please try it.

Michael

> BTW I could run some tests on joints_axes3, just tell me what needs to be
> tested.
>
> Andrew
> --
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[Gene Heskett]

On Tuesday 23 April 2013 16:16:48 Stuart Stevenson did opine:

> Gene - The parabole function may not work but I see the hyperbole button
> is working. :)

ROTFLMAO!  Thanks for the flowers Stuart!

But, really, does it HAVE to be that picky?  Seems to me that .0001" or 
.005mm really ought to be close enough for the girls we go with. :)
 
> On Tue, Apr 23, 2013 at 10:57 AM, Gene Heskett  wrote:
> > On Tuesday 23 April 2013 11:30:19 andy pugh did opine:
> > > On 23 April 2013 15:57, Gene Heskett  wrote:
> > > > The biggest confusion to me yet is that even though
> > > > the machine is in absolute movement mode, G2/3 moves are still
> > > > made relative to where it starts from,
> > > 
> > > See G91.1
> > > http://www.linuxcnc.org/docs/html/gcode/gcode.html#sec:G90_1-G91_1
> > 
> > I did look at that Andy, and did do one move successfully with G90.1
> > in effect.  But when I tried to do the next pass, cutting a little
> > more off incrementally, because it was all from only one side of the
> > ellipse, it all went to hell with the error message saying I wasn't
> > off a thou, but 70+ percent of the radii for a 3 thou change in the
> > end point location.
> > 
> > I finally gave up and went to the discouraged r mode, which worked
> > perfectly once I had figured out how to pre-calc the r parts.
> > 
> > The I/J/K mode may be the preferred mode, but if not accurate out to
> > about 16 or 17 decimal places, it prefers to throw essentially
> > meaningless errors at you instead of carving metal, or even video
> > screen pixels.
> > 
> > What is the path tolerance thing, G64? In any event, it has no effect,
> > you can tell it to use a .5000" tolerance but you cannot get past the
> > brick wall that is the error checker for g2-3.  Unless you are trying
> > to carve a new mirror for the Hubble, I see zero practical use for
> > sub-angstrom level accuracy checking.  Besides, that would not be an
> > arc, but parabola, something I don't know that linuxcnc can do all as
> > a canned function.
> > 
> > Simply put, you cannot /see/ the error because it won't even draw it
> > in the backplot.
> > 
> > Put even simpler, its a pain in the ass.  The absolute worst, hands
> > down by a very wide margin, command to "get right" in linuxcnc.  When
> > you finally make it work, you've got enough time in 12 lines of code
> > out of 300 to have bought lottery tickets and won $2, for only a
> > $10,000 investment.  Whoopy Ding, I won!  NOT...
> > 
> > Cheers, Gene
> > --
> > 
> > "There are four boxes to be used in defense of liberty:
> >  soap, ballot, jury, and ammo. Please use in that order."
> > 
> > -Ed Howdershelt (Author)
> > My web page:  is up!
> > My views
> > 
> > Results are not typical.
> > A pen in the hand of this president is far more
> > dangerous than 200 million guns in the hands of
> > 
> >  law-abiding citizens.
> > 
> > --
> >  Try New Relic Now & We'll Send You this Cool Shirt
> > New Relic is the only SaaS-based application performance monitoring
> > service that delivers powerful full stack analytics. Optimize and
> > monitor your browser, app, & servers with just a few lines of code.
> > Try New Relic and get this awesome Nerd Life shirt!
> > http://p.sf.net/sfu/newrelic_d2d_apr
> > ___
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users


Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
My web page:  is up!
My views 

BOFH excuse #421:

Domain controller not responding
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
 law-abiding citizens.

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machine documentation: software, best practices

[Gene Heskett]

On Tuesday 23 April 2013 15:59:53 Kent A. Reed did opine:

> On 4/22/2013 10:40 PM, Gene Heskett wrote:
> > On Monday 22 April 2013 22:25:37 Kent A. Reed did opine:
> >> On 4/16/2013 7:23 PM, Gene Heskett wrote:
> >>>  >>> configurations
> >>> 
> >>> There was a thingy that I assume worked, I installed it last fall,
> >>> but it had one fatal flaw.  It tried to make the whole diagram fit
> >>> on a single sheet of paper, when, in order to have been able to
> >>> read it with a magnifying glass, it would have had to be done in
> >>> multi-page poster style that would have likely used 54 to 100+
> >>> sheets of paper to be taped together before the text in one of the
> >>> teeny little logic boxes would have been big enough to read.
> >>> <...>
> >>> 
> >>> Looking on the lathes box, I find a ~/gene/src/RockHopper directory
> >>> that looks like one of those usual suspects, was that it?
> >>> 
> >>> Cheers, Gene
> >> 
> >> Sorry for coming to the party a week late. I've been very distracted
> >> lately.
> > 
> > Your lady's health?  My sympathies.  I hope she is better now.
> 
> No.
> 
> >> Me thinks you exaggerate the problem just a bit. The Rockhopper
> >> server creates a diagram in SVG. One could save* the diagram to file
> >> from the browser, open the saved file in Inkscape,
> > 
> > I didn't know it was svg.  Hitting ctl+ in the browser did not magnify
> > it enough to be useful so I assume it was spitting out postscript at
> > 72 dpi.
> 
> Unfortunately, the svg rendering engine in Firefox and Chrome doesn't
> support the magnification trick. I don't know if it ever will. There was
> a lot to like about Adobe's SVG Viewer plugin but Adobe killed the
> product 4 years ago.
> 
> >> and either go through
> >> machinations to print it in tiles
> > 
> > I like that idea, and will check it out, but now it will likely be
> > later next week as we're headed to NYS over the weekend, Thursday I'm
> > told.
> 
> Well, I gave alternatives because I think the required "machinations"
> are clumsy. I forget the search term I used to find instructions out
> there in Internet land, some mix of poster and svg I suppose.
> 
> > <...>
> > 
> >> you could modify the Rockhopper Python script
> >> so that in addition to creating the SVG file Graphviz would also
> >> create and save a PDF file for further use.
> > 
> > That would be useful only if the PS engine can be convinced to output
> > a 2400 dpi image before its compressed from the 2+ gigabytes of raw
> > data that would generate.  72 dpi won't cut it.
> 
> You're thinking of an uncompressed bitmap. The diagram is line art that
> is mostly white space. Compression techniques deal with that. As well,
> cranking the resolution up to 2400 dpi shouldn't be necessary.
> 
> For the example configuration file configs/sim/axis/axis.ini, the size
> of the halgraph.svg file saved from the browser is of order 50 KB. Open
> it in Inkscape and "save as" a PDF file with default settings (I forget
> exactly; something less than 100 dpi IIRC). The PDF file size is less
> than 25 KB. Open the PDF file in the Gnome Document Viewer. Zoom to 400
> percent and wait a while for the renderer to catch up. Result looks
> great. Unfortunately, the Document Viewer doesn't know how to print out
> tiled pages...let me rephrase that...I don't know how to make the
> Document Viewer print out tiled pages and as far as I know my printer
> driver doesn't support that function either.
> 
> Hope your trip goes well. To quote the old Hill Street Blues show,
> "let's be careful out there."
> 
> Regards,
> Kent
> 
Well, they took care of the recalls, but of course they have to go over it 
with their fine toothed comb, last time it was a leaking water pump they 
could replace for about $400, (no leak visible, and I haven't added a drop 
of coolant to it yet in 5 moths since that pronouncement of doom) this time 
the front brake pads are down to about 3mm thick.  The disks of course are 
rusty, so they want to replace it all, and then do the rear end another 10k 
down the log.

Gotta find something to help pay for the recalls don'tcha know. :(

I'll look at the pads when I get back, ATM its stopping on a dime, very 
quietly, and giving me back 9 cents change.  I asked if 44k miles was about 
right to do that and the mechanic said no, its usually in the later 30k 
range.  Good drivers are off the loud pedal far enough back they don't have 
to do stoppies at the next traffic light.  And now State Farm wants to put 
a computer tattle tale on it, claiming that will lower our rates.  Sure, 
and that one time a year that I give it a whole cup of good gulf to get out 
of the way of some idiot yakking on a cell phone else get hit, will cost me 
$200 year?  I think not.  That computer thingy might work 99% of the time, 
but it can't see the reason I did that, so screw'em.

Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershel

Re: [Emc-users] Machine documentation: software, best practices

[Gene Heskett]

On Tuesday 23 April 2013 15:55:19 Kent A. Reed did opine:

> On 4/22/2013 6:49 PM, andy pugh wrote:
[...]> > 
> > We have 46,000 labels to calibrate and 5000 pages of these diagrams.
> 
> I don't know which is worse---dealing with the diagrams themselves or
> trying to grok the designers' intent. I'd say "poor you" but maybe being
> able to do this guarantees a job for life, akin to being able to program
> in COBOL :-)
> 
> Regards,
> Kent

Thanks Kent, fresh back from getting a couple recalls on our Toy taken care 
of, and needed a good belly laugh, and that was it. :)

Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
My web page:  is up!
My views 

Truth never comes into the world but like a bastard, to the ignominy
of him that brought her birth.
-- Milton
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
 law-abiding citizens.

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Two hal+gladevcp questions

[Viesturs Lācis]

2013/4/23 John Kasunich 
>
>
> "halcmd show" will show pin names as well.  You can also
> use halcmd in interactive mode;  type "halcmd -f" at the
> shell prompt, and halcmd will display a "halcmd: " prompt.
> You can then use commands like "show pin", complete
> with wildcards, tab completion, and all the goodies.
>

I already found halcmd man page and tried also "halcmd show pin" command.
Here is what I get in terminal:

vie@vie:~$ halrun -f /home/vie/linuxcnc/configs/counter/counter.hal

(gladevcp:1516): libglade-WARNING **: Expected .  Got
.

(gladevcp:1516): libglade-WARNING **: did not finish in PARSER_FINISH state
 GLADE VCP INFO:Not a libglade project, trying to load as a GTK
builder project
/home/vie/linuxcnc/configs/counter/counter.hal:21: parameter or pin
'gladevcp.length' not found
Traceback (most recent call last):
  File "/usr/lib/pymodules/python2.6/gladevcp/led.py", line 105, in expose
cr.stroke_preserve()
KeyboardInterrupt

vie@vie:~$ halcmd show pin
RTAPI: ERROR: could not open shared memory (errno=2)
Segmentation fault
vie@vie:~$


At first I started the HAL+Glade. The output shows that the meter widget
pin also is not found.
But the glade panel shows up and everything else seems fine.
Then, with glade panel open, I ran the second command of halcmd and
received segmentation fault error.

I would appreciate any hints, why are the glade panel pins not found, when
they are linked in HAL file and what does the segmentation fault mean.
I am running LinuxCNC in VirtualBox, if that is important. I would gladly
do whatever tests required to help identify the cause.

-- 
Viesturs

If you can't fix it, you don't own it.
http://www.ifixit.com/Manifesto
--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[Stuart Stevenson]

Gene - The parabole function may not work but I see the hyperbole button is
working. :)


On Tue, Apr 23, 2013 at 10:57 AM, Gene Heskett  wrote:

> On Tuesday 23 April 2013 11:30:19 andy pugh did opine:
>
> > On 23 April 2013 15:57, Gene Heskett  wrote:
> > > The biggest confusion to me yet is that even though
> > > the machine is in absolute movement mode, G2/3 moves are still made
> > > relative to where it starts from,
> >
> > See G91.1
> > http://www.linuxcnc.org/docs/html/gcode/gcode.html#sec:G90_1-G91_1
>
> I did look at that Andy, and did do one move successfully with G90.1 in
> effect.  But when I tried to do the next pass, cutting a little more off
> incrementally, because it was all from only one side of the ellipse, it all
> went to hell with the error message saying I wasn't off a thou, but 70+
> percent of the radii for a 3 thou change in the end point location.
>
> I finally gave up and went to the discouraged r mode, which worked
> perfectly once I had figured out how to pre-calc the r parts.
>
> The I/J/K mode may be the preferred mode, but if not accurate out to about
> 16 or 17 decimal places, it prefers to throw essentially meaningless errors
> at you instead of carving metal, or even video screen pixels.
>
> What is the path tolerance thing, G64? In any event, it has no effect, you
> can tell it to use a .5000" tolerance but you cannot get past the brick
> wall that is the error checker for g2-3.  Unless you are trying to carve a
> new mirror for the Hubble, I see zero practical use for sub-angstrom level
> accuracy checking.  Besides, that would not be an arc, but parabola,
> something I don't know that linuxcnc can do all as a canned function.
>
> Simply put, you cannot /see/ the error because it won't even draw it in the
> backplot.
>
> Put even simpler, its a pain in the ass.  The absolute worst, hands down by
> a very wide margin, command to "get right" in linuxcnc.  When you finally
> make it work, you've got enough time in 12 lines of code out of 300 to have
> bought lottery tickets and won $2, for only a $10,000 investment.  Whoopy
> Ding, I won!  NOT...
>
> Cheers, Gene
> --
> "There are four boxes to be used in defense of liberty:
>  soap, ballot, jury, and ammo. Please use in that order."
> -Ed Howdershelt (Author)
> My web page:  is up!
> My views
> 
> Results are not typical.
> A pen in the hand of this president is far more
> dangerous than 200 million guns in the hands of
>  law-abiding citizens.
>
>
> --
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>



-- 
dos centavos
--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] LinuxCNC would be very nice on the UDOO board.

[Jon Elson]

Matt Shaver wrote:
> On Mon, 22 Apr 2013 21:19:19 -0500
> Jon Elson  wrote:
>
>   
>> I sent one of these EPP
>> emulator/converters to Dale Grover, I think it ended up with Matt
>> Shaver.
>> 
>
> Nope, I think Dale's still got it. I was discouraged by the lack of
> RTAI for ARM, but that seems to be a non-issue with Xenomai available,
> plus all the breakthroughs in running things in user space.
>
> Well, if anyone wants one of these, I could make some more.
>   
But, it is Beagle Board (not Bone) only.  I'm getting fairly serious 
about getting
either a Pi or a Bone Black, but am waiting to see which way people go with
this.

Jon

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] "Black BeagleBoard: available at $45" (was LinuxCNC would be very nice on the UDOO board.)

[Jon Elson]

Ralph Stirling wrote:
> One show stopper I found with the RPi was the high probability
> of SD card corruption during power cycles (power failure without
> full OS shutdown first).  Is the BeagleBone Black immune to this
> problem?  The quoted ten second shutdown is good, but I'd like
> to have the flash image survive unforeseen power failures.  If
> it handles those fine, then I'm ready to get excited about it.
>   
This is a problem that has been around for some time.  it is almost 
certainly more
complicated than at first glance.  it may be some kind of interaction 
between the
writes to the SD card and the wear leveling algorithm in the card.  
Apparently,
some people have been running for years with no trouble, and others have
SD cards become completely unusable within weeks.  i have a couple units
in the field with the original Beagle Board, and have had some minor trouble
with corrupted SD cards, but my fix is to plug them into a desktop PC 
with card reader
and do an fsck on them.  That has always brought them back.  Our units run
in an industrial equipment rack and are powered on and off willy-nilly by
the users, and nobody in the field has reported a dead one, yet.

If you read the R Pi and Beagle discussion lists, there is a LOT of
traffic about this problem.  I have been using ext3 and noatime settings
on the fstab, which seem to help.  some people report ext4 has some
kind of serious interaction that kills SD cards (it isn't supposed to...)

Jon

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual

[Sebastian Kuzminsky]

On 04/23/2013 09:37 AM, Kent A. Reed wrote:
> That issue aside, I just did what I should have done yesterday and
> compared the HTML and PDF versions of the Integrator  manual. The
> subheadings Rudy and I questioned are present in the HTML version
> (http://www.linuxcnc.org/docs/2.5/html/config/ini_config.html) and not
> in the PDF version (Integrator Manual V2.5, 2013-03-26).

I recently added the "Servo" and "Stepper" subsections to the 
"[AXIS_]" section based on your suggestion.  If you fetch a newer 
pdf you'll find the new subsections there too.


-- 
Sebastian Kuzminsky

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machine documentation: software, best practices

[Kent A. Reed]

On 4/22/2013 10:40 PM, Gene Heskett wrote:
> On Monday 22 April 2013 22:25:37 Kent A. Reed did opine:
>
>> On 4/16/2013 7:23 PM, Gene Heskett wrote:
>>> >> configurations
>>>
>>> There was a thingy that I assume worked, I installed it last fall, but
>>> it had one fatal flaw.  It tried to make the whole diagram fit on a
>>> single sheet of paper, when, in order to have been able to read it
>>> with a magnifying glass, it would have had to be done in multi-page
>>> poster style that would have likely used 54 to 100+ sheets of paper
>>> to be taped together before the text in one of the teeny little logic
>>> boxes would have been big enough to read.
>>> <...>
>>>
>>> Looking on the lathes box, I find a ~/gene/src/RockHopper directory
>>> that looks like one of those usual suspects, was that it?
>>>
>>> Cheers, Gene
>> Sorry for coming to the party a week late. I've been very distracted
>> lately.
> Your lady's health?  My sympathies.  I hope she is better now.

No.

>   
>> Me thinks you exaggerate the problem just a bit. The Rockhopper server
>> creates a diagram in SVG. One could save* the diagram to file from the
>> browser, open the saved file in Inkscape,
> I didn't know it was svg.  Hitting ctl+ in the browser did not magnify it
> enough to be useful so I assume it was spitting out postscript at 72 dpi.

Unfortunately, the svg rendering engine in Firefox and Chrome doesn't 
support the magnification trick. I don't know if it ever will. There was 
a lot to like about Adobe's SVG Viewer plugin but Adobe killed the 
product 4 years ago.

>
>> and either go through
>> machinations to print it in tiles
> I like that idea, and will check it out, but now it will likely be later
> next week as we're headed to NYS over the weekend, Thursday I'm told.

Well, I gave alternatives because I think the required "machinations" 
are clumsy. I forget the search term I used to find instructions out 
there in Internet land, some mix of poster and svg I suppose.

> <...>
>> you could modify the Rockhopper Python script
>> so that in addition to creating the SVG file Graphviz would also create
>> and save a PDF file for further use.
> That would be useful only if the PS engine can be convinced to output a
> 2400 dpi image before its compressed from the 2+ gigabytes of raw data that
> would generate.  72 dpi won't cut it.
>

You're thinking of an uncompressed bitmap. The diagram is line art that 
is mostly white space. Compression techniques deal with that. As well, 
cranking the resolution up to 2400 dpi shouldn't be necessary.

For the example configuration file configs/sim/axis/axis.ini, the size 
of the halgraph.svg file saved from the browser is of order 50 KB. Open 
it in Inkscape and "save as" a PDF file with default settings (I forget 
exactly; something less than 100 dpi IIRC). The PDF file size is less 
than 25 KB. Open the PDF file in the Gnome Document Viewer. Zoom to 400 
percent and wait a while for the renderer to catch up. Result looks 
great. Unfortunately, the Document Viewer doesn't know how to print out 
tiled pages...let me rephrase that...I don't know how to make the 
Document Viewer print out tiled pages and as far as I know my printer 
driver doesn't support that function either.

Hope your trip goes well. To quote the old Hill Street Blues show, 
"let's be careful out there."

Regards,
Kent


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[Gene Heskett]

On Tuesday 23 April 2013 11:30:19 andy pugh did opine:

> On 23 April 2013 15:57, Gene Heskett  wrote:
> > The biggest confusion to me yet is that even though
> > the machine is in absolute movement mode, G2/3 moves are still made
> > relative to where it starts from,
> 
> See G91.1
> http://www.linuxcnc.org/docs/html/gcode/gcode.html#sec:G90_1-G91_1

I did look at that Andy, and did do one move successfully with G90.1 in 
effect.  But when I tried to do the next pass, cutting a little more off 
incrementally, because it was all from only one side of the ellipse, it all 
went to hell with the error message saying I wasn't off a thou, but 70+ 
percent of the radii for a 3 thou change in the end point location.

I finally gave up and went to the discouraged r mode, which worked 
perfectly once I had figured out how to pre-calc the r parts.

The I/J/K mode may be the preferred mode, but if not accurate out to about 
16 or 17 decimal places, it prefers to throw essentially meaningless errors 
at you instead of carving metal, or even video screen pixels.

What is the path tolerance thing, G64? In any event, it has no effect, you 
can tell it to use a .5000" tolerance but you cannot get past the brick 
wall that is the error checker for g2-3.  Unless you are trying to carve a 
new mirror for the Hubble, I see zero practical use for sub-angstrom level 
accuracy checking.  Besides, that would not be an arc, but parabola, 
something I don't know that linuxcnc can do all as a canned function.

Simply put, you cannot /see/ the error because it won't even draw it in the 
backplot.

Put even simpler, its a pain in the ass.  The absolute worst, hands down by 
a very wide margin, command to "get right" in linuxcnc.  When you finally 
make it work, you've got enough time in 12 lines of code out of 300 to have 
bought lottery tickets and won $2, for only a $10,000 investment.  Whoopy 
Ding, I won!  NOT...
   
Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
My web page:  is up!
My views 

Results are not typical.
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
 law-abiding citizens.

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Machine documentation: software, best practices

[Kent A. Reed]

On 4/22/2013 6:49 PM, andy pugh wrote:
> On 22 April 2013 19:33, Kent A. Reed  wrote:
>
>>   We both suffer from the problem
>> that the size and complexity of a HAL diagram is potentially unbounded.
> Not just you.
>
> My job basically revolves around peering at Simulink diagrams (the
> code _is_ the documentation) trying to work out what the heck Bosch /
> Delphi / Ford / Volvo coders were trying to do.
> There are many pages in the docs I am given that can not be read even
> when zoomed to the limits of PDF.
>
> We have 46,000 labels to calibrate and 5000 pages of these diagrams.
>

I don't know which is worse---dealing with the diagrams themselves or 
trying to grok the designers' intent. I'd say "poor you" but maybe being 
able to do this guarantees a job for life, akin to being able to program 
in COBOL :-)

Regards,
Kent


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual

[Kent A. Reed]

On 4/23/2013 7:34 AM, John Thornton wrote:
> The biggest problem is none of the items after the warning are actually
> used by LinuxCNC. When the subheading was there it implied that LinuxCNC
> actually used them somehow. IMHO each one of those entries really
> belongs with what ever component might use them, but I have no way of
> knowing which component might read them from the ini file. I think I
> have at least narrowed down which entries are used by Axis and separated
> them from the rest.

John:

We obviously have different viewpoints. I think of the collection of 
components as being part of LinuxCNC so I find the warning statement 
confusing.

That issue aside, I just did what I should have done yesterday and 
compared the HTML and PDF versions of the Integrator  manual. The 
subheadings Rudy and I questioned are present in the HTML version 
(http://www.linuxcnc.org/docs/2.5/html/config/ini_config.html) and not 
in the PDF version (Integrator Manual V2.5, 2013-03-26).

Regards,
Kent


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] LinuxCNC would be very nice on the UDOO board.

[Charles Steinkuehler]

On 4/22/2013 11:40 AM, Matt Shaver wrote:
> On Sun, 21 Apr 2013 19:30:23 +0200
> Michael Haberler  wrote:
>  
>> We'll have a ready-to-go SD card image with xenomai, all prerequisite
>> packages and linuxcnc installed in a week or two; that will include
>> Charles' PRU stepgen and Sergey's/GP Orcullo's emcweb, and should
>> reduce the guru requirements to get things going
> 
> So, I'm ordering a new Beagle Bone Black this morning & I was wondering
> what you guys are doing for a field I/O interface for this computer. Is
> there a cape I should get? Or are we at the perfboard and wire-wrap
> wire stage? :)

The BeBoPr board is available now and has support for both the Pololu
drivers common in the 3D printer world and a header to drive out-board
stepper drivers with a more "oomph".  From the BeBoPr maunal:

  J5 – all stepper signals are present on this connector. It connects
  directly to the 15 pin sub-D connector of a TB6560-4V5 3A stepper
  driver board sold on e-Bay.

There's also the Replicape, but AFAIK it's on-board stepper driver only
(Pololu class, but using TI driver chips).

There's nothing I'm aware of yet that works for analog servos, but it
should be possible, at least for a few channels (the CPU has 3 hardware
encoder inputs, and the PRU could support more in "software" at lower
pulse rates).

-- 
Charles Steinkuehler
char...@steinkuehler.net

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[andy pugh]

On 23 April 2013 15:57, Gene Heskett  wrote:
> The biggest confusion to me yet is that even though
> the machine is in absolute movement mode, G2/3 moves are still made
> relative to where it starts from,

See G91.1
http://www.linuxcnc.org/docs/html/gcode/gcode.html#sec:G90_1-G91_1

-- 
atp
If you can't fix it, you don't own it.
http://www.ifixit.com/Manifesto

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[Gene Heskett]

On Tuesday 23 April 2013 10:35:41 John Thornton did opine:

> At one time there were 4 manuals and 2 of them contained almost the same
> content. After years of work I put the stuff needed for a newbee to get
> started into the Getting Started Manual pdf and the stuff needed to USE
> LinuxCNC in the User Manual pdf and the stuff needed to integrate
> LinuxCNC into the Integrators Manual pdf that left the HAL tutorial
> which fluctuated a bit but is/was for using HAL standalone and last but
> not least the developer information which is in the Developer Manual.
> 
> And then there is the secret man pages that only Linux Guru's know about
> and thanks to Jeff's magic are included in the HTML documents.
> 
> On the bright side the entire manual is in HTML from start to finish
> including the secret man pages this way you get the manual any way you
> want.
> 
> What did I miss?
> 
> John

IMO not much John, Thanks.

I've been using FF to check things in the newest html as I work, but whats 
missing is stuff I might be doing that isn't covered so I've invented the 
lugs nuts I need. No big deal, we all have to make those parts. :)

There is some ambiguity in language, but that seems unavoidable when trying 
to be concise, and a few re-reads generally, or a question on IRC, will 
clear it up for me.  The biggest confusion to me yet is that even though 
the machine is in absolute movement mode, G2/3 moves are still made 
relative to where it starts from, and that is a translation in concept that 
I have a hard time wrapping this aged brain around.

It would help me if there were some examples of the variable pre use 
calculations needed to properly arrive at the correct i/j/k values needed 
to go from y=3.111 to y=5.09 with x fixed at 4.33 for a 180 rounded curve 
on the end of something.  I fought with that for several days while trying 
to do that cocking handle in brass 2 weeks back.

Cheers, Gene
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
My web page:  is up!
My views 

You're not an alcoholic unless you go to the meetings.
A pen in the hand of this president is far more
dangerous than 200 million guns in the hands of
 law-abiding citizens.

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[Peter C. Wallace]

On Tue, 23 Apr 2013, Andrew wrote:

> Date: Tue, 23 Apr 2013 12:56:30 +0300
> From: Andrew 
> Reply-To: "Enhanced Machine Controller (EMC)"
> 
> To: "Enhanced Machine Controller (EMC)" 
> Subject: [Emc-users]  Joints_axes3 branch
> 
> Hello gentlemen,
>
> I'm trying joints_axes3 for my delta robot. It has Yaskawa Sigma II servo
> motors with analog control and MESA 7i43+7i48. Working config is converted
> from 2.5.2 to ja3 (attached).
> I should say that ja3 is much better than 2 years ago when I tried it.
> Generally the machine works just like on 2.5.2, but the error message
> appears after machine is on and keeps showing later:
>
> hm2/hm2_7i43.0: Encoder 0: quadrature count error
>
> Though nothing seems to happen, I'm afraid that the encoder might lose
> impulses or smth. What's the problem? Nothing similar on 2.5.2 with exactly
> same settings, so the hardware should be ok.
>
> Thanks,
> Andrew
>


A HostMot2 encoder error means that the encoder counter sensed that the A and 
B input changed at the same time (this is an illegal condition for a 
quadrature encoder). This can be because of a faulty encoder, or more likely 
electrical noise. I would try rearranging power/encoder cabling, use 
differerential signaling if you are not already doing that, checking that the 
7I48 is jumpered properly for the encoder interface mode (TTL or 
differential), enabling the encoder filter, and if nothing else helps, slow 
the encoder digital filter as low as you can.



Peter Wallace
Mesa Electronics


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[Andrew]

2013/4/23 Yishin Li 

> AFAIK, the world mode jogging had been fixed on araisrobo's ja3 branch.
> This is the github repository:
> https://github.com/araisrobo/linuxcnc/commits/feature/usb-rigid-tap
> (We are working on "feature/usb-rigid-tap" branch.)
> This is the guide to pull and build it:
> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?ArtekBranch
>
> Thanks Yishin!
It says 12.04, I'm still on 10.04. Does it matter?
Also, what about S-curves? They might be very useful for a fast delta-robot.

2013/4/23 Michał Geszkiewicz 

> I know where with world jogging is. I'll fix it on the evening.
>

Great, Michał, thanks!
Please let me know.
BTW I could run some tests on joints_axes3, just tell me what needs to be
tested.

Andrew
--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual - Why one Document

[John Thornton]

At one time there were 4 manuals and 2 of them contained almost the same 
content. After years of work I put the stuff needed for a newbee to get 
started into the Getting Started Manual pdf and the stuff needed to USE 
LinuxCNC in the User Manual pdf and the stuff needed to integrate 
LinuxCNC into the Integrators Manual pdf that left the HAL tutorial 
which fluctuated a bit but is/was for using HAL standalone and last but 
not least the developer information which is in the Developer Manual.

And then there is the secret man pages that only Linux Guru's know about 
and thanks to Jeff's magic are included in the HTML documents.

On the bright side the entire manual is in HTML from start to finish 
including the secret man pages this way you get the manual any way you want.

What did I miss?

John

On 4/22/2013 3:21 PM, Greg Bentzinger wrote:
> There are multible manuals - This is the integrators manual.
>   
> LCNC is whatever you choose to make it.
>   
> There are many configs which combine stepper and servo methods of control, be 
> it a servo running step/dir signals or otherwise.
>   
> A mill I have in progress uses servos via +/-10V signals for X & Y axis, and 
> stepper drive for Z & A (4th) axis, topped off with VFD feedback for the 
> spindle. All these will have to work togather and LCNC has the flexability to 
> make it happen. The info required to configure this does not need to be 
> spread all over.
>   
> If anything LCNC could benefit from more consolidation, but for the most part 
> LCNC continues to grow just as is.
> --
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Integrators Manual

[John Thornton]

The biggest problem is none of the items after the warning are actually 
used by LinuxCNC. When the subheading was there it implied that LinuxCNC 
actually used them somehow. IMHO each one of those entries really 
belongs with what ever component might use them, but I have no way of 
knowing which component might read them from the ini file. I think I 
have at least narrowed down which entries are used by Axis and separated 
them from the rest.

The old lead in paragraph suggested that in some way LinuxCNC used the 
entries which could lead to much confusion.

LinuxCNC is so flexible it is difficult to document everything 
especially the ini file which can contain anything and be used by 
anything. And the worst case is things can be placed in ini files by 
wizards and not even be used, talk about confusing.


On 4/22/2013 8:59 AM, Kent A. Reed wrote:
> On 4/22/2013 7:05 AM, John Thornton wrote:
>> Did you skip past the Warning?
>>
>>
>> On 4/22/2013 2:06 AM, Rudy du Preez wrote:
>>> I find pages 20 to 23 of the Integrators manual V2.5, 2013-03-26 very
>>> confusing:
>>>
>>> Page 20 is headed 3.2.10.1 Homing.
>>>
>>> It discusses HOME related parameters and the suddenly jumps into PID
>>> settings. Is that correct? At least a heading is missing.
> John:
>
> In the v2.4 docs, there was a subheading just above the line which has
> been turned into a warning.
>
> The subheading began
>
> ---snip---
> 4.2.9.2 Servo
>
> The following items are for servo-based systems and servo-like systems.
> This description assumes
> that the units of output from the PID component are volts.
>
> DEADBAND = 0.15 (HAL) How close is close enough the consider the
> motor in position."
> ---snip---
>
> The result is we have lost not only the subheading but also the slightly
> redundant information contained in the old lead sentence.
>
>>> Please read the paragraph on OUTPUT_OFFSET = 0.000 carefully and see if it
>>> makes sense with regards to units. How can the second value (offset) first
>>> be subtracted from the computed value (in volts) and the be divided by the
>>> first value (scale)?
>>>
>>> Lower down the equation is given which shows that the subtraction is done
>>> before scaling.
> The second sentence somewhat awkwardly repeats what the equation says.
> Further, it does not use the actual names of the parameters. Perhaps the
> sentence could be changed to say something like
>
> "The second value (output_offset, in volts) is subtracted from the
> computed output (in volts); the result is divided by the first value
> (output_scale, in true volts per DAC output volts)."
>
> Other parts of the paragraph would be adjusted to suit.
>
> My personal preference, however, would be to state the units for each
> parameter right up front
>
> "OUTPUT_SCALE = 1.000 [true volts per DAC output volts] -
> OUTPUT_OFFSET = 0.000 [volts] - These two values"
>
> and then drop them from the ensuing text.
>
> However, this ought to be done everywhere. I don't have the time to
> check every place a parameter is defined in the docs and propose an
> editorial change; my guess is you're pretty busy too:-)
>
> Note also that the next subheading in the document is also missing
> having been replaced in the same way by a boilerplate warning.
>
> ---snip---
> 4.2.9.3 Stepper
>
> The following items are Stepper related items.
> ---snip---
>
>>> Rudy
> Thanks, Rudy.
>
> Regards,
> Kent
>
>
> --
> Precog is a next-generation analytics platform capable of advanced
> analytics on semi-structured data. The platform includes APIs for building
> apps and a phenomenal toolset for data science. Developers can use
> our toolset for easy data analysis & visualization. Get a free account!
> http://www2.precog.com/precogplatform/slashdotnewsletter
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[Michał Geszkiewicz]

W dniu 23.04.2013 12:42, Andrew pisze:
> 2013/4/23 andy pugh 
>
>> On 23 April 2013 10:56, Andrew  wrote:
>>
>>>   hm2/hm2_7i43.0: Encoder 0: quadrature count error
>>>
>>> Though nothing seems to happen, I'm afraid that the encoder might lose
>>> impulses or smth. What's the problem? Nothing similar on 2.5.2 with
>> exactly
>>> same settings, so the hardware should be ok.
>> I get that with my machine (home-made encoder). it is a new error
>> message added by this commit:
>>
>> http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=commit;h=a3719bd5a9a6c64326a4d1ac9c32fe06385f07cf
>> And is based on a fault bit raised by the Mesa firmware.
>>
>> It doesn't appear to have been in the 2.5.2 release, so it seems that
>> the issue has always been there, but you just weren't being told.
>>
>> Thanks, Andy! I just exchanged connectors from servo drives 0 and 1 to
> 7i48, the message still appears on encoder 0. That might be 7i48 problem
> then. I will try to switch encoder 0 to another channel on 7i48.
>
> Another problem appeared: world mode always jogs full speed. If it was not
> for world mode limits, that could be painful.
>
> Andrew
Hi,

I know where with world jogging is. I'll fix it on the evening.

regards,
Michael


>
>   --
>> atp
>> If you can't fix it, you don't own it.
>> http://www.ifixit.com/Manifesto
>>
>>
>> --
>> Try New Relic Now & We'll Send You this Cool Shirt
>> New Relic is the only SaaS-based application performance monitoring service
>> that delivers powerful full stack analytics. Optimize and monitor your
>> browser, app, & servers with just a few lines of code. Try New Relic
>> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>
> --
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
>


--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[Yishin Li]

On Tue, Apr 23, 2013 at 6:42 PM, Andrew wrote:

>
> Another problem appeared: world mode always jogs full speed. If it was not
> for world mode limits, that could be painful.
>

AFAIK, the world mode jogging had been fixed on araisrobo's ja3 branch.
This is the github repository:
https://github.com/araisrobo/linuxcnc/commits/feature/usb-rigid-tap
(We are working on "feature/usb-rigid-tap" branch.)
This is the guide to pull and build it:
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?ArtekBranch

Yishin
--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[Andrew]

2013/4/23 andy pugh 

> On 23 April 2013 10:56, Andrew  wrote:
>
> >  hm2/hm2_7i43.0: Encoder 0: quadrature count error
> >
> > Though nothing seems to happen, I'm afraid that the encoder might lose
> > impulses or smth. What's the problem? Nothing similar on 2.5.2 with
> exactly
> > same settings, so the hardware should be ok.
>
> I get that with my machine (home-made encoder). it is a new error
> message added by this commit:
>
> http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=commit;h=a3719bd5a9a6c64326a4d1ac9c32fe06385f07cf
> And is based on a fault bit raised by the Mesa firmware.
>
> It doesn't appear to have been in the 2.5.2 release, so it seems that
> the issue has always been there, but you just weren't being told.
>
> Thanks, Andy! I just exchanged connectors from servo drives 0 and 1 to
7i48, the message still appears on encoder 0. That might be 7i48 problem
then. I will try to switch encoder 0 to another channel on 7i48.

Another problem appeared: world mode always jogs full speed. If it was not
for world mode limits, that could be painful.

Andrew

 --
> atp
> If you can't fix it, you don't own it.
> http://www.ifixit.com/Manifesto
>
>
> --
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
>
--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Joints_axes3 branch

[andy pugh]

On 23 April 2013 10:56, Andrew  wrote:

>  hm2/hm2_7i43.0: Encoder 0: quadrature count error
>
> Though nothing seems to happen, I'm afraid that the encoder might lose
> impulses or smth. What's the problem? Nothing similar on 2.5.2 with exactly
> same settings, so the hardware should be ok.

I get that with my machine (home-made encoder). it is a new error
message added by this commit:
http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=commit;h=a3719bd5a9a6c64326a4d1ac9c32fe06385f07cf
And is based on a fault bit raised by the Mesa firmware.

It doesn't appear to have been in the 2.5.2 release, so it seems that
the issue has always been there, but you just weren't being told.

-- 
atp
If you can't fix it, you don't own it.
http://www.ifixit.com/Manifesto

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Joints_axes3 branch

[Andrew]

Hello gentlemen,

I'm trying joints_axes3 for my delta robot. It has Yaskawa Sigma II servo
motors with analog control and MESA 7i43+7i48. Working config is converted
from 2.5.2 to ja3 (attached).
 I should say that ja3 is much better than 2 years ago when I tried it.
Generally the machine works just like on 2.5.2, but the error message
appears after machine is on and keeps showing later:

 hm2/hm2_7i43.0: Encoder 0: quadrature count error

Though nothing seems to happen, I'm afraid that the encoder might lose
impulses or smth. What's the problem? Nothing similar on 2.5.2 with exactly
same settings, so the hardware should be ok.

Thanks,
Andrew


ja3-delta.hal
Description: Binary data


ja3-delta.ini
Description: Binary data
--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] "Cubieboard: available at 49$" (was: LinuxCNC would be very nice on the UDOO board.)

[Fox Mulder]

Am 23.04.2013 00:07, schrieb Kasey Matejcek:
> Has anyone tried to play a video file from something like a Gopro camera is
> there a enough power to play these type of files ?
> Or isn't there enough video or cpu power to do this
> I have costumer that is asking if I could put something together that would
> play the files after their done recording and they don't want to spend the
> money on a laptop

There are some android xbmc versions which "should" work with hardware
accelerated video decoding. But it seems not to be ready for the masses.

So if you want just an little arm box for video output with hardware
full-hd support i would suggest you buy a rasperry-pi modell A (25$) or
B (35$), depending on the need for ethernet support. For these exist
multiple xbmc versions (openelec, raspbmc, ...) with hardware
accelerated video output. I own one myself and use it as standalone
media player.

Newer gopro cameras use h264 encoded video format which is supported by
the rpi gpu. But you should first be sure that the video format is
supported before you order one. ;)

Ciao,
 Rainer

--
Try New Relic Now & We'll Send You this Cool Shirt
New Relic is the only SaaS-based application performance monitoring service 
that delivers powerful full stack analytics. Optimize and monitor your
browser, app, & servers with just a few lines of code. Try New Relic
and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users