Date   
Re: Unable to open include file

John Johnson <john6060842@...>
 

I appreciate the kind words Kirk.
I'm still surprised at how many people around the world find my little
program useful. I used to have a frappr! map where people could tag
themselves on a map so I could see where everyone is, but frappr! became a
subscription service.
I also like that we've built a respectful and helpful community around the
program. I guess we've really only had one twit since I created the
pcb-gcode group on Yahoo! in December 2005. I greatly appreciate the users
that support other users when I'm busy with school.

Regards,
JJ

On Sun, Aug 7, 2011 at 2:23 PM, k_yarina <kc8mmu@...> wrote:

**


Thanks for all the help! I did a clean install on another PC, took Kurt's
advice and put in #HOME/eagle/ulp/pcb-gcode, turned progress off, and it
generated usable gcode. I cut the gcode on top of a couple pieces of scrap
plywood.

Now to mill a plastic cutting board into a flat surface, try some real
laminate, and expand my drill collection.

I've got a problem with one trace having a corner connected to ground, but
it's probably a clearance/isolation issue. Also, EMC2 isn't turning the
spindle on unless I run another part first, which is pretty weird. I don't
see the problem (there's an M3 where it should be), but it's livable and
I'll track it down when I get time.

JJ - it's pretty impressive you continue to support something you wrote
some years ago and distributed for free. I'm grateful for your help. Also, I
have a borrowed Mini, and my daughter's Macbook Pro (she bought a PC when it
was her money, not mine) on it's way here. It'll join the dozen or so
Windows machines and the collection of Unix and Linux systems lying around,
and be used to try to beat Xcode4 into shape.

Kirk


--- In pcb-gcode@..., "Kurt Poulsen" <kurt@...> wrote:

Hi Kirk

I used following definition $HOME&#92;Eagle&#92;ulp2 in the Control Panel/
Options/Directives/User Language Programmes where the $HOME points to
&#92;users&#92;kirk&#92;documents&#92; in your case and then if you have installed the
content of the zip file in a directive called pcb-gcode (I created a ulp2
directive) under the "eagle" directive, then you path should be
$HOME&#92;Eagle&#92;pcb-gcode

It worked at least for me having the same problem.

Kind regards

Kurt



Fra: pcb-gcode@... [mailto:pcb-gcode@...] P
vegne
af k_yarina
Sendt: 5. august 2011 20:37
Til: pcb-gcode@...
Emne: [pcb-gcode] Re: Unable to open include file





Everything should be in &#92;users&#92;kirk&#92;documents&#92;eagle&#92;pcb-gcode; nothing
from
pcb-gcode was installed in the program file directory.

I'll install pcb-gcode on my CNC w7 desktop at home and try it without
the
progress menu. The test machine is at work and not available. Is eagle
looking in the wrong place, since I found pcb-gcode-prg.scr under
&#92;users&#92;kirk&#92;documents&#92;eagle&#92;pcb-gcode&#92;(i think)source ? Another path
issue?

IIRC W7 (and maybe Vista, too) creates shadow directories to allow user
changes to system directories. This is for compatibility with older
programs
that muck around with system files. It does make it really hard to figure
out where .ini files are for old windows ham radio programs. However,
since
pcb-gcode isn't in the program file directory it shouldn't be writing or
looking there.

Thanks again! Isn't Windoze fun?

Kirk

--- In pcb-gcode@... <mailto:pcb-gcode%40yahoogroups.com> ,
John
Johnson <john6060842@> wrote:

This is probably Windows trying to keep you from getting a virus.
Windows
keeps an eye on directories like c:&#92;program files&#92;*. If a new file
appears
it may be removed. If an existing file changes, it may be replaced with
an
older version.
There are two ways you can try to fix this:
In pcb-gcode-setup go to the Other tab and turn off Show progress menu.
or
Move everything to a directory that isn't monitored, like c:&#92;documents
and
settings&#92;whatever. This might be a good idea as the previewer needs to
create a temporary file which Windows might replace on the fly, meaning
you
could get the wrong preview.

Let us know how it goes.

Regards,
JJ


On Fri, Aug 5, 2011 at 12:02 AM, k_yarina <kc8mmu@> wrote:



--
Sent from a MacBook Pro


[Non-text portions of this message have been removed]

Re: Auto-leveller -- another IMPORTANT fix revision 66.

Gaston Gagnon
 

On 2011-08-07 14:19, essele1 wrote:

HI Gaston,

This is good news - although it should be no different than the original
pcb-gcode behaviour now.
Sorry, I had not looked at the drill part before.

In the new version I have the option of using tool change (Tnn/M05 codes)
or Operator Pause (M00) or both, plus I have better control of the
spindle stop/start and speeds.
Are you keeping the form M6Tnn though?

Your remarks about positioning actually aren't my code, this is the
original behaviour of pcb-gcode ... there's a "tool change" location
setting which sets the X, Y, and Z positions. Unfortunately they default
to zero, so they can cause problems with the z-axis, I always change Z to
+25mm at least, although 40mm is probably easier to actually get the
tools changed.
I see what you mean at line 3 but at line 5 Z is sent back to 0.0 then at
line 6 it is moved to location (0,0)
As we are presuming that the pcb is warped, it seem to me that the bit
could/will drag on the pcb surface on its way to 0,0. What do you think?

1. G00 X-3.1600 Y1.6100
2. G01 Z-0.0320 F10
3. G00 Z0.1000
4. M05
5. G00 Z0.0000
6. G00 X0.0000 Y0.0000
7. M06 T03 ; 0.0440
8. G00 Z0.1000
9. M03
10. G04 P3.000000
11. G00 X-1.4200 Y0.3400

The delay is also original code, based on spindle spin-up time ... I
think this is necessary as my spindle does take a while to spin up. In
any case it's configurable.
You are right I did not think of the spindle time to reach speed since I
control it manually.


--
Gaston
http://public.fotki.com/Gaston-Gagnon/

Re: Unable to open include file

Kirk
 

Thanks for all the help! I did a clean install on another PC, took Kurt's advice and put in #HOME/eagle/ulp/pcb-gcode, turned progress off, and it generated usable gcode. I cut the gcode on top of a couple pieces of scrap plywood.

Now to mill a plastic cutting board into a flat surface, try some real laminate, and expand my drill collection.

I've got a problem with one trace having a corner connected to ground, but it's probably a clearance/isolation issue. Also, EMC2 isn't turning the spindle on unless I run another part first, which is pretty weird. I don't see the problem (there's an M3 where it should be), but it's livable and I'll track it down when I get time.

JJ - it's pretty impressive you continue to support something you wrote some years ago and distributed for free. I'm grateful for your help. Also, I have a borrowed Mini, and my daughter's Macbook Pro (she bought a PC when it was her money, not mine) on it's way here. It'll join the dozen or so Windows machines and the collection of Unix and Linux systems lying around, and be used to try to beat Xcode4 into shape.


Kirk

--- In pcb-gcode@..., "Kurt Poulsen" <kurt@...> wrote:

Hi Kirk

I used following definition $HOME&#92;Eagle&#92;ulp2 in the Control Panel/
Options/Directives/User Language Programmes where the $HOME points to
&#92;users&#92;kirk&#92;documents&#92; in your case and then if you have installed the
content of the zip file in a directive called pcb-gcode (I created a ulp2
directive) under the "eagle" directive, then you path should be
$HOME&#92;Eagle&#92;pcb-gcode

It worked at least for me having the same problem.

Kind regards

Kurt



Fra: pcb-gcode@... [mailto:pcb-gcode@...] På vegne
af k_yarina
Sendt: 5. august 2011 20:37
Til: pcb-gcode@...
Emne: [pcb-gcode] Re: Unable to open include file





Everything should be in &#92;users&#92;kirk&#92;documents&#92;eagle&#92;pcb-gcode; nothing from
pcb-gcode was installed in the program file directory.

I'll install pcb-gcode on my CNC w7 desktop at home and try it without the
progress menu. The test machine is at work and not available. Is eagle
looking in the wrong place, since I found pcb-gcode-prg.scr under
&#92;users&#92;kirk&#92;documents&#92;eagle&#92;pcb-gcode&#92;(i think)source ? Another path issue?

IIRC W7 (and maybe Vista, too) creates shadow directories to allow user
changes to system directories. This is for compatibility with older programs
that muck around with system files. It does make it really hard to figure
out where .ini files are for old windows ham radio programs. However, since
pcb-gcode isn't in the program file directory it shouldn't be writing or
looking there.

Thanks again! Isn't Windoze fun?

Kirk

--- In pcb-gcode@... <mailto:pcb-gcode%40yahoogroups.com> , John
Johnson <john6060842@> wrote:

This is probably Windows trying to keep you from getting a virus. Windows
keeps an eye on directories like c:&#92;program files&#92;*. If a new file appears
it may be removed. If an existing file changes, it may be replaced with an
older version.
There are two ways you can try to fix this:
In pcb-gcode-setup go to the Other tab and turn off Show progress menu.
or
Move everything to a directory that isn't monitored, like c:&#92;documents and
settings&#92;whatever. This might be a good idea as the previewer needs to
create a temporary file which Windows might replace on the fly, meaning
you
could get the wrong preview.

Let us know how it goes.

Regards,
JJ


On Fri, Aug 5, 2011 at 12:02 AM, k_yarina <kc8mmu@> wrote:

Re: Auto-leveller -- another IMPORTANT fix revision 66.

essele1
 

HI Gaston,

This is good news - although it should be no different than the original pcb-gcode behaviour now.

In the new version I have the option of using tool change (Tnn/M05 codes) or Operator Pause (M00) or both, plus I have better control of the spindle stop/start and speeds.

Your remarks about positioning actually aren't my code, this is the original behaviour of pcb-gcode ... there's a "tool change" location setting which sets the X, Y, and Z positions. Unfortunately they default to zero, so they can cause problems with the z-axis, I always change Z to +25mm at least, although 40mm is probably easier to actually get the tools changed.

The delay is also original code, based on spindle spin-up time ... I think this is necessary as my spindle does take a while to spin up. In any case it's configurable.

Regards,

Lee.

--- In pcb-gcode@..., Gaston Gagnon <gaston.gagnon@...> wrote:

Hi Lee,
I just tried r66 and the tool change work correctly for Mach3 but users
have to enable Tool change before. To do this, go in Config menu then
select General Logic Configuration then check "Stop Spindle. Wait for Cycle
Start." in Tool change (top left, second line).

A few remarks:
Before each Tool Change you move Z to 0 then move to X0Y0. Don't you think
that may scratch the pcb on its way to 0,0 if the pcb is warped?

Beside it may not be necessary to move to X0X0 at all if people are using
drills with stop ring, raise Z to a safe level could be enough.

With Tool Change enabled, G04 P3.0 delay (I suppose you have added that
line to give time to people to stop their machine) is not necessary anymore.

Gaston



On 2011-08-07 08:51, essele1 wrote:

Hi,

I've just uploaded another fix to the alpha1 tree of the leveller. It
seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to make
sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very
nearly ready.

Cheers,

Lee.


--
Gaston
http://public.fotki.com/Gaston-Gagnon/



[Non-text portions of this message have been removed]

Re: Auto-leveller -- another IMPORTANT fix revision 66.

Gaston Gagnon
 

Hi Lee,
I just tried r66 and the tool change work correctly for Mach3 but users
have to enable Tool change before. To do this, go in Config menu then
select General Logic Configuration then check "Stop Spindle. Wait for Cycle
Start." in Tool change (top left, second line).

A few remarks:
Before each Tool Change you move Z to 0 then move to X0Y0. Don't you think
that may scratch the pcb on its way to 0,0 if the pcb is warped?

Beside it may not be necessary to move to X0X0 at all if people are using
drills with stop ring, raise Z to a safe level could be enough.

With Tool Change enabled, G04 P3.0 delay (I suppose you have added that
line to give time to people to stop their machine) is not necessary anymore.

Gaston

On 2011-08-07 08:51, essele1 wrote:

Hi,

I've just uploaded another fix to the alpha1 tree of the leveller. It
seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to make
sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very
nearly ready.

Cheers,

Lee.

Re: Auto-leveller -- another IMPORTANT fix revision 66.

Art Eckstein <art.eckstein@...>
 

Lee,
Be glad to help. As you look at the turbocnc.pp, I am the one that is listed in the credits. I have been active in modifying and enhancing Turbocnc for quite some time and look forward to making this work. Hopefully, my router will be done by the time you get to this phase.

Art
Country bubba

At 10:55 AM 8/7/2011, you wrote:
Hi Art,

I haven't even looked at the turbocnc profile yet, but I have been working to simplify the changes, so it should be easier to implement new machine support.

Once I get the new code out it would be good if you could help to test the turbocnc version.

If you have any info about turbocnc's gcode support that would be useful also - I'll put together a first-guess and we can iterate on getting it working.

Cheers,

Lee.

--- In pcb-gcode@..., Art Eckstein <art.eckstein@...> wrote:

Lee,
I have been following the thread and hope to implement your system as
soon as I get my router finished. I see most comments relating to
Mach and EMC and I will be using Turbocnc. I see the post info for
Turbocnc is included in the zip file. Is there any additional changes
that will have to be made to make the system work with Turbocnc?

Thanks in advance.

Art
Country Bubba


At 08:51 AM 8/7/2011, you wrote:
Hi,

I've just uploaded another fix to the alpha1 tree of the leveller.
It seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to
make sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very
nearly ready.

Cheers,

Lee.

Re: Auto-leveller -- another IMPORTANT fix revision 66.

essele1
 

Hi Art,

I haven't even looked at the turbocnc profile yet, but I have been working to simplify the changes, so it should be easier to implement new machine support.

Once I get the new code out it would be good if you could help to test the turbocnc version.

If you have any info about turbocnc's gcode support that would be useful also - I'll put together a first-guess and we can iterate on getting it working.

Cheers,

Lee.

--- In pcb-gcode@..., Art Eckstein <art.eckstein@...> wrote:

Lee,
I have been following the thread and hope to implement your system as
soon as I get my router finished. I see most comments relating to
Mach and EMC and I will be using Turbocnc. I see the post info for
Turbocnc is included in the zip file. Is there any additional changes
that will have to be made to make the system work with Turbocnc?

Thanks in advance.

Art
Country Bubba


At 08:51 AM 8/7/2011, you wrote:
Hi,

I've just uploaded another fix to the alpha1 tree of the leveller.
It seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to
make sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very
nearly ready.

Cheers,

Lee.

Re: Auto-leveller -- another IMPORTANT fix revision 66.

essele1
 

Hi Bruce,

Yep .. fair point ...

The most recent version is available here:

http://svn.randomthinker.org/trac/changeset/67/pcbgcode/tags/2.0-alpha1?old_path=%2F&format=zip

Anything newer than 66 is correct at present.

Regards,

Lee.

--- In pcb-gcode@..., bruce q <brc_q@...> wrote:

Lee,
I have been watching this thread on the leveler pcb-gcode and hope to try it out soon..
But may I make a suggestion here..
When you post that you have uploaded a new revision could you please include the link
to the new rev?
I have the original link in an post that you did many, many posts ago and would have to go through each
one trying to find it. So it would sure be easy to catch the latest rev if you included that link
in your recent update post.
Thanks..
Bruce



________________________________
From: essele1 <lee.essen@...>
To: pcb-gcode@...
Sent: Sunday, August 7, 2011 5:51 AM
Subject: [pcb-gcode] Auto-leveller -- another IMPORTANT fix revision 66.


 
Hi,

I've just uploaded another fix to the alpha1 tree of the leveller. It seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to make sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very nearly ready.

Cheers,

Lee.





Re: Viewer

Jojo
 

Hi John,

Sent the attachment via email.

Cheers!

On Sun, Aug 7, 2011 at 9:42 PM, John Johnson <john6060842@...> wrote:

Hi Jojo,

Attachments are stripped from messages.
You can email it directly to me ( pcbgcode@... ), or send it to
me
on Facebook.
Thanks for the kind words!

Regards,
John

On Sat, Aug 6, 2011 at 10:32 PM, Jojo Dijamco <jojod818@...> wrote:

**


Hi John!

First of all, thanks for this wonderful software you are sharing!

I've attached a screen shot of my setup page, I can't seem to find the
preview check box??? The only check box I see is Generate Milling. You
will
also notice that I have unchecked the Top side of the pcb since I only do
single-sided boards. :)

Warm Regards,

Jojo


On Sun, Aug 7, 2011 at 12:55 AM, John Johnson <john6060842@...>
wrote:

Hi Jojo,

Do you have Show preview checked under Generation Options | Board?
If so, the viewer should automatically open as each etch file is
created.

Regards,
JJ

On Sat, Aug 6, 2011 at 9:10 AM, Jojo <jojod818@...> wrote:

**


Hi everyone!

I've just completed my very first Eagle > PCB Gcode > my diy CNC pcb
milling process and it rocks!

Many thanks to PCB Gcode, otherwise I would have been amiss with all
the
gcode commands.

I have one question though and that I can't seem to view my board
using
the
Viewer. How do I go about it? I'm using EMC2 profile btw.

Many thanks and more power!

Jojo




--
Sent from a MacBook Pro






------------------------------------

Find us on Facebook:
http://www.facebook.com/group.php?gid=324296011726&ref=ts


Also check out the Phorum http://www.pcbgcode.org/

Yahoo! Groups Links








--
Sent from a MacBook Pro






------------------------------------

Find us on Facebook:
http://www.facebook.com/group.php?gid=324296011726&ref=ts


Also check out the Phorum http://www.pcbgcode.org/

Yahoo! Groups Links



Re: Auto-leveller -- another IMPORTANT fix revision 66.

Art Eckstein <art.eckstein@...>
 

Lee,
I have been following the thread and hope to implement your system as soon as I get my router finished. I see most comments relating to Mach and EMC and I will be using Turbocnc. I see the post info for Turbocnc is included in the zip file. Is there any additional changes that will have to be made to make the system work with Turbocnc?

Thanks in advance.

Art
Country Bubba

At 08:51 AM 8/7/2011, you wrote:
Hi,

I've just uploaded another fix to the alpha1 tree of the leveller. It seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to make sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very nearly ready.

Cheers,

Lee.

Re: Auto-leveller -- another IMPORTANT fix revision 66.

bruce q
 

Lee,
I have been watching this thread on the leveler pcb-gcode and hope to try it out soon..
But may I make a suggestion here..
When you post that you have uploaded a new revision could you please include the link
to the new rev?
I have the original link in an post that you did many, many posts ago and would have to go through each
one trying to find it. So it would sure be easy to catch the latest rev if you included that link
in your recent update post.
Thanks..
Bruce



________________________________
From: essele1 <lee.essen@...>
To: pcb-gcode@...
Sent: Sunday, August 7, 2011 5:51 AM
Subject: [pcb-gcode] Auto-leveller -- another IMPORTANT fix revision 66.


 
Hi,

I've just uploaded another fix to the alpha1 tree of the leveller. It seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to make sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very nearly ready.

Cheers,

Lee.

Re: Viewer

John Johnson <john6060842@...>
 

Hi Jojo,

Attachments are stripped from messages.
You can email it directly to me ( pcbgcode@... ), or send it to me
on Facebook.
Thanks for the kind words!

Regards,
John

On Sat, Aug 6, 2011 at 10:32 PM, Jojo Dijamco <jojod818@...> wrote:

**


Hi John!

First of all, thanks for this wonderful software you are sharing!

I've attached a screen shot of my setup page, I can't seem to find the
preview check box??? The only check box I see is Generate Milling. You will
also notice that I have unchecked the Top side of the pcb since I only do
single-sided boards. :)

Warm Regards,

Jojo


On Sun, Aug 7, 2011 at 12:55 AM, John Johnson <john6060842@...>
wrote:

Hi Jojo,

Do you have Show preview checked under Generation Options | Board?
If so, the viewer should automatically open as each etch file is created.

Regards,
JJ

On Sat, Aug 6, 2011 at 9:10 AM, Jojo <jojod818@...> wrote:

**


Hi everyone!

I've just completed my very first Eagle > PCB Gcode > my diy CNC pcb
milling process and it rocks!

Many thanks to PCB Gcode, otherwise I would have been amiss with all
the
gcode commands.

I have one question though and that I can't seem to view my board using
the
Viewer. How do I go about it? I'm using EMC2 profile btw.

Many thanks and more power!

Jojo




--
Sent from a MacBook Pro


[Non-text portions of this message have been removed]



------------------------------------

Find us on Facebook:
http://www.facebook.com/group.php?gid=324296011726&ref=ts


Also check out the Phorum http://www.pcbgcode.org/

Yahoo! Groups Links



[Non-text portions of this message have been removed]




--
Sent from a MacBook Pro

Auto-leveller -- another IMPORTANT fix revision 66.

essele1
 

Hi,

I've just uploaded another fix to the alpha1 tree of the leveller. It seems I broke the tool change code, so this reverts it.

This is a change to the EMC and Mach profiles, so you'll need to make sure you reselect the profile.

The correct revisision is now r66.

Hopefully this will be the last one of these as the new code is very nearly ready.

Cheers,

Lee.

Re: Auto pcb-levelling plus a few other things...

essele1
 

Hi Richard,

Thanks for this. As I was working through my tool change code today I also noticed a problem here.

It seems that I misunderstood the original behaviour because of a badly labelled variable, and hence I've now made things worse!

There was originally an "OPERATOR PAUSE" variable marked at M06, for EMC there's actually a M60 which is pause operation, so I assumed (incorrectly) that this was a type and "fixed" them in both emc and mach profiles.

It turns out that M06 is actually the tool change request (which does have manual interaction in EMC and I assume Mach) ... interestingly though it's called before the Tnn to select the tool, so it's wrong in any case.

I've completely reworked this in my new code, but I'll upload a fix to the alpha tree to revert this behaviour otherwise it could could problems.

Cheers,

Lee.

--- In pcb-gcode@..., "Richard A Blick" <rab@...> wrote:

OK............. My fault for not reporting fully on what happened during the drilling phase of my test PCB. I have to say, at the time my main focus was on the levelling and probing side of things, so sorry about that, if as a result it's since caused problems for anyone.................!

From memory, the drill events...............

After completion of the etching phase, on the test PCB, the program halted for a short pause but then continued straight on into the drill cycle - (so it didn't stop for me either).

I was however waiting to change the tool and was able to manually stop Mach and catch the program from running any further.

Rechecked the GCode and had a very quick look for what a M60 comand was and found nothing, but decided to continue the drill process on anyway. (Was going to get back to you about that Lee!).

Replaced the tool in the spindle with a 0.8mm dril and reset the Z DRO height to 0 for the new tool and continued the drill cycle.

(Note: After a tool change and when resetting the Z height to 0, I always move the X,Y axis and then the Z, so the tool is positioned a couple of mm above a section of the PCB so I can check that there is going to be continuity between the probe and board by using a steel rule to bridge between the two, before actually zero'ing. It's a good test to catch any leads that might have come loose or fallen off and believe you me it happens!.)


Part way through the drill cycle there was a further stop and change required for a larger drill, which I was watching out for, but the program didn't stop for that either, (but I noted the programe line number)............and .this time I let the cycle continue on with the smaller 0.8mm drill, through to completion.

I then changed the drill size to the larger 1.0mm, rezeroed the Z axis, and reran Mach using the "Run from here" option - setup on the drill change line number previously noted, and the program ran from that point through to completion, without problem.

All of my boards are single or double sided - definitely not into multilayer boards, so whether I drill or etch first is not an issue for me either, although I do spot drill during the etch cycle - probably more as a visual check and reassurance for me prior to actually doing the drilling. It's just something I do now.

Also for testing I fit a cork floor tile to the router table and carpet tape the PCB to the cork tile - the tiles have a good flat and firm surface and any drill or mill tool "plunges" have a better chance of survival. Has worked well for me in the past. It's a good test for the leveller code as well.

Hope that helps.

Richard



Re: Viewer

Jojo
 

Hi John!

First of all, thanks for this wonderful software you are sharing!

I've attached a screen shot of my setup page, I can't seem to find the
preview check box??? The only check box I see is Generate Milling. You will
also notice that I have unchecked the Top side of the pcb since I only do
single-sided boards. :)

Warm Regards,

Jojo

On Sun, Aug 7, 2011 at 12:55 AM, John Johnson <john6060842@...> wrote:

Hi Jojo,

Do you have Show preview checked under Generation Options | Board?
If so, the viewer should automatically open as each etch file is created.

Regards,
JJ

On Sat, Aug 6, 2011 at 9:10 AM, Jojo <jojod818@...> wrote:

**


Hi everyone!

I've just completed my very first Eagle > PCB Gcode > my diy CNC pcb
milling process and it rocks!

Many thanks to PCB Gcode, otherwise I would have been amiss with all the
gcode commands.

I have one question though and that I can't seem to view my board using
the
Viewer. How do I go about it? I'm using EMC2 profile btw.

Many thanks and more power!

Jojo




--
Sent from a MacBook Pro






------------------------------------

Find us on Facebook:
http://www.facebook.com/group.php?gid=324296011726&ref=ts


Also check out the Phorum http://www.pcbgcode.org/

Yahoo! Groups Links



Re: Leveller -- tool change and auto-calibrate?

Z P
 

Hello Lee,

I do not use pcb-gcode. I have started a small prototyping service and as such I need to be able to read in Gerber and Excellon drill files.
I use a commercial Gerber editor which allows me to generate isolation contours. The contours are then output in 274X Gerber format followed by a bit of conversion software to get from 274X gerber to 274N ( i think thats the official label ) for CNC G code language.
Similarly the excellon drill file format is parsed to convert it to cnc g code.

There are a number of methods of activating copper holes.

Simplest and incrementally cheapest is use of conductive inks. The PCB is drilled... holes filled with ink and cleared of excessive amounts of ink followed by a drying/curing in an oven... and then electroplated in copper acid tank. ( see Think'n'tinker website  The fellow from Soigeneris who is a member also distributes Think'n'tinker milling and drilling bits I have no affiliation with the businesses).

I use a more complex process based on palladium colloidal preparation based on preconditioning, etchin, cleaning developing chain of process tanks. This approach allows me to electroplate teflon substrate.
Palladium is maintained under constant mixing and its redox potential constantly monitored and a reducing reagent added to maintain the redox potnetial at the setpoint.

The PCB is pretreated through wetting agents, FR4  microetch agent,copper preetching agent.. and then dipped in the paladium solution.

Paladium solution adheres to pre wetted fibreglas base, it is then developed whihc pops open packets of paladiumon non copper surface and is then placed in the copper acid electrolytic bath for plating.

All up it takes a half an hour through the activation line ... and about 80 minute in the copper electrolytic plating bath for a mirro finish half ounce copper plate layer on top of what ever was there.

As for probing I use a floating head. Initially i used a fixed head and the ( lack of ) quality of laminate was immediately apparent. I then fabricated a floating head ( which is a couple of kilos in weight and its full weight is supported by the PCB being machined. I get a good intimate contact between PCB and milling machine base and a steady depth of tool penetration on the top side.

While I do not use Eagle I find the group has some interesting thinkers whose take on problems and ideas had helped me develop my approach



Zig




________________________________
From: essele1 <lee.essen@...>
To: pcb-gcode@...
Sent: Sunday, 7 August 2011 2:54 AM
Subject: [pcb-gcode] Re: Leveller -- tool change and auto-calibrate?


 
Hi Zig,

Wow ... it definitely sounds like you are leagues ahead of my efforts, I haven't even braved double sided yet!

I have added code to make the order an option, although I can see another problem...

You drill, then remove the board to activate/plate, therefore you're really going to need to probe again (assuming you bothered for the drilling phase) and you have holes all over the board!

Now, we could get sophisticated, we know where the holes are and how big they are, so we could find the nearest non-holed spot to probe and adjust the interpolation accordingly but it's certainly a more complex approach.

Once the basic system is up and running, I could certainly look at this if there's any interest.

I the meantime I'd love to hear what methods you use for activating and plating?

Lee.

--- In pcb-gcode@..., Z P <chatnick_au@...> wrote:

Well there are a couple of differing opinions on whether it is better to drill first or not.


They are based on perceived benefits of not wasting the drill effort should milling fail.

Milling will fail as soon as You get serious about high density,  "industrial grade" copper annuls around the hole.

Milling process would obliterate the "donut" if too close to the hole edge.


I drill down to 0.3mm dia and mill down to 0.1mm isolation clearance ( for the imperially minded members 0.1mm is a tad under 0.004"), additionally I electroplate copper on inside holes.

It would be impracticable to mill -> drill-> activate holes ->plate.
The process from my perspective is drill -> activate holes -> electroplate -> mill.

In making multilayer panels I mill core and inner layers. This is followed by laminating them into a finished panel which is then drilled and electroplated. Final stage in the process is milling of external layers.

This is my take on mechanical structuring of boards


Zig


PS most recent effort four layer board, 0.1mm isolation, 0.1mm tracks. 0.3mm hole inside 0201 SMD pads. Board thickness 0.44mm board width 0.91mm. attrition through the process 69 out of 80. Lesson learned, expect attrition rate to be 1 in 10 in second batch. Approximately two third failed due to a bad 0.4mm end mill which cut 0.1mm oversize.



________________________________
From: essele1 <lee.essen@...>
To: pcb-gcode@...
Sent: Saturday, 6 August 2011 2:35 AM
Subject: [pcb-gcode] Re: Leveller -- tool change and auto-calibrate?


 
Hi Zig,

Interesting ... I used to drill first, but I found that the most likely problem area was the etching, hence I didn't want to waste time drilling if the etching was going to subsequently fail, hence I tend to etch first.

With the leveller, it's definitely much more consistent, so I probably could switch round ... in fact, this sounds like a good option to have in the setup ui.

The optimiser is very simple, it iterates through all the paths and finds the closest one to the current location each time ... if it's closer than the tool width (as it would be for a multi-pass etch) then it doesn't lift.

Similar approach to the hole drilling, although obviously optimised for each group of holes based on size.

Lee.

--- In pcb-gcode@..., Z P <chatnick_au@> wrote:

Lee,


A small observation. Your thinking is that You ought to be milling copper first which is then followed by drilling.

In my experience it needs to be the other way around.  Drill first since drilling of small pads tends to be more aggressive on the laminate than milling. That is to say drill first while copper foil is intact and strong and only then mill.


This is also a prerequisite in case of plating holes.

On the question of optimisation do you intend to start at the beginning of file and look for nearest contour start point ( doing the radial distance calculation or do you intend to do something else in the way of optimisation??

Regards

Zig




________________________________
From: essele1 <lee.essen@>
To: pcb-gcode@...
Sent: Friday, 5 August 2011 8:15 PM
Subject: [pcb-gcode] Leveller -- tool change and auto-calibrate?


 
Hi,

I'm getting very close to releasing the next version of the leveller with path and drill optimisation included.

It's a bit of a major restructure of the pcb-gcode approach but I think still works ok.

Basic premise is that you only want to run the probe phase once for each side of the board, therefore the drilling and milling code needs to be included in the same file as the etch code (with the option of including each in either the top or the bottom.)
(I haven't looked at the text or fill stuff yet)

So, if it's all in one file you need a way to change tools (you'd need this for the drilling phase anyway), you obviously have the probe, so I am thinking the following...

1. Provide a maximum tool size delta parameter (i.e. how much longer is your longest tool than your shortest tool)
2. Provide a tool change location where you can sensibly change the tool.

For this example, we'll use a tool change Z of 40mm and a max tool delta of 10mm.

The start is actually a bit more awkward, so lets pretend we are happily etching away with the z-axis zeroed nicely at the board surface.

So now we need to change to a drill for the start of the drilling phase.

1. Move to the tool change location (Z=40mm) [plus X and Y need to be above a non-etched part of the board (0,0 is probably ok)
2. Either use 'T' codes to change tool, or Operator pause and a message (I have this is an option currently)
3. Move down to the surface position less the max delta, i.e. z=+10mm ... this way if this tool is 10mm longer than the last we are still clear.
4. Probe for 2 x max_tool_delta ... so we can cope with this tool being 10mm less than the last.
5. Reset z=0 once the probe triggers and start the first drilling phase.

Repeat for each tool change.

Does this sound reasonable? In non-leveller mode it would remain pretty much the same as it is now, although would have the option to use operator pause.

The initial start is a bit more awkward since you don't have a reference Z to work with, my thought is to keep similar to now, ensure that the tool is within a reasonable distance of the surface and probe.

Any thoughts appreciated. I'm just about to code this up .. the rest of it is largely done, and getting good optimisation results. I've also added a tool section to the setup ui, so you actually edit the tools here rather than as a separate file ... I think it's more intuitive, plus allows me to implement different speeds/feeds for different tools, and ultimately the option to spiral mill larger holes.

Cheers,

Lee.




[Non-text portions of this message have been removed]



[Non-text portions of this message have been removed]



[Non-text portions of this message have been removed]

Re: Auto pcb-levelling plus a few other things...

R A Blick
 

OK............. My fault for not reporting fully on what happened during the drilling phase of my test PCB. I have to say, at the time my main focus was on the levelling and probing side of things, so sorry about that, if as a result it's since caused problems for anyone.................!

From memory, the drill events...............

After completion of the etching phase, on the test PCB, the program halted for a short pause but then continued straight on into the drill cycle - (so it didn't stop for me either).

I was however waiting to change the tool and was able to manually stop Mach and catch the program from running any further.

Rechecked the GCode and had a very quick look for what a M60 comand was and found nothing, but decided to continue the drill process on anyway. (Was going to get back to you about that Lee!).

Replaced the tool in the spindle with a 0.8mm dril and reset the Z DRO height to 0 for the new tool and continued the drill cycle.

(Note: After a tool change and when resetting the Z height to 0, I always move the X,Y axis and then the Z, so the tool is positioned a couple of mm above a section of the PCB so I can check that there is going to be continuity between the probe and board by using a steel rule to bridge between the two, before actually zero'ing. It's a good test to catch any leads that might have come loose or fallen off and believe you me it happens!.)


Part way through the drill cycle there was a further stop and change required for a larger drill, which I was watching out for, but the program didn't stop for that either, (but I noted the programe line number)............and .this time I let the cycle continue on with the smaller 0.8mm drill, through to completion.

I then changed the drill size to the larger 1.0mm, rezeroed the Z axis, and reran Mach using the "Run from here" option - setup on the drill change line number previously noted, and the program ran from that point through to completion, without problem.

All of my boards are single or double sided - definitely not into multilayer boards, so whether I drill or etch first is not an issue for me either, although I do spot drill during the etch cycle - probably more as a visual check and reassurance for me prior to actually doing the drilling. It's just something I do now.

Also for testing I fit a cork floor tile to the router table and carpet tape the PCB to the cork tile - the tiles have a good flat and firm surface and any drill or mill tool "plunges" have a better chance of survival. Has worked well for me in the past. It's a good test for the leveller code as well.

Hope that helps.

Richard

Re: Auto pcb-levelling plus a few other things...

Danny Miller <dannym@...>
 

Mach3 has many options to monkey with how commands are implemented.
Unfortunately that system is rather confusing and difficult to track
exactly what's going to happen when you've got multiple profiles
floating around.

But M60 is NOT "pause during tool change". It's not listed under my
Mach3 M-code page at all, so I assume it's not supported. Wikipedia
lists M60 as:
M60 Automatic pallet change (APC) M For machining centers with
pallet changers


Does the leveling code actually use M60? That can't be right... it
sounds like an unnecessarily obscure code for a delay, and I don't see
how "pallet changers" has anything to do with the job at hand.

Danny

On 8/6/2011 7:09 AM, Mail wrote:

Hi.
I have been following this thread for some time now and rapidly
discovered a few problems when using Mach-3, which I see have been
overcome with the latest releases of the leveller code. I must point
out one that I dont think has been mentioned, Mach-3 doesnt respond to
M60 for pause during tool change so I have changed the default setting
to M0.

I ruined one board because I hadnt noticed that the 0,0 corner had
been damaged and the initial probe test for Z=0 went lower than it
should thereby 'etching' everything too deep. I therefore now test for
initial Z=0 at 0.5,0.5 (mm).

Another point regarding the 'drill before mill' question. I tried this
and promptly ruined another board because my probe tip gave false
results when it landed on top of a hole.

After all this I must congratulate Lee and all the 'beta-testers' on
their efforts to produce a piece of software which makes mechanical
etching of circuit boards a reality for people with home built and
other 'economy' machines.

Dick...



Re: Auto pcb-levelling plus a few other things...

dick williams
 

Hi Lee.

The problem with the low probe was totally of my own making.
I was probing with the cutting tool and not watching what I was doing.
I had my Z-up set to 0.5 mm, because of the damaged corner Z-zero got set to slightly lower than this consequently I carved a lovely pattern accross the pcb.(smack wrist).

As for mill/drill I usually probe and etch the top, then probe and etch the bottom and finnish off with the drilling. I am not concerned about the auto levelling for the drilling, I just go through into a sacrificial layer under the pcb.

Keep up the good work...
Dick...

Hi Dick,

Thanks for the note about M60 ... nobody had noticed that ... interestingly it was originally M06 in both emc and mach profiles and hence >didn't seem to have any effect at all.

I'm a little concerned about your 0,0 probe problem though .. the initial probe at 0,0 only serves to set a reference Z position which is then >used to set the height for the rest of the probes ... if it was low then the probes should compensate?

On the drill-before-mill probe-hitting-hole comment ... you would probe once before any of the drill/mill cycles ... you cannot probe >afterwards, either you run the risk of hitting a hole or you end up in an isolated section and don't have probe continuity (been there, done that!)

Cheers,

Lee.

Re: Viewer

John Johnson <john6060842@...>
 

Hi Jojo,

Do you have Show preview checked under Generation Options | Board?
If so, the viewer should automatically open as each etch file is created.

Regards,
JJ

On Sat, Aug 6, 2011 at 9:10 AM, Jojo <jojod818@...> wrote:

**


Hi everyone!

I've just completed my very first Eagle > PCB Gcode > my diy CNC pcb
milling process and it rocks!

Many thanks to PCB Gcode, otherwise I would have been amiss with all the
gcode commands.

I have one question though and that I can't seem to view my board using the
Viewer. How do I go about it? I'm using EMC2 profile btw.

Many thanks and more power!

Jojo




--
Sent from a MacBook Pro