Topics

removing trace isolation swarf #mill #development

John Ferguson
 

I suppose swarf isn't the best word, but I'm curious about how you guys get rid of the chips which lodge in cuts.  I've been chasing them out with a needle, buy it's tedious.

I've got 90 psi air in the shop. Maybe I should try that.

I've got my production system running very smoothly. Design In Eagle, generate Gcode in pcpcode, sometimes include autocad millwork with gcode generated in SheetCAM, set up work piece, set height on cutter and go.

I make all of my boards from 2.5x4.0 workpieces. I made an aluminum fixture with a depression which fist these workpieces and has a lid which locates 4 holes one in each corner.  I drill the holes and then screw the board down to an HDPE fixture bolted to the table on my 6040 router.  The HDPE has 4 holes which take 6-32 nylon socket head caps screws.  I face off the surface the board sits on every 4th or 5th board to maintain level. 

I adjust the bit height by rolling a piece of .250 drill rod back and forth until the bit just clears and then dropping it one click on the servo so it doesn't quite clear.  I have vertical "home" set at Z=1.000 on the machine so I have to raise the spindle .750 before I start.  I suppose I could lower the home to .250 and eliminate the "raise" but having it at one inch gives me enough time to catch some kinds of screw-ups.

I've been delighted with this bit. I buy them from Precise Bits and they seem to last pretty well. 

EM3E8-0100-15V 0.0100in. (0.254mm) tip dia., 3-flute 15° tapered stub trace-isolation bit, 0.020 in. DOC, 1.42 in. (36mm) OAL 2 $16.95 $33.90

The size of the bur on the edge of cuts seems to increase as they wear out, but only downside of this seems to be more dressing of the finished board. I use 400 grit paper and final clean up with 800 crocus cloth.

I've been tinning the finished boards with MG Chemical 421.  I'm a bit mystified why I get better tinning if i rough up the board with 400 grit before plating it.  I wash board with acetone before tinning.

After I've soldered everything (all through hole - no smds yet), I go over the board with a loupe and remove any shorts with braid and clean out any suspicious cuts with a needle.

and they always work, at least the ones I haven't messed up the circuits on.

If you read this and see anything stupid, let me know.  Also I'm still like a better way to get rid of chips.

regards,

John ferguson  delray beach

 

Probably starting a religious war, but I've actually had better luck with engraving bits that actual mills for traces. I change over to a real mill for milling operations.

What I really need to do is master gcode well enough to have a routine that re-zero's the height after changing a bit.

Bob


On Thu, Mar 21, 2019 at 4:26 PM John Ferguson via Groups.Io <jferg977=aol.com@groups.io> wrote:

I suppose swarf isn't the best word, but I'm curious about how you guys get rid of the chips which lodge in cuts.  I've been chasing them out with a needle, buy it's tedious.

I've got 90 psi air in the shop. Maybe I should try that.

I've got my production system running very smoothly. Design In Eagle, generate Gcode in pcpcode, sometimes include autocad millwork with gcode generated in SheetCAM, set up work piece, set height on cutter and go.

I make all of my boards from 2.5x4.0 workpieces. I made an aluminum fixture with a depression which fist these workpieces and has a lid which locates 4 holes one in each corner.  I drill the holes and then screw the board down to an HDPE fixture bolted to the table on my 6040 router.  The HDPE has 4 holes which take 6-32 nylon socket head caps screws.  I face off the surface the board sits on every 4th or 5th board to maintain level. 

I adjust the bit height by rolling a piece of .250 drill rod back and forth until the bit just clears and then dropping it one click on the servo so it doesn't quite clear.  I have vertical "home" set at Z=1.000 on the machine so I have to raise the spindle .750 before I start.  I suppose I could lower the home to .250 and eliminate the "raise" but having it at one inch gives me enough time to catch some kinds of screw-ups.

I've been delighted with this bit. I buy them from Precise Bits and they seem to last pretty well. 

EM3E8-0100-15V 0.0100in. (0.254mm) tip dia., 3-flute 15° tapered stub trace-isolation bit, 0.020 in. DOC, 1.42 in. (36mm) OAL 2 $16.95 $33.90

The size of the bur on the edge of cuts seems to increase as they wear out, but only downside of this seems to be more dressing of the finished board. I use 400 grit paper and final clean up with 800 crocus cloth.

I've been tinning the finished boards with MG Chemical 421.  I'm a bit mystified why I get better tinning if i rough up the board with 400 grit before plating it.  I wash board with acetone before tinning.

After I've soldered everything (all through hole - no smds yet), I go over the board with a loupe and remove any shorts with braid and clean out any suspicious cuts with a needle.

and they always work, at least the ones I haven't messed up the circuits on.

If you read this and see anything stupid, let me know.  Also I'm still like a better way to get rid of chips.

regards,

John ferguson  delray beach

Art Eckstein
 

Bob,
Been there and done that!  I use a probing function on a fixed tool pad that has +5V on it and a ground clip on the tool to make sure we have a good contact.  The touch pad is on G53 and I have managed to make tool offsets work in Turbocnc.  I believe others have done it in Mach and other controllers.  Another way to do it is to have a dedicated area on your blank pcb and use the electrical method to touch off and not worry about fixture offsets. 

In this case, use a probe function and call a G92 Z0 at the probe.  Raise the tool to clear everything and forge on.

Country


 At 3/21/2019 04:52 PM, bownes wrote:
Probably starting a religious war, but I've actually had better luck with engraving bits that actual mills for traces. I change over to a real mill for milling operations.

What I really need to do is master gcode well enough to have a routine that re-zero's the height after changing a bit.

Bob


On Thu, Mar 21, 2019 at 4:26 PM John Ferguson via Groups.Io <jferg977=aol.com@groups.io> wrote:

I suppose swarf isn't the best word, but I'm curious about how you guys get rid of the chips which lodge in cuts.  I've been chasing them out with a needle, buy it's tedious.

I've got 90 psi air in the shop. Maybe I should try that.

I've got my production system running very smoothly. Design In Eagle, generate Gcode in pcpcode, sometimes include autocad millwork with gcode generated in SheetCAM, set up work piece, set height on cutter and go.

I make all of my boards from 2.5x4.0 workpieces. I made an aluminum fixture with a depression which fist these workpieces and has a lid which locates 4 holes one in each corner.  I drill the holes and then screw the board down to an HDPE fixture bolted to the table on my 6040 router.  The HDPE has 4 holes which take 6-32 nylon socket head caps screws.  I face off the surface the board sits on every 4th or 5th board to maintain level. 

I adjust the bit height by rolling a piece of .250 drill rod back and forth until the bit just clears and then dropping it one click on the servo so it doesn't quite clear.  I have vertical "home" set at Z=1.000 on the machine so I have to raise the spindle .750 before I start.  I suppose I could lower the home to .250 and eliminate the "raise" but having it at one inch gives me enough time to catch some kinds of screw-ups.

I've been delighted with this bit. I buy them from Precise Bits and they seem to last pretty well. 
EM3E8-0100-15V 0.0100in. (0.254mm) tip dia., 3-flute 15° tapered stub trace-isolation bit, 0.020 in. DOC, 1.42 in. (36mm) OAL 2 $16.95 $33.90

The size of the bur on the edge of cuts seems to increase as they wear out, but only downside of this seems to be more dressing of the finished board. I use 400 grit paper and final clean up with 800 crocus cloth.

I've been tinning the finished boards with MG Chemical 421.  I'm a bit mystified why I get better tinning if i rough up the board with 400 grit before plating it.  I wash board with acetone before tinning.

After I've soldered everything (all through hole - no smds yet), I go over the board with a loupe and remove any shorts with braid and clean out any suspicious cuts with a needle.

and they always work, at least the ones I haven't messed up the circuits on.

If you read this and see anything stupid, let me know.  Also I'm still like a better way to get rid of chips.

regards,

John ferguson  delray beach

John Ferguson
 

I thought I was asking about swarf. 

On 3/21/19 5:30 PM, Art Eckstein wrote:
Bob,
Been there and done that!  I use a probing function on a fixed tool pad that has +5V on it and a ground clip on the tool to make sure we have a good contact.  The touch pad is on G53 and I have managed to make tool offsets work in Turbocnc.  I believe others have done it in Mach and other controllers.  Another way to do it is to have a dedicated area on your blank pcb and use the electrical method to touch off and not worry about fixture offsets. 

In this case, use a probe function and call a G92 Z0 at the probe.  Raise the tool to clear everything and forge on.

Country


 At 3/21/2019 04:52 PM, bownes wrote:
Probably starting a religious war, but I've actually had better luck with engraving bits that actual mills for traces. I change over to a real mill for milling operations.

What I really need to do is master gcode well enough to have a routine that re-zero's the height after changing a bit.

Bob


On Thu, Mar 21, 2019 at 4:26 PM John Ferguson via Groups.Io <jferg977=aol.com@groups.io> wrote:

I suppose swarf isn't the best word, but I'm curious about how you guys get rid of the chips which lodge in cuts.  I've been chasing them out with a needle, buy it's tedious.

I've got 90 psi air in the shop. Maybe I should try that.

I've got my production system running very smoothly. Design In Eagle, generate Gcode in pcpcode, sometimes include autocad millwork with gcode generated in SheetCAM, set up work piece, set height on cutter and go.

I make all of my boards from 2.5x4.0 workpieces. I made an aluminum fixture with a depression which fist these workpieces and has a lid which locates 4 holes one in each corner.  I drill the holes and then screw the board down to an HDPE fixture bolted to the table on my 6040 router.  The HDPE has 4 holes which take 6-32 nylon socket head caps screws.  I face off the surface the board sits on every 4th or 5th board to maintain level. 

I adjust the bit height by rolling a piece of .250 drill rod back and forth until the bit just clears and then dropping it one click on the servo so it doesn't quite clear.  I have vertical "home" set at Z=1.000 on the machine so I have to raise the spindle .750 before I start.  I suppose I could lower the home to .250 and eliminate the "raise" but having it at one inch gives me enough time to catch some kinds of screw-ups.

I've been delighted with this bit. I buy them from Precise Bits and they seem to last pretty well. 
EM3E8-0100-15V 0.0100in. (0.254mm) tip dia., 3-flute 15° tapered stub trace-isolation bit, 0.020 in. DOC, 1.42 in. (36mm) OAL 2 $16.95 $33.90

The size of the bur on the edge of cuts seems to increase as they wear out, but only downside of this seems to be more dressing of the finished board. I use 400 grit paper and final clean up with 800 crocus cloth.

I've been tinning the finished boards with MG Chemical 421.  I'm a bit mystified why I get better tinning if i rough up the board with 400 grit before plating it.  I wash board with acetone before tinning.

After I've soldered everything (all through hole - no smds yet), I go over the board with a loupe and remove any shorts with braid and clean out any suspicious cuts with a needle.

and they always work, at least the ones I haven't messed up the circuits on.

If you read this and see anything stupid, let me know.  Also I'm still like a better way to get rid of chips.

regards,

John ferguson  delray beach

Dan.Staver
 

What step size do you use and what cutter diameter?
I’ve found if I use a step of >=50% the cutter foam, I get these “leftovers”, but, if I set a step of 30-40%, it tends to be clean. Takes longer, of course.

Dan Staver
Tave Tech Corp.
3130 Hollycrest Dr.
Colorado Springs, CO 80920
+1-719-359-5352 - office
+1-719-502-1675 - cell
tavetech - Skype
W3QDO

On Mar 21, 2019, at 6:15 PM, John Ferguson via Groups.Io <jferg977@...> wrote:

I thought I was asking about swarf. 

On 3/21/19 5:30 PM, Art Eckstein wrote:
Bob,
Been there and done that!  I use a probing function on a fixed tool pad that has +5V on it and a ground clip on the tool to make sure we have a good contact.  The touch pad is on G53 and I have managed to make tool offsets work in Turbocnc.  I believe others have done it in Mach and other controllers.  Another way to do it is to have a dedicated area on your blank pcb and use the electrical method to touch off and not worry about fixture offsets. 

In this case, use a probe function and call a G92 Z0 at the probe.  Raise the tool to clear everything and forge on.

Country


 At 3/21/2019 04:52 PM, bownes wrote:
Probably starting a religious war, but I've actually had better luck with engraving bits that actual mills for traces. I change over to a real mill for milling operations.

What I really need to do is master gcode well enough to have a routine that re-zero's the height after changing a bit.

Bob


On Thu, Mar 21, 2019 at 4:26 PM John Ferguson via Groups.Io <jferg977=aol.com@groups.io> wrote:

I suppose swarf isn't the best word, but I'm curious about how you guys get rid of the chips which lodge in cuts.  I've been chasing them out with a needle, buy it's tedious.

I've got 90 psi air in the shop. Maybe I should try that.

I've got my production system running very smoothly. Design In Eagle, generate Gcode in pcpcode, sometimes include autocad millwork with gcode generated in SheetCAM, set up work piece, set height on cutter and go.

I make all of my boards from 2.5x4.0 workpieces. I made an aluminum fixture with a depression which fist these workpieces and has a lid which locates 4 holes one in each corner.  I drill the holes and then screw the board down to an HDPE fixture bolted to the table on my 6040 router.  The HDPE has 4 holes which take 6-32 nylon socket head caps screws.  I face off the surface the board sits on every 4th or 5th board to maintain level. 

I adjust the bit height by rolling a piece of .250 drill rod back and forth until the bit just clears and then dropping it one click on the servo so it doesn't quite clear.  I have vertical "home" set at Z=1.000 on the machine so I have to raise the spindle .750 before I start.  I suppose I could lower the home to .250 and eliminate the "raise" but having it at one inch gives me enough time to catch some kinds of screw-ups.

I've been delighted with this bit. I buy them from Precise Bits and they seem to last pretty well. 
EM3E8-0100-15V 0.0100in. (0.254mm) tip dia., 3-flute 15° tapered stub trace-isolation bit, 0.020 in. DOC, 1.42 in. (36mm) OAL 2 $16.95 $33.90

The size of the bur on the edge of cuts seems to increase as they wear out, but only downside of this seems to be more dressing of the finished board. I use 400 grit paper and final clean up with 800 crocus cloth.

I've been tinning the finished boards with MG Chemical 421.  I'm a bit mystified why I get better tinning if i rough up the board with 400 grit before plating it.  I wash board with acetone before tinning.

After I've soldered everything (all through hole - no smds yet), I go over the board with a loupe and remove any shorts with braid and clean out any suspicious cuts with a needle.

and they always work, at least the ones I haven't messed up the circuits on.

If you read this and see anything stupid, let me know.  Also I'm still like a better way to get rid of chips.

regards,

John ferguson  delray beach

John Ferguson
 

HI Dan,

I've been cutting my "grooves" in a single pass. I'm assuming "steps" means offsets, maybe cutting a groove and then making another pass .001 nearer the trace to clean it up. Is that it?

I had the idea of repeating the cuts assuming that the next pass would throw the swarf out of the groove. 

If I want to do steps, I assume I would need to increase trace separation in Eagle and then set the step in pcbgcode.  If this is approximately right, I'll go back to the manual and try it.

my boards, except for one, are single sided with no vias.  they have a number pf dip sockets and traces must pass between pins and there isn't a lot of space there - .010 pin separation and I like to use the long pads; tight.

what do you think?


john

On 3/21/19 9:35 PM, Dan.Staver wrote:
What step size do you use and what cutter diameter?
I’ve found if I use a step of >=50% the cutter foam, I get these “leftovers”, but, if I set a step of 30-40%, it tends to be clean. Takes longer, of course.

Dan Staver
Tave Tech Corp.
3130 Hollycrest Dr.
Colorado Springs, CO 80920
+1-719-359-5352 - office
+1-719-502-1675 - cell
tavetech - Skype
W3QDO

John Johnson
 

I had the idea of repeating the cuts assuming that the next pass would throw the swarf out of the groove.

Yes, that’s the idea. I got tired of cleaning the grooves myself.

John T. Johnson
PhD Candidate - Cognitive Motor Control Laboratory
Lab TA NEURO 2001 Principles
School of Biological Sciences
Georgia Institute of Technology

678-575-2093
john.johnson@...

On 22 Mar 2019, at 8:26, John Ferguson via Groups.Io wrote:

HI Dan,

I've been cutting my "grooves" in a single pass. I'm assuming "steps" means offsets, maybe cutting a groove and then making another pass .001 nearer the trace to clean it up. Is that it?

I had the idea of repeating the cuts assuming that the next pass would throw the swarf out of the groove.

If I want to do steps, I assume I would need to increase trace separation in Eagle and then set the step in pcbgcode.  If this is approximately right, I'll go back to the manual and try it.

my boards, except for one, are single sided with no vias.  they have a number pf dip sockets and traces must pass between pins and there isn't a lot of space there - .010 pin separation and I like to use the long pads; tight.

what do you think?


john

On 3/21/19 9:35 PM, Dan.Staver wrote:

What step size do you use and what cutter diameter?
I’ve found if I use a step of >=50% the cutter foam, I get these “leftovers”, but, if I set a step of 30-40%, it tends to be clean. Takes longer, of course.

Dan Staver
Tave Tech Corp.
3130 Hollycrest Dr.
Colorado Springs, CO 80920
+1-719-359-5352 - office
+1-719-502-1675 - cell
tavetech - Skype
dan.staver@... <mailto:dan.staver@...>
www.tavetech.com <http://www.tavetech.com>
W3QDO

John Ferguson
 

sounds like the old shampoo ad. Cut, repeat. 

While I'm at it i want to again thank you for your wonderful code.  As I wrote when I first surfaced the swarf problem, I've gotten my design and cut routine down cold - everything works.

best regards,

john ferguson

On 3/22/19 1:21 PM, John Johnson wrote:

I had the idea of repeating the cuts assuming that the next pass would throw the swarf out of the groove.

Yes, that’s the idea. I got tired of cleaning the grooves myself.

John T. Johnson
PhD Candidate - Cognitive Motor Control Laboratory
Lab TA NEURO 2001 Principles
School of Biological Sciences
Georgia Institute of Technology

678-575-2093
john.johnson@...

On 22 Mar 2019, at 8:26, John Ferguson via Groups.Io wrote:

HI Dan,

I've been cutting my "grooves" in a single pass. I'm assuming "steps" means offsets, maybe cutting a groove and then making another pass .001 nearer the trace to clean it up. Is that it?

I had the idea of repeating the cuts assuming that the next pass would throw the swarf out of the groove.

If I want to do steps, I assume I would need to increase trace separation in Eagle and then set the step in pcbgcode.  If this is approximately right, I'll go back to the manual and try it.

my boards, except for one, are single sided with no vias.  they have a number pf dip sockets and traces must pass between pins and there isn't a lot of space there - .010 pin separation and I like to use the long pads; tight.

what do you think?


john

On 3/21/19 9:35 PM, Dan.Staver wrote:

What step size do you use and what cutter diameter?
I’ve found if I use a step of >=50% the cutter foam, I get these “leftovers”, but, if I set a step of 30-40%, it tends to be clean. Takes longer, of course.

Dan Staver
Tave Tech Corp.
3130 Hollycrest Dr.
Colorado Springs, CO 80920
+1-719-359-5352 - office
+1-719-502-1675 - cell
tavetech - Skype
dan.staver@... <mailto:dan.staver@...>
www.tavetech.com <http://www.tavetech.com>
W3QDO

John Johnson
 

Glad to hear everything is working, John.
Maybe one day I’ll get back to it…

Regards,

John T. Johnson
PhD Candidate - Cognitive Motor Control Laboratory
Lab TA NEURO 2001 Principles
School of Biological Sciences
Georgia Institute of Technology

678-575-2093
john.johnson@...

On 22 Mar 2019, at 15:00, John Ferguson via Groups.Io wrote:

sounds like the old shampoo ad. Cut, repeat.

While I'm at it i want to again thank you for your wonderful code.  As I wrote when I first surfaced the swarf problem, I've gotten my design and cut routine down cold - everything works.

best regards,

john ferguson

On 3/22/19 1:21 PM, John Johnson wrote:

I had the idea of repeating the cuts assuming that the next pass
would throw the swarf out of the groove.

Yes, that’s the idea. I got tired of cleaning the grooves myself.

John T. Johnson
PhD Candidate - Cognitive Motor Control Laboratory
Lab TA NEURO 2001 Principles
School of Biological Sciences
Georgia Institute of Technology

678-575-2093
john.johnson@...

On 22 Mar 2019, at 8:26, John Ferguson via Groups.Io wrote:

HI Dan,

I've been cutting my "grooves" in a single pass. I'm assuming
"steps" means offsets, maybe cutting a groove and then making
another pass .001 nearer the trace to clean it up. Is that it?

I had the idea of repeating the cuts assuming that the next pass
would throw the swarf out of the groove.

If I want to do steps, I assume I would need to increase trace
separation in Eagle and then set the step in pcbgcode.  If this is
approximately right, I'll go back to the manual and try it.

my boards, except for one, are single sided with no vias. they
have a number pf dip sockets and traces must pass between pins and
there isn't a lot of space there - .010 pin separation and I like
to use the long pads; tight.

what do you think?


john

On 3/21/19 9:35 PM, Dan.Staver wrote:

What step size do you use and what cutter diameter?
I’ve found if I use a step of >=50% the cutter foam, I get
these “leftovers”, but, if I set a step of 30-40%, it tends to
be clean. Takes longer, of course.

Dan Staver
Tave Tech Corp.
3130 Hollycrest Dr.
Colorado Springs, CO 80920
+1-719-359-5352 - office
+1-719-502-1675 - cell
tavetech - Skype
dan.staver@... <mailto:dan.staver@...>
www.tavetech.com <http://www.tavetech.com>
W3QDO

mariob_1960@...
 

In bcnc there is a new complement (I do not know if it is already available) to clean the traces of pcb with laser (remove black paint for acid later) which I have helped with my small contribution

mariob_1960@...
 

sorry if I return with the same theme: it is useful to use your software to generate the trace layout with a bit of 0.10mm and then increase the separation between tracks with a bit of 0.7mm.
Can you generate the residual areas where the bit of 0.70 does not arrive (angles, sectors where should bit 0.7 be cut because it invades another track) that should be done by pocket operation for copper cleaning? I have managed to solve it with bcnc, but it is not a practical solution,

John Ferguson
 

I liked idea of running same bit a second time.  I'm going to try that.

thanks for your thoughts.

john

On 3/24/19 11:11 AM, mariob_1960 via Groups.Io wrote:
sorry if I return with the same theme: it is useful to use your software to generate the trace layout with a bit of 0.10mm and then increase the separation between tracks with a bit of 0.7mm.
Can you generate the residual areas where the bit of 0.70 does not arrive (angles, sectors where should bit 0.7 be cut because it invades another track) that should be done by pocket operation for copper cleaning? I have managed to solve it with bcnc, but it is not a practical solution,

mariob_1960@...
 

Hola John, a pesar de mi pequeño conocimiento de programación, consigo progresos. Volviendo al tema del uso de dos fresas, la preocupación es la siguiente, si me puede guiar: ¿en qué archivo trabajar para ver cómo se generan y pueden modificar las rutas, pcb-gcode.ulp?
El objetivo es:
1- Principal: Obtenga el diagrama de intersección en las rutas generadas por la superposición (es decir, lo que su programa elimina cuando realiza una unión sólida de dos islas con bits de gran diámetro)
Esto ya tendría más del 60% de la meta alcanzada.
2- Generar archivo con nombre de intersección.
3- El siguiente paso debe hacer una diferencia entre lo que se obtuvo y la ruta de
bit mínimo
En el peor de los casos, podría trabajar en esto desde bcnc con python
Como te dije, el objetivo que he logrado funciona bien para algunos rastros, pero es demasiado lento para una pequeña PCB
Te consulto: ¿Tu software funciona solo dentro de Eagle, o puede funcionar en diferentes programas que generan Gerber?
Lo siento si no era el lugar adecuado para estas consultas.
Abrazo mario

mariob_1960@...
 

Hi John, despite my little programming knowledge, I get progress. Returning to the issue of using two strawberries, the concern is the following, if you can guide me: on what file to work to see how the routes are generated and able to modify, pcb-gcode.ulp?
The objective is:
1- Main: Obtain the intersection plot on the routes generated by overlap (that is, what your program eliminates when you make solid union of two islands with large diameter bits)
This would already have more than 60% of the goal achieved.
2- Generate file with intersect name.
3- The next step should make a difference between what was obtained and the route of
minimum bit
In the worst case, I could work on this from bcnc with python
As I told you, the objective I have achieved, it works well for few traces, but it is too slow for a small pcb
I consult you: Does your software work only within Eagle, or can it work on different programs that generate Gerber?
Sorry if it was not the right place for these queries.
Hug Mario.