--- In DynoMotion@yahoogroups.com, <dynomotion@yahoogroups.com> wrote:

Hi EC,

Strange. I can't think what that might be if not the Collinear Tolerance.

Please send the GCode and all of your Trajectory Planner Settings including Resolution, Vel, Acceleration.

When you look at the G-code viewer are you looking after an actual run or after a simulation?  You might do both and compare.  Simulation shows more the original GCode where after an actual run it plots the results after Collinear Tolerance, Corner rounding, segment divisions, etc...

There is a C example that you can run to capture the final generated trajectory to a disk file.  CaptureXYZMotionToFile.c

When debugging I assign this to an MCode so I can easily capture specific portions of a GCode file.  You can then import and plot the data in a program like Excel to view the exact path that was being generated.  If the generated Trajectory doesn't show a problem you can also change the program to also capture what the axes actually did (Axis Position) if your system has encoders.  

Regards

TK

---

From: ericncn <ericnc@...>
To: DynoMotion@yahoogroups.com
Sent: Monday, August 5, 2013 12:15 AM
Subject: [DynoMotion] Re: what you see in G Code viewer is not what you get

 

Hi TK,

my Collinear Tolerance was set to 0.0005
then I changed it to 5e-006
but the result is basically the same.

The problem initially showed machining an arc with an endmill having a diameter 0.95 times the diameter of the arc being machined.
Then I tried a smaller endmill (half of the diameter of the arc being machined) but the result didn't change, and reducing the Collinear Tolerance didn't change it as well.
Also, I've machined several parts and the result is still always the same.
I mean there's no randomness and it's not due to vibrations or lack of stiffness in the machine. The resulting "polygon" is always the same.

I would think it's wrong G-code but the graphical G-code viewer in KmotionCNC shows the arc nicely smooth even if magnify it a lot hence I think the input G-code is correct.

Any ideas?
I can send the G-code of that arc, if that helps

EC

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi EC,
>
> A large Collinear Tolerance could cause something like that.  What are your Trajectory Planner Settings?
>
> Regards
> TK
>
>
>
> ________________________________
> From: ericncn <ericnc@...>
> To: DynoMotion@yahoogroups.com
> Sent: Sunday, July 28, 2013 6:27 AM
> Subject: [DynoMotion] what you see in G Code viewer is not what you get
>
>
>
>  
> I am trying to mill a curve (2D milling of a contour) and the result looks a bit "polygon-ized" (nothing dramatic, but after a close inspection moving the machined part against light and looking for reflections, you can see the curve (roughly 90 degrees) is made by 3-4 pieces that are flatter than they should be and nicely blended together. No sharp angles).
>
> The same curve shows nicely in KmotionCNC's G Code viewer instead, even if I magnify it a lot I can see no artifacts so I think the G-code is OK.
>
> I tried reducing the "collinear tolerance" by 100 times in the trajectory planner screen, but the result doesn't change.
>
> Is anything else I could do to solve this issue?
>
> EC
>

--- In DynoMotion@yahoogroups.com, <dynomotion@yahoogroups.com> wrote:

Hi EC,

Strange. I can't think what that might be if not the Collinear Tolerance.

Please send the GCode and all of your Trajectory Planner Settings including Resolution, Vel, Acceleration.

When you look at the G-code viewer are you looking after an actual run or after a simulation?  You might do both and compare.  Simulation shows more the original GCode where after an actual run it plots the results after Collinear Tolerance, Corner rounding, segment divisions, etc...

There is a C example that you can run to capture the final generated trajectory to a disk file.  CaptureXYZMotionToFile.c

When debugging I assign this to an MCode so I can easily capture specific portions of a GCode file.  You can then import and plot the data in a program like Excel to view the exact path that was being generated.  If the generated Trajectory doesn't show a problem you can also change the program to also capture what the axes actually did (Axis Position) if your system has encoders.  

Regards

TK

---

From: ericncn <ericnc@...>
To: DynoMotion@yahoogroups.com
Sent: Monday, August 5, 2013 12:15 AM
Subject: [DynoMotion] Re: what you see in G Code viewer is not what you get

 

Hi TK,

my Collinear Tolerance was set to 0.0005
then I changed it to 5e-006
but the result is basically the same.

The problem initially showed machining an arc with an endmill having a diameter 0.95 times the diameter of the arc being machined.
Then I tried a smaller endmill (half of the diameter of the arc being machined) but the result didn't change, and reducing the Collinear Tolerance didn't change it as well.
Also, I've machined several parts and the result is still always the same.
I mean there's no randomness and it's not due to vibrations or lack of stiffness in the machine. The resulting "polygon" is always the same.

I would think it's wrong G-code but the graphical G-code viewer in KmotionCNC shows the arc nicely smooth even if magnify it a lot hence I think the input G-code is correct.

Any ideas?
I can send the G-code of that arc, if that helps

EC

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi EC,
>
> A large Collinear Tolerance could cause something like that.  What are your Trajectory Planner Settings?
>
> Regards
> TK
>
>
>
> ________________________________
> From: ericncn <ericnc@...>
> To: DynoMotion@yahoogroups.com
> Sent: Sunday, July 28, 2013 6:27 AM
> Subject: [DynoMotion] what you see in G Code viewer is not what you get
>
>
>
>  
> I am trying to mill a curve (2D milling of a contour) and the result looks a bit "polygon-ized" (nothing dramatic, but after a close inspection moving the machined part against light and looking for reflections, you can see the curve (roughly 90 degrees) is made by 3-4 pieces that are flatter than they should be and nicely blended together. No sharp angles).
>
> The same curve shows nicely in KmotionCNC's G Code viewer instead, even if I magnify it a lot I can see no artifacts so I think the G-code is OK.
>
> I tried reducing the "collinear tolerance" by 100 times in the trajectory planner screen, but the result doesn't change.
>
> Is anything else I could do to solve this issue?
>
> EC
>

--- In DynoMotion@yahoogroups.com, <dynomotion@yahoogroups.com> wrote:

Sorry for answering so late, was away from the machine.

Now I had the chance to veryfy that:
1) yes, the G-code viewer isn't showing same thing in a simulation and in an actual run of the same G-code.  My curve looked smooth in an actual run but doing a simulation, it looks "polygonalized" same as the actual part machined

and in fact

2) once I pinned down the piece of G-code relative to that curve, I found it was actually made by a sequence of segments and not tangent arcs like I believed.

So, sorry for making you waste time, after looking at it better it looks KmotionCNC has no fault, the poligonalization happens elsewhere.

Still, I'm not sure that the fact the Gcode viewer displays differently the same Gcode depending on it being a simulation or an actual run, is good.

EC

--- In DynoMotion@yahoogroups.com, <dynomotion@yahoogroups.com> wrote:

Hi EC,

Strange. I can't think what that might be if not the Collinear Tolerance.

Please send the GCode and all of your Trajectory Planner Settings including Resolution, Vel, Acceleration.

When you look at the G-code viewer are you looking after an actual run or after a simulation?  You might do both and compare.  Simulation shows more the original GCode where after an actual run it plots the results after Collinear Tolerance, Corner rounding, segment divisions, etc...

There is a C example that you can run to capture the final generated trajectory to a disk file.  CaptureXYZMotionToFile.c

When debugging I assign this to an MCode so I can easily capture specific portions of a GCode file.  You can then import and plot the data in a program like Excel to view the exact path that was being generated.  If the generated Trajectory doesn't show a problem you can also change the program to also capture what the axes actually did (Axis Position) if your system has encoders.  

Regards

TK

---

From: ericncn <ericnc@...>
To: DynoMotion@yahoogroups.com
Sent: Monday, August 5, 2013 12:15 AM
Subject: [DynoMotion] Re: what you see in G Code viewer is not what you get

 

Hi TK,

my Collinear Tolerance was set to 0.0005
then I changed it to 5e-006
but the result is basically the same.

The problem initially showed machining an arc with an endmill having a diameter 0.95 times the diameter of the arc being machined.
Then I tried a smaller endmill (half of the diameter of the arc being machined) but the result didn't change, and reducing the Collinear Tolerance didn't change it as well.
Also, I've machined several parts and the result is still always the same.
I mean there's no randomness and it's not due to vibrations or lack of stiffness in the machine. The resulting "polygon" is always the same.

I would think it's wrong G-code but the graphical G-code viewer in KmotionCNC shows the arc nicely smooth even if magnify it a lot hence I think the input G-code is correct.

Any ideas?
I can send the G-code of that arc, if that helps

EC

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi EC,
>
> A large Collinear Tolerance could cause something like that.  What are your Trajectory Planner Settings?
>
> Regards
> TK
>
>
>
> ________________________________
> From: ericncn <ericnc@...>
> To: DynoMotion@yahoogroups.com
> Sent: Sunday, July 28, 2013 6:27 AM
> Subject: [DynoMotion] what you see in G Code viewer is not what you get
>
>
>
>  
> I am trying to mill a curve (2D milling of a contour) and the result looks a bit "polygon-ized" (nothing dramatic, but after a close inspection moving the machined part against light and looking for reflections, you can see the curve (roughly 90 degrees) is made by 3-4 pieces that are flatter than they should be and nicely blended together. No sharp angles).
>
> The same curve shows nicely in KmotionCNC's G Code viewer instead, even if I magnify it a lot I can see no artifacts so I think the G-code is OK.
>
> I tried reducing the "collinear tolerance" by 100 times in the trajectory planner screen, but the result doesn't change.
>
> Is anything else I could do to solve this issue?
>
> EC
>

--- In DynoMotion@yahoogroups.com, <dynomotion@yahoogroups.com> wrote:

This is a good example of why I ended up changing the KMotionCNC software to produce different colors for arc and lines.  Mach3 does it and several times it allowed me to catch an error where I thought a curve was setup to use a curve but in fact was a series of lines.

--- In DynoMotion@yahoogroups.com, <dynomotion@yahoogroups.com> wrote:

Sorry for answering so late, was away from the machine.

Now I had the chance to veryfy that:
1) yes, the G-code viewer isn't showing same thing in a simulation and in an actual run of the same G-code.  My curve looked smooth in an actual run but doing a simulation, it looks "polygonalized" same as the actual part machined

and in fact

2) once I pinned down the piece of G-code relative to that curve, I found it was actually made by a sequence of segments and not tangent arcs like I believed.

So, sorry for making you waste time, after looking at it better it looks KmotionCNC has no fault, the poligonalization happens elsewhere.

Still, I'm not sure that the fact the Gcode viewer displays differently the same Gcode depending on it being a simulation or an actual run, is good.

EC

--- In DynoMotion@yahoogroups.com, <dynomotion@yahoogroups.com> wrote:

Hi EC,

Strange. I can't think what that might be if not the Collinear Tolerance.

Please send the GCode and all of your Trajectory Planner Settings including Resolution, Vel, Acceleration.

When you look at the G-code viewer are you looking after an actual run or after a simulation?  You might do both and compare.  Simulation shows more the original GCode where after an actual run it plots the results after Collinear Tolerance, Corner rounding, segment divisions, etc...

There is a C example that you can run to capture the final generated trajectory to a disk file.  CaptureXYZMotionToFile.c

When debugging I assign this to an MCode so I can easily capture specific portions of a GCode file.  You can then import and plot the data in a program like Excel to view the exact path that was being generated.  If the generated Trajectory doesn't show a problem you can also change the program to also capture what the axes actually did (Axis Position) if your system has encoders.  

Regards

TK

---

From: ericncn <ericnc@...>
To: DynoMotion@yahoogroups.com
Sent: Monday, August 5, 2013 12:15 AM
Subject: [DynoMotion] Re: what you see in G Code viewer is not what you get

 

Hi TK,

my Collinear Tolerance was set to 0.0005
then I changed it to 5e-006
but the result is basically the same.

The problem initially showed machining an arc with an endmill having a diameter 0.95 times the diameter of the arc being machined.
Then I tried a smaller endmill (half of the diameter of the arc being machined) but the result didn't change, and reducing the Collinear Tolerance didn't change it as well.
Also, I've machined several parts and the result is still always the same.
I mean there's no randomness and it's not due to vibrations or lack of stiffness in the machine. The resulting "polygon" is always the same.

I would think it's wrong G-code but the graphical G-code viewer in KmotionCNC shows the arc nicely smooth even if magnify it a lot hence I think the input G-code is correct.

Any ideas?
I can send the G-code of that arc, if that helps

EC

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi EC,
>
> A large Collinear Tolerance could cause something like that.  What are your Trajectory Planner Settings?
>
> Regards
> TK
>
>
>
> ________________________________
> From: ericncn <ericnc@...>
> To: DynoMotion@yahoogroups.com
> Sent: Sunday, July 28, 2013 6:27 AM
> Subject: [DynoMotion] what you see in G Code viewer is not what you get
>
>
>
>  
> I am trying to mill a curve (2D milling of a contour) and the result looks a bit "polygon-ized" (nothing dramatic, but after a close inspection moving the machined part against light and looking for reflections, you can see the curve (roughly 90 degrees) is made by 3-4 pieces that are flatter than they should be and nicely blended together. No sharp angles).
>
> The same curve shows nicely in KmotionCNC's G Code viewer instead, even if I magnify it a lot I can see no artifacts so I think the G-code is OK.
>
> I tried reducing the "collinear tolerance" by 100 times in the trajectory planner screen, but the result doesn't change.
>
> Is anything else I could do to solve this issue?
>
> EC
>