Yesterday I did a 3D model of a Celtic knot. In the last 2 minutes of a 2 hour job, 2 passes resulted in the Z axis being lower than it should have been. Nothing in Cut3D showed this happening. Over lunch today, I imported the Cut3D file into VCarve 5.5 (don't want to add 6.0 to the mix just yet) and the preview is MUCH cleaner than Cut3D. In fact, it was kinda scary how close it showed all the bumps and left over areas to the final results. I will be using VCarve from now on after generating from Cut3D. However, it did not show the grooves that appeared. These grooves only appeared on the final step, not the roughing step.
A photo can be seen here. http://www.cncsharktalk.com/download/fi ... &mode=view The problem area is the upper right hand corner.
I would rather not try trial and error method of redoing the project as it takes 2 hours. Although as I type this, I can recreate the tap file and compare it against the old to see if there are any changes.
What would be nice is to be able to take the tap file as an input file and display it graphically to see if it was generated properly.
I don't think it was the machine as it started working properly after the grooves.
The finishing file is 1.31mb. Around 17k+ lines if I remember correctly.
Troubleshooting tap file
Moderators: al wolford, sbk, Bob, Kayvon
Re: Troubleshooting tap file
Update: I found out that Google Sketchup can import xyz coordinates thru a plugin. So, I wrote a program to extract the xyz from the tap file creating a new file. It worked pretty well. 45k+ points. Crazy number.
So far not seeing anything obvious but it is difficult as performance of moving around is pretty poor.
So far not seeing anything obvious but it is difficult as performance of moving around is pretty poor.
Re: Troubleshooting tap file
That's a pretty interesting idea and innovation. So, my suggestion would be to import only the last 10% or less of that file, and you'll have less than 45K points in sketchup. You may have already thought about that.
Re: Troubleshooting tap file
I hid the areas I did not care about.
In hindsight, I should have deleted the first have the file I created of the XYZ coordinates as I knew it was the last part that I cared about. Spent about 5 hours on this project last night including reworking the original design.
Nothing was obvious with the points but doing it this way does not tell me the path between the points.
Anyway, it did show me that when I made this design in Google SketchUP, I did have lines in that corner that I did not have in the other 3. The way it was sloped was different than the other three. It did not look like it should have caused a problem, but it is something to try. So last night I reworked that corner to be more like the other 3 and tonight I will recreate the tap file and try it again.
In hindsight, I should have deleted the first have the file I created of the XYZ coordinates as I knew it was the last part that I cared about. Spent about 5 hours on this project last night including reworking the original design.
Nothing was obvious with the points but doing it this way does not tell me the path between the points.
Anyway, it did show me that when I made this design in Google SketchUP, I did have lines in that corner that I did not have in the other 3. The way it was sloped was different than the other three. It did not look like it should have caused a problem, but it is something to try. So last night I reworked that corner to be more like the other 3 and tonight I will recreate the tap file and try it again.
Re: Troubleshooting tap file
Have you tried or used MeshLab http://meshlab.sourceforge.net/. You can import Sketchup work as well as other formats and it will export to the proper file save for Cut3D. I build or model in Sketchup then export in Collada and open in MeshLab. MeshLap has a better model viewing tool in my opinion. I am sure Sketchup Pro is better tooled than the basic but I found what I needed in MeshLab without putting any $$$ out. If I was designing more then I believe it fair to buy but I'm a one-zie, two-zie a month 3D project.
Bobby
Bobby
Re: Troubleshooting tap file
Bobby,
I had seen a post somewhere (they all start to run together) about someone who used Sketchup and then MeshLab to create the file to import into Vcarve. I am currently using those steps.
However, what I was looking to find out was to see if the final toolpath in the tap file was being generated like I would expect. I wanted to see if the XYZ points in the tap file were correctly reflecting what I was seeing graphically.
So I wanted to take the tap file and pull it back into a 3D modeling program to see if I could detect why I was getting the problem. I did not even think of seeing if MeshLab had an import option of XYZ points. However, I did successfully import the tap file's XYZ points back into Sketchup after I wrote my own custom program to clean up the tap file so that the plugin I used in Sketchup could import those points.
By doing this, I did see that I had extra vectors in that corner that I did not have in the other corners. So I have redone the 3D image in SketchUp (free version) to look like the other 3 corners from a line/vector standpoint. Now I am going to run that back through and see if I have the same problem.
Adam
I had seen a post somewhere (they all start to run together) about someone who used Sketchup and then MeshLab to create the file to import into Vcarve. I am currently using those steps.
However, what I was looking to find out was to see if the final toolpath in the tap file was being generated like I would expect. I wanted to see if the XYZ points in the tap file were correctly reflecting what I was seeing graphically.
So I wanted to take the tap file and pull it back into a 3D modeling program to see if I could detect why I was getting the problem. I did not even think of seeing if MeshLab had an import option of XYZ points. However, I did successfully import the tap file's XYZ points back into Sketchup after I wrote my own custom program to clean up the tap file so that the plugin I used in Sketchup could import those points.
By doing this, I did see that I had extra vectors in that corner that I did not have in the other corners. So I have redone the 3D image in SketchUp (free version) to look like the other 3 corners from a line/vector standpoint. Now I am going to run that back through and see if I have the same problem.
Adam
Re: Troubleshooting tap file
Adam, I would do the same as you in the technique of checking the .tap file but my personal experience with Sketchup I have found some quirks (my own I am sure) that were drawn in there like a reference measuring point that became part of my model. I don't see it in Sketchup after a while because I used the "hide" feature and it doesn't show in the model but when I convert to Collada and open in MeshLab I can see my errant line 
I am all too new at CNC and especially 3D modeling but when I find a recipe that works I practically burn it into palm. Like most of us, my scrap pile is a lot taller than my accomplished pile.
I am all too new at CNC and especially 3D modeling but when I find a recipe that works I practically burn it into palm. Like most of us, my scrap pile is a lot taller than my accomplished pile.
Re: Troubleshooting tap file
Well, did the knot pattern again. Did not get the mess in the corner like last time so that was good. However, got different issues. I got some lines that went deeper than I expected, but I am not surprised. There was a 3 hour pause as we went out to eat and see a movie. I also used the .125 ball bit instead of the .25 so that "errors" were not as deep.
I see that the controller uses the points to say, "go run to that point". The smaller the spaces to between points, the slopes and curvers you have. Makes sense, it was just interesting to see the points not scroll on the control panel while the router continued to move to the next point.
In order for me to truly see what the tap file does, I would have to recreate the lines between the points to see path the router will take.
I see that the controller uses the points to say, "go run to that point". The smaller the spaces to between points, the slopes and curvers you have. Makes sense, it was just interesting to see the points not scroll on the control panel while the router continued to move to the next point.
In order for me to truly see what the tap file does, I would have to recreate the lines between the points to see path the router will take.
Re: Troubleshooting tap file
Hi Adam,
Based on your last description, I'm wondering if you're not running into a condition where the feed rate and/or plunge rate are too high for this setup, and the outcome is what you are getting. I'm still learning the behavior of the machine. I've been increasing feed rates and plunge rates to see where the knee in the curve is. One of the things I learned over the last few days, even working in pine, is that sometimes a high plunge and or a high feed rate results in the things you described.
In the case of too high a plunge rate, the bit can't cut as fast as it is travelling in negative z, so it is actually being pushed against by uncut wood. But the machine thinks it actually "made it" to the desired z, but in reality it is higher. But, it retains that thinking from that point forward. In one instance, I set the FRO to 50% at the start of the project, and let it get a few passes in before I set the FRO back to 100%. As I only had one instance of the tool starting in a new place for this particular project, I didn't have to set and catch the next "new place."
In the case of too high a feed rate, the same kinds of errors could be induced into x and y - and you get crappy cuts in either case.
I can imagine that this kind of error inducement can be exaggerated over time and possibly result in the original condition you described.
I know from reading the VCarve documentation that you can somehow go in and edit the FR and Plunge rates in the generated tool path, but haven't taken the time to figure that out yet. This would remove the need to manually adjust the FRO.
As to the effects of the pause (sounds like a perceived difference in z after restart) - I'm not surprised. I have noticed this kind of behavior between jobs (finished job, tool ended up at 0,0,0.8), stroked move to 0,0,0 and it wasn't at z0 - looked at panel and sure enough, machine has something other than 0 for z).
Of course, it could all be bugs in the tap file, but thought I'd share this finding with you as it may produce similar results - and the tap file may be fine.
Based on your last description, I'm wondering if you're not running into a condition where the feed rate and/or plunge rate are too high for this setup, and the outcome is what you are getting. I'm still learning the behavior of the machine. I've been increasing feed rates and plunge rates to see where the knee in the curve is. One of the things I learned over the last few days, even working in pine, is that sometimes a high plunge and or a high feed rate results in the things you described.
In the case of too high a plunge rate, the bit can't cut as fast as it is travelling in negative z, so it is actually being pushed against by uncut wood. But the machine thinks it actually "made it" to the desired z, but in reality it is higher. But, it retains that thinking from that point forward. In one instance, I set the FRO to 50% at the start of the project, and let it get a few passes in before I set the FRO back to 100%. As I only had one instance of the tool starting in a new place for this particular project, I didn't have to set and catch the next "new place."
In the case of too high a feed rate, the same kinds of errors could be induced into x and y - and you get crappy cuts in either case.
I can imagine that this kind of error inducement can be exaggerated over time and possibly result in the original condition you described.
I know from reading the VCarve documentation that you can somehow go in and edit the FR and Plunge rates in the generated tool path, but haven't taken the time to figure that out yet. This would remove the need to manually adjust the FRO.
As to the effects of the pause (sounds like a perceived difference in z after restart) - I'm not surprised. I have noticed this kind of behavior between jobs (finished job, tool ended up at 0,0,0.8), stroked move to 0,0,0 and it wasn't at z0 - looked at panel and sure enough, machine has something other than 0 for z).
Of course, it could all be bugs in the tap file, but thought I'd share this finding with you as it may produce similar results - and the tap file may be fine.
Re: Troubleshooting tap file
I stand corrected - changing the FR and/or Plunge rate on a generated toolpath is a feature request, not an existing feature. So, for now you can either establish that in the tool when you choose it or influence it by manipulating the FRO in the Basic Panel. Sorry to raise any hopes in that regard.