Newbie here, but I looked through the archives and don’t see the topic addressed.
I’m trying to create recessed or raised text that follows the contour of a 3d model defined in the 3d model section of CC Pro build 514. Simple example…imagine a domed lid on a box (dome = Round/45 degrees/Add) and you want to cut a name into the dome as a pocketing-type cut.
The only way I’ve found to do this is to 3d model the text vectors themselves as a 3d model Add or Subtract operation to the existing model, but the quality of the edges of the text is horrible due to the raster toolpath used for all 3d machining.
Is there a better approach that results in clean cut edges for text vectors when following the terrain of a 3d model?
(I’ve also used the workaround of cutting the text as a simple 2D pocketing cut and just cutting it deep across the 3d surface, but this only works on shallow 3d surfaces).
I use a couple finishing passes to smooth things out a bit with the smallest ball nose bit I can lay my hands on. Step over on those finishing passes is also super important, keep it small. I use .008" with 1/8" ball and .004" on a 1/16" ball. Font type is also important, keep it simple. I’ve attached a design file with my usual approach, it may give you some ideas that may help.
do you have design examples and/or pics of the (bad) results you are getting ?
I was going to comment on stepover like @Ed.E mentioned, and add that you can also use multiple finishing passes at different angles (typically one at 0° and one at 90°), but re-reading your post I’m sure I undertand your usecase correctly.
I think the OP is referring to ‘depth sensitive’ text that cuts deeper or shallower depending on the height of the underlying surface. In Vectrics this option appears during toolpath setup as ‘project onto 3D model surface’ and does what I think is being referred to here. I don’t have CC Pro so can’t comment on what CC is capable of doing
Here’s an example…this is the base 3d model (yeah, I’m a big softy and into that I’ve modelled text vectors as components in CC Pro to get them to follow the surface (“depth sensitive text” is a good description).
Here’s a shot of 3D finish pass done with a 2mm ballnose and 0.007 stepover…run time is 61 minutes. The large text elements (“Love”) look ok, but the smaller text element (fyi, kanji) immediately below it looks like it was pecked by birds and not the clean lines it should be.
Since the text elements are not strongly aligned to any single axis, I assume another finishing pass at 90 degrees to the first will result in some improvement, but still not nice and clean.
hm so the tool I built for the friday evening project took an STL as input for the 3D model… but it;s not that much more complex to take gcode as input I think… how hard could it be (most dangerous words in computer science)… if I had that then you could just make the text as a flat design and then “modulate” it on top of (or rather into) the gcode for the heart.
I’ll ponder what it’d take… probably an hour or two coding at least.
yeah that’s the idea.
surprisingly the hard part is knowing the size of the various tools in the gcode of the first cut. As a general thing that’s not viable, but carbide create puts in some special comments that have some of the tool info at least… so very likely this is only going to work with gcode created by carbide create.
(or similar tools that put in the same or similar comments about tool info)
Just curious, what feeds and speeds are you using the finishing pass? One can often push them very far in cases like this “simple” round surfaces where thé finishing pass removes little material
I’m running a very conservative 35 ipm to avoid any issues. Could probably go much faster, especially for the 2nd pass. I’m ok trading off speed for quality, but after the second finish pass I didn’t get either.
As predicted text that’s strongly aligned to the direction of the cut saw improvement from a second pass, but not so much the angled characters.
I’ll have to clean up some of those lines with a veining tool.
I’d try 100ipm (x3) and reducing the stepover by x3: for the same runtime, you should end up with much less artefacts.
but anyway by now @fenrus must be almost done
G code was invented in the 1950s, while Wikipedia tells a different story the first one I heard was that Gerber Scientific (the baby food company) developed it to drive their pen plotters and is where the G came from. It has it’s own standard, RS-274x, and multiple attempts to come up with a “better” way have all failed. OT: Gcode was originally punched into paper tape (hence the rewind commands), I spent some time developing a paper tape utility (using a paper tape punch on a Univac 9300 connected to an Univac 1110) at the NASA JSC (then MSC) for CNC machines in the early 70s, although like my previous summer job elsewhere they still wouldn’t let the smartass punk kid anywhere near the actual hardware. Mylar/aluminum/mylar sandwich was the material of choice, way outlasted paper.
So, just to be clear, you’re looking to define a vector and then project it onto the model and cut it at some defined depth relative to the model’s surface. This would be a single-line font/engraving, not a pocketing operation?
has the early version of the tool. It only takes carbide create (regular/pro) gcode nothing else,
and the latest update is 10x too slow but fixes some math accuracy issues… patience required.
I’m currently cutting a design that I’ll stick in the gallery if it comes out ok
and into that I’ve modelled text vectors as components in CC Pro to get them to follow the surface (“depth sensitive text” is a good description).
