A trio of Cutter Cutting Plotter Blade Holders arrived:
Despite the name, they’re not well-suited for drag knife blades, because they’re collets gripping a 2 mm shaft. The blade doesn’t rotate unless the plotter / cutter rotates the entire holder, which is actually a thing.
I got ’em because the snout of a common ball-point pen refill measures about 2 mm:
The glob around the tip comes from plotting too fast for conditions; about 1500 mm/min works better for continuous lines and 250 mm/min improves text.
The stock MPCNC adapter has a single recess suited for Genuine Plotter Pens, but the knurled lock ring on these cheapies sticks out far enough to make them wobbly. This being an inconvenience up with which I need not put, a few lines of OpenSCAD tweak the stock STL:
The original STL is ivory, new cuts are cyan, and additions are reddish.
The two support beams are now 1.6 mm = four thread widths, for improved slicing with a 0.35 mm nozzle and a higher spring constant.
It’s by-and-large indistinguishable from the old adapter:
Which I was using upside-down, because the flange fit better.
The MPCNC works reasonably well as a pen plotter with a genuine ballpoint pen:
The OpenSCAD source code as a GitHub Gist:
| // Adding clearance for eBay collet pen holder | |
| MPCNC_OD = 12.0; // pen holder OD (matches STL curvature) | |
| MPCNC_Z = 8.9; // Z offset of pen axis | |
| Pen_OD = 11.5; // actual pen body OD | |
| ID = 0; | |
| OD = 1; | |
| LENGTH = 2; | |
| Flange = [11.5,15.7,2.2]; // actual pen body flange | |
| Locknut = [16.0,16.0,2.8]; // knurled locknut | |
| Locknut_Offset = 4.5; // flange center to locknut | |
| Wall = [41.0,4 * 0.4,9.0]; // thicker walls for more spring and better fill | |
| $fn = 32; // default cylinder sides | |
| difference() { | |
| translate([-(101.3 + MPCNC_OD/2),-111.9,0]) // put pen axis above Y axis, flange centered on X axis | |
| import("/mnt/bulkdata/Project Files/Mostly Printed CNC/Accessories/Tool Holders/MPCNC_525_Drag_Knife-1860310.STL", | |
| convexity=5); | |
| if (true) // improve holder-to-mount fit if needed | |
| translate([0,60/2,8.90]) | |
| rotate([90,0,0]) | |
| cylinder(d=MPCNC_OD,h=60); | |
| translate([0,0,MPCNC_Z]) // improve flange slot clearance | |
| rotate([90,0,0]) | |
| cylinder(d=Flange[OD] + 1.0,h=Flange[LENGTH] + 0.5,center=true); | |
| translate([0,Locknut_Offset - 0.5,MPCNC_Z]) // add locknut clearance | |
| rotate([-90,0,0]) | |
| cylinder(d=Locknut[OD] + 1.5,h=Locknut[LENGTH] + 1.5,center=false); | |
| } | |
| # translate([-(27.0 + Wall.x/2),0,0]) { // embiggen walls for higher spring constant | |
| translate([0,-24.4,0]) | |
| cube(Wall); | |
| translate([0,20.6 - Wall.y,0]) | |
| cube(Wall); | |
| } |
The Smell of Molten Projects in the Morning 
Do you have it hang so far out because those struts act like a spring to regulate pressure on the pen?
Exactly: the pressure is linearly related to the cut depth and you get enough compliance to handle a reasonable amount of misalignment.
Ryan designed it to make the pen concentric with the DW660 router spindle and the spring beams turned out to have roughly the right pressure for plotter pens at a reasonable depth of cut.
Great minds, etc.: http://scruss.com/blog/2014/04/06/making-cheap-hp-plotter-pens-yet-another-hp-gl-viewer/
I cut the ends off mine so they’d fit in a short-pen HP7470A. I found using a ballpoint pen unsatisfying, as the output was uneven. Where they’ve found the most use is with 2 mm graphite to use as guides for illustration and painting. Finding the right grade of lead is tough: HB is a little faint, 2B wears a little fast. With eye-searing lighting and a magnifier, I can just make out non-photo-blue lines, but it’s not worth the effort. Best of all, though, are carmine red Ticonderoga Erasable cores: visible, reasonably hard-wearing, erasable and cheap (if you don’t mind attacking dollar-store pencils with a craft knife).
Of course, there are real ½ mm plotter pencils with spring-loaded tips and auto-advance available new-old-stock, but each of those costs more than I paid for my plotter.
You’re a braver man than I: even thinking of cutting an ink tube sends me screaming from the workbench!
I think more pressure = better performance with ballpoint pens, because a few hundred grams produced a really nice line. The XY motion doesn’t need much compliance, so the force remains constant over the whole plot area and it looks OK.
Modulo getting nasty globs before I reduced the speed, natch.