Big CNC as a Plotter

Sorry about that, it looks like I lost some progress in Fusion. The collet should be able to slide into the bottom piece without gluing. Both top and bottom should have snap fits that hold the spring in place:

Putting a pencil in the collet piece and sliding that into the bottom holds the pencil in place. The top and bottom are held together by the spring that is held in place by the snap fits.

It has been my plan to get a spring that is shorter and made from a thicker gauge of wire to add stability to the assembly for the next iteration, V1 was built around the only spring Home Depot had with an inner diameter big enough for a writing instrument.

Here’s a version that might have been closer to the final version, with access properly set this time.

This is still a work in progress, but not something I’ve done much with lately.

I’d love to have a version of this that uses a sharpie. It’d be nice to be able to write a series of annotations on parts before cutting them. I even have some sharpie minis (somewhere) that’d be perfect for this.

Yeah a pencil generally needs to be applied at an angle like that- 45 deg or more- but then it has to caster. But castering makes it do a twist on corners and the point of contact is behind the axis of rotation on the spindle. And if the travel just reverses straight back, there’s no telling if it will caster CW or CCW and it can fail to caster for a short distance and stab into the work instead.

Castering requires the point of contact at the tip can’t be directly under the axis of rotation. It has to lag or else it won’t rotate and follow reliably.

A marker could be mounted vertically with no caster, but it will wear the tip real fast. A ballpoint could do it, in theory, but ballpoints may not mark as well as you want

My goal for a sharpie is to stamp dots in areas that are safe for hold downs. Writing was never the intention of this design beyond an X/Y reference line. I won’t be trying to solve for the issues Danny raises.

but it will wear the tip real fast.

Good point Danny. I’ll view the marker as a consumable that I’ll be killing after a few uses. That’s kind of why I was hoping for the spring loaded version.

Check Mcmaster Carr for the spring. Its easiest to use on something like this.

I’ve not tried it but I like the idea of using a CNC as a plotter to mark out the design before cutting. It looks like you guys have a good prototype, but another solution might be treat the pen/pencil/marker like a drag knife allowing it to sit at an angle for drawing lines… CAM programs can (I think) create the X/Y and rotational offsets.

@Iammikecohen,
Thanks for v2, I’ve downloaded it and am printing it now. Would you have the time and ability to put a link to the specific Home Depot spring that you are using for your design?
@bwatt.
P.S. @EricP, thanks I had forgotten about McMaster-Carr. What I need to know now is the spring size that Mike used to find an equivalent at McMaster-Carr.

To all listening in on this conversation, I’m early in the experimental stage and so far I haven’t had time on the big CNC to use any of the pen holder designs yet. However, I’m listening to all comments and appreciate the feedback. Caster vs no-caster is still an open question. Nevertheless I especially appreciate @dannym’s input because of his deep expertise.

Follow-on thoughts and musings… I’ve been thinking about how to create the Gcode to drive the CNC. Presently I use a CAD program and it has the ability to create a flat 2D projection of the 3D design. Then I can export this projection to an SVG file. From that there are programs on the Internet to convert SVG to Gcode which is what I hope to use to drive the big CNC to be a plotter. Does this make sense? Thoughts?

Side note… Here is an existing pen mount for a CNC router and instructions for using it with VCarve, which we use for the Laser cutters, but not for the big CNC. However, the instructions are informative. Should we put VCarve on the big CNC? Personally I don’t think so. Since the big CNC is a Gcode machine I’d prefer to drive it using Gcode. What do you think?

I want to build this one. Stuff Made Here

Nice one! I’ve enjoyed “Stuff Made Here” for a while now, but this one is especially enjoyable. Do you think this is even possible in our environment?

Sorry to be late to this thread and I love all the creative ideas and solutions but when I want safe hold down points marked on material for the CNC router I just have a layer in my model that has 1/8” holes marked and I use the CNC to just kiss those points using a 1/8” bit and then get up on the table and drill screws in to hold the material down.

1 Like

This is allowed but I would remind folks to use the kreg pocket jig screws or other high quality screw. Cheap drywall screws break and it’s a big pain to remove

@bwatt Vcarve cannot be “put” in a machine, it does not drive the axes. VCarve generates gcode for the machine.

Since the pen does caster like a drag knife, that means the marking spot lags the spindle axis center, so you may want to use the “drag knife” function under Fillets. This keeps from rounding off corners as the lagging knife tries to follow as the spindle has to take a 90 deg turn. However, it’s not perfect.

Also you’d select the “drag knife” postprocessor on the machine, as it omits the Spindle Start command from the final G-code.

@dannym, James told us in class that you want to keep the LinuxCNC machine isolated from Vcarve and I understand that entirely. Oops I goofed up above. What I meant to say is “Should we put Lightburn on the VCarve machine?”. And then I immediately withdrew the question.

Well I came in yesterday night to make my first stumbling attempts at getting the big CNC to be a plotter. Here are some notes and thoughts about what happened.

I removed the dust-brush magnetic lucite plate and attached the pencil and its mount to the router. (See IMG_2530 above for what it looks like). It somewhat fit, but I’ve got to redesign it a bit by making the conical section slightly taller. The wrap-around loop that tightens onto the router housing worked well. However I have to redesign how the ER20 collet and nut is held, because I blocked access to the router shaft for tightening the nut. The plastic spring that’s part of the mount also worked well. I could even use the Z-probe to set the Z-home.

Although the pencil is at an angle and the mount is firmly attached to the router itself, that is, the mount does not move or swing as the pencil is dragged, I had no problem manually drawing simple lines and squares using the handheld CNC controller.

I mounted a scrap 2’x4’ panel to the table and put tracing paper on it. I was ready to plot!

Next I tried a “flattened” SVG file from my CAD system (FreeCAD). Over on the VCarve machine I fed in the SVG file and looked around to see what toolpath I could use. I chose “engrave” and generated the GCode. I “sneaker-net” the USB drive to the CNC machine next to it. When it came up on the CNC screen the origin was off the table. Not good. It had been at 0,0 in CAD. What did VCarve do to it? So back and forth I went from VCarve machine to CNC machine until I had finally positioned it ON the table. Not knowing exactly where VCarve thought home was, i had to again position the CNC home over-and-over until it stopped giving me limit errors on the CNC screen. Next, when I ran the GCode I forgot about the spindle speed and quickly stopped it when the spindle started-up. Luckily I hadn’t been able to tighten the collect otherwise it would have torn itself to pieces. On the CNC machine I overrode sprindle percentage to 0%. Please remember that I’m new to all this so it takes me time to reason out what went wrong and to attempt various ways to correct the issue. Bottom line, after resolving home & spindle the pencil would move, lower but then stop and go no further. Ugh. No matter what I changed it would get stuck in the same GCode. I can’t decode GCode yet so I gave up and I tried plan B.

My next attempt was to just use VCarve directly. In other words, don’t use my SVG, but a simple drawing directly with VCarve. I drew a circle, star, curvy line and text. Then back through the pipeline that I was getting to know quite well: select toolpath, calculate toolpath, save to USB drive, unplug USB on VCarve, plug USB into CNC machine, find mount, load GCode, and attempt to run it. In time I got it to start drawing on the paper. It would draw each geometric shape three times on top of itself (which I assumed was caused by the “engrave” toolpath). It did the text, the circle, and started the star and then it would stop. It never drew the curvy line. But it was beginning to work!

In the meanwhile another ASMBLY member stopped by and suggested I use a Drag Knife Toolpath. So onto Google I go to look up VCarve Drag Knife. I watch a short Youtube video about it. I find it’s a VCarve Gadget and locate it on the menus. I try to use it, but I’m confused about how to save the Toolpath. Finally I find save-toolpath. And then went through my pipeline to the CNC. However when I run the GCode it does the initial move and stops on a GCode line. Nada! Repeat. Same thing! Ugh.

It’s late. I close up, clean up, and go home.

Summary/To-Do’s

  • Rework Mount
  • SVG to GCode is my stumbling block - do I need specific GCode for this application?
  • Be able to interpret GCode when things stop
1 Like

When I suggested the Drag Knife tool, I didn’t realize that the pencil was fixed; I through it would rotate behind the motion like a drag knife does. Since it doesn’t, that probably won’t help you.

(That the gcode for the drag knife stops is odd and a concern for me, since I’m considering making one for this CNC.)

It needs to caster so the pencil angle is correct. Otherwise it stabs forward and digs in in some directions. But even with a caster, the drag has problems at high reversal angles- e.g. you go one way with the knife dragging correctly, but if the G-code doesn’t have any special toolpath tricks, it could do a straight 180 reversal that makes the tool dig in then anchor itself at some point and twist around that point, digging in hard, before assuming the opposite drag angle again.

Castering does create its own problems. It is possible to have a servo-driven knife on a cutter that drives its angle to a commanded point.

I’m away for a bit. But I hear ya and need to rethink things based upon your comments.

I’m working on the drag knife I mentioned, and I think something similar might work for this. The upper section with the bearings would be the same. The lower part would be in two pieces with springs in between: one with the shaft, the second holding the pen. See my conceptual part here: Puzzle Prototype - #11 by mgmoore