Before I tell the full story, it is perhaps best to clarify, that although this post is about failing, it does not describe a failure… Riddle me this? OK here goes
Until recently I was the proud owner of a shed bound CNC3040. The CNC 3040 is a relatively inexpensive (and not the best quality) desktop CNC. Mine was sitting unloved in the garden shed far away from power supplies and therefore cutting action. At a previous address it had pride of place on a garage workbench cutting plywood components for model rockets and other projects. But after a house move it has been languishing waiting for a loft based workshop to be built.
Well recently that changed. My workshop was recently completed and the CNC3040 has come back out to play. Trouble is I have yet to build a dust/noise containing enclosure for the thing yet. So I got round to thinking, what else can I do with a desktop CNC? Vinyl cutting, that’s what. More importantly for me, I needed a project that will test and develop my making skills (I have been a year without a workshop now), as well as let me learn a new software package (Fusion 360). The mini-project was born, an attachment for a CNC machine that adds a vinyl cutter. It would be 3D printable and let me use the CNC while I design and build and enclosure for messy and noisy cutting. It would also let me use and learn Fusion 360. I was still using OpenScad for 3D design, and I have been wanting to use something more capable and modern for some time.
I found vinyl cutters straight away. My google searching for vinyl cutter blades brought me to many e-bay vendors and Chinese online stores selling spare parts. Being unaware really as to what makes a good vinyl cutter tool, I opted for the scientific approach as to “what was cheapest/most impressive looking for the money”. The one I settled on was described as “50mm Length Small Blade Holder for Liyu Vinyl Cutter”. It is fair to say, that I had no real appreciation for how a vinyl cutter worked having never used one, but how hard can it be, it’s just a knife that moves around surely?
The first failure was not unexpected. With the cutting tool in my hand I decided to make a first prototype. The purpose of this prototype was to answer a few questions for me:
So a design was born, and of course did not work. The 2nd question on my list was the one that I failed on. I intended to place the tool holder in basically a tube, but the problem I had was it had a protruding collar. I thought I was being smart by shaping the holder in a certain way to allow me to insert the cutter then hold it tight with a tie wrap, but the tolerances were off, and the mount would not hold it tight… Oh well.
More successful was the mount to the CNC machine, a ring that tightened around the body of the spindle. I designed it with a complex tightening system that could be held in place by either bulldog clip, tie wrap or M4 nut and bolt. I ended up though settling for the nut and bolt, it was just easier and less wasteful.
As for the PLA. Well I got it working, even though my hand built 3D printer was designed (by me) to print ABS, by leaving the door open to the heated chamber I could squeeze good quality prints.
Armed with my new found knowledge and more importantly rapidly improving Fusion 360 skills, I embarked on the next design… This design sported a new tool holder, no longer awkward to load, and much more able to hold the cutter securely in place. It clamped well to the CNC machine, it was even reasonably stiff (having used a denser infill when printing it) I printed it off in an evening, though needed to sleep before I got a chance to test it…
So where was the fail? My complete misunderstanding of how a vinyl cutter works. During my lunch hour at work I began an internet search on how to set up the cut. I figured that a good place to start was the operating instructions for the vinyl cutter the tool was designed to work in. I found a YouTube video that demonstrated the setting up of the machine and then the penny dropped.
You see when I designed the accessory I had looked at photos of the machine, and the cutting head looked to be pretty solid. I guessed that it simply moved the knife up and down when it wanted to cut. This was a mistake. In the setup video, as well as setting the cut height of the blade (I had worked out that bit for myself) the user also set the pressure in grams. This was when I realised my mistake. The photos of the machine did not make it clear that the head had vertical movement. It was being moved down to cut… That bit I guessed, but it was not solidly mounted, it was free to move up and down, held down with some pressure.
So this design I never even tried.
Back to the drawing board, by now I was becoming half competent in using Fusion 360. I needed to come up with a new design that would allow constant pressure to be applied to the cutting tool. I discovered through more googling and some Facebook advice from someone who had actually used a vinyl cutter that machines were using springs to exert the pressure. This caused me a problem. Where to buy springs, and what spec would I need. The answers were online and no idea! This project was supposed to a cheap and simple diversion. Purchasing springs and testing (and perhaps failing) them was going to take time and money. I had no doubt the manufacturers of commercially available vinyl cutters would have spent time and money developing their products. I needed something cheap and adjustable.
What I opted for was a design that used gravity to apply the cutting pressure. Best of all, it used some parts that I had in my spares box… 8mm stainless steel sooth rod and LM8UU bearings. Users of 3D printers would recognise these, they are widely used in many designs.
In my design I calculated that the weight of the smooth rods (cut to a set length) plus the weight of the rest of the moving assembly would exert 90gm of pressure on the cutting knife tip… No springs would be required, and if necessary I could either shorten the rods lower the pressure or add weights to make it more… Awesome!
For the first cut I simply lowered the cutting knife onto the vinyl and using the jog controls of the CNC machine manually cut a crude shape. It worked! I got clean cuts and perfect depth adjustment. So where is the failure? One word… Bearings!
I put together some g-code using Fusion 360 (yet something else I had not done before with this software) and attempted to cut the openhardware logo. The machine did its thing, but the cut was far from perfect. Why? LM8UU bearings have allow the smooth rod to twist and move laterally in this design. And I don’t know if it is something that can be fixed. If you watch the video you can see the cutting tool moving around. This mal-forms the shapes and makes it impossible to cut detail….
So far I have considered adding a second set of bearings to constrain the movement of the rods. But to do this I need to lengthen the steel rods. I do that and the carefully designed in cutting pressure is increased. Not sure that I am going to pursue this design any further. Instead I have ordered a linear bearing that should be more accurate (I hope)
All but Edison would have to consider the design a failure. My lack of knowledge, coupled with lack of research and lack of engineering prowess have led me to spend hours designing and printing a pretty useless attachment. But I do not consider this a failure… here is why:
You can see some of my older projects at
I also document projects at