The stock spoilerboard included with the MDF version of the Mega V uses t-nuts for workholding, but they’re pretty sparse to begin with (example here)
I knew that I wanted to add a supplemental wasteboard that could be surfaced and replaced easily, and started looking for other options.
From what I can tell, the most common options here are:
No special workholding: Screw, glue, nail, or even plastic pin nail in your pieces and/or workholding devices directly into the spoilboard. This means that setting up the wasteboard is easy and inexpensive, but it does suffer some extra wear.
T-nut and holes: Pretty easy, plenty strong, not too expensive.
Aluminum t-track slots: The track is fairly easy to set up, but does require some extra t-track hardware, as well as the tracks themselves. The board between the tracks can use strips, allowing better use of large material sheets.
I was planning to use aluminum t-tracks, which were only marginally more expensive than the t-nuts, but stumbled on a couple of examples of using dovetail grooves instead:
These slots have a couple of cool features. They’re cheaper (depending on what you use for fixture hardware) and allow for very dense workholding, in any direction, at any spacing.
They do have some (significant) disadvantages though, some I recognized during planning, and some only later.
When using these grooves with MDF wasteboard, you must, 100% of the time, use clamps which press directly against the wasteboard surface at the clamping point, or the MDF can absolutely shear out. I was easily able to tear out an entire 4” square section by hand with the clamping pressure spread away from the slot. This eliminates many of the most common types of clamps.
I’ve reduced my available space a few ways with this board. First, by putting another 3/4” board on top of the existing wasteboard, I’ve lost a pretty good chuck of my machinable Z height. I’m already planning a new stock and secondary wasteboard to reduce the height.
In order to route the slots with the CNC itself, using a 1/2” dovetail bit, I needed a little clearance on all sides for the bit to enter and exit the edge of the board, so this reduced my X and Y range as well.
Lastly, I wanted the router to overhang the end of my spoilboard a bit so that I can (hopefully) cut dovetails in upright stock, so I slid both spoilboards back a couple of inches, which reduces the Y range a little more.
I have a bit of extra room, but I’m limited it how many times, and how deeply I can resurface my wasteboard before the inserts no longer fit correctly. I can reprint them easily when that happens though.
Ease of use
While these can be easy to set up, they rely on a pretty good fit inside the slot. This means, though, that they’re harder to get in and out when the slots fill up with chips.
They get stuck
More on this later, still trying to figure it out, but I often have to unjam these from the slots after use.
First attempt (spoiler)
First, I cut a section of MDF, a bit oversized, and temporarily fastened it to my wasteboard using drywall screws around the outside edge. I really want to find a better fastener for this but haven’t yet.
Then, I set up some operations in Fusion 360. First, I surfaced the area of the wasteboard that the CNC could reach, enough to take care of any real high spots and/or warping. This also showed just how out of tram my router was, with a huge nod forward. After some investigating, most of this nod was coming from badly off-square router mount. Once that was shimmed back to square, there was enough play at the router to get it close to trammed using a simple tramming tool.
Then, using a 90 degree chamfer bit, I made some spot markings for the holes that would fasten the new spoilboard to the stock one.
These I then hand drilled, deeply countersunk, dropped some CA glue into for a little more longevity, and finally fastened down.
I then ran a 2D contour toolpath to cut the excess MDF free and removed the temporary screws.
After the wasteboard was fastened and cut free, I made some relief cuts for the dovetails just under the final depth with a 1/4” downcut endmill, and then ran the dovetails in the same slots.
Here I had some trouble, and had to start over.
First, with the relief cuts and dovetails, the MDF wanted very badly to curl up. Some of my screws were around 5” from the edge of the board, and that was enough for it to want to lift.
Secondly, somewhere along the way, I lost some steps in the Y direction, and a couple of my dovetail slots were both the wrong shape and size (due to them not being centered on the relief cuts).
I went back to fusion and moved the screw holes around. I increased the density of the screws, and made sure they were fastened close to the edge as well.
Since the process of manually drilling these was such a pain, I also set up a toolpath to bore these down to final depth instead of just marking them, leaving me to only need to countersink, drill, CA glue, and screw each of them, which was still a pain, but one step fewer.
Afterward, I ran through the same steps as before. I’m not sure when the missed steps happened. Maybe I crashed something. At any rate, I didn’t end up with any missed steps for the second attempt.
I also took the time to set up an extra toolpath to cut beveled openings in all of the slots. It was really satisfying to see these run successfully.
After assembling the Mega V, I ordered some endmills and MDF for a supplemental wasteboard, and while waiting for those to arrive, I made a few pen holders to get a feel for the machine and sotfware.
For software, I settled on CNCjs because I had a spare raspberry pi available, and it’s open-source and easy to modify.
I’m currently connecting to it with my phone, but have tried with an old android tablet with mixed success.
For small touchscreens, this pendant interface is probably your best bet: cncjs-shopfloor-tablet. It features a simplified interface designed for touchscreens, making it harder to crash while jogging with tiny buttons (though I’ve still managed to).
I ziptied a sharpie to the router first. This kind of works, but since I hadn’t surfaced my spoilboard yet, and it’s a bit warped, the pen can’t follow its contours. Heres what slowly destroying a sharpie looks and sounds like.
To add a little more flexibility (and because I was still waiting for endmills), I modified a design for quick release router mount to allow for some modular attachments, and then started iterating on pen holders for it.
And here it is with a small bracket to hold a sharpened nail (I used it to measure and calibarte motor steps).
Pen holder the first
This one is very simple. It uses a collar with a set screw to keep the pen from sliding out, and a cop with a slot to hold a rubber band for tension. The pen can move up and down a bit in case of uneven surfaces, and getting the heigh zeroed out exactly isn’t critical.
Here it is running some gcode. I generated the paths for this by passing a photo to blackstripes, and then importing the SVG into Fusion 360 and then running a contour toolpath with no cut compensation.
This pen has a bit of a taper though, so the holes that guide it are a bit oversized, and this causes pen backlash when drawing.
Also, lots of little dots without much sideways movement prevents the ink from flowing a bit. I’d want to find a way to easily skip paths under a certain minimum distance.
Pen holder the second
To reduce some of the backlash, I used a couple of bolts with smooth shafts to guide a pen assembly up and down, without relying on the pen’s diameter. I also tried adding a bit of an angle to the pen, more in line with a natural writing angle. However, this applied a lot of friction between the bolts and the 3d printed plastic when pressing down due to the torque it applied, and the pen got stuck very easily.
Pen holder the third
I reused the sliding base for this one, but turned the pen back to straight upright, so that it can slide up and down more easily.
This gave the best result so far, but you can still see some of the play in the system as it “flicks” back into position at the end of each line.
Here’s a comparison of the line quality
The issue with this last design is that it both relies on a tight fit between the sliding plastic carriage and the bolts, but is also hampered by any friction there. To reduce play, these holes should be as tight as possible, but then pen won’t be able to move up and down smoothly.
I’ve purchased some teeny, tiny linear rails and will attempt to replace the moving components of this last design with them instead when I get the chance.
I have a small workspace in the corner of my garage, so naturally I ripped out most of it to make room for the 35” machine instead of something smaller and more practical for my space.
When I bought it, it wasn’t scheduled to ship for a few months While waiting, I took a stab at a basic table design to keep the machine at a comfortable working height, and to allow some storage for tools and dust collection.
This is a very simple 2x4 and 4x4 box (in line with my woodworking ability), with a plywood top and MDF sides. Eventually, the front will have cabinet doors, and the top will be enclosed with a hinged lid to keep the sound level down.
I went back and forth about putting the whole thing on casters, whether they should be raisable so that the box is sitting on solid wood when stationary, how much the box and table would weigh, and most importantly how well they could be used with the table tucked into a corner when only of the feet/casters might be accessible.
They work well, but are difficult raise and lower once weighed down. I’d like to replace them but haven’t yet found an alternative.
Got all my parts. I think I was missing a couple of screws, and there was a little shipping damage to the MDF spoilboard
I put together what I could before building a cabinet for the CNC but it didn’t take long.
I started building the table right around the time of my state’s stay-at-home order. I let my local big box store select and deliver some of the worst warped, cupped, and twisted, wet wood, I’ve seen. It was still cheaper than my local lumberyards (by a lot) but quality isn’t great unless you can spend time picking through their stacks.
Mega V assembly
Assembly was easier than I expected, largely owing to the prebuilt control box. I think the longest part was just running cables through the cable chains.
There are some issues getting the Mega V set up that I (and other people) have run into, that I wish I’d known about when I started.
At the time of building (and writing), the assembly instructions are limited to two assembly youtube videos. They’re not bad, but some sections are not clear, or slightly out of order (esp related to the endstop switches) and in general it’s a pain in the butt to watch video, pause, assemble, rewind and repeat, when written instructions with diagrams would have been easier in most cases.
The instructions call for hammering these into the v-wheels. I found it was much easier to use this 3d-printed bearing press. I printed this ages ago, and I use it for all kinds of things now. This may be be the first time I’ve used it to press bearings into something though.
There are a couple of issues I ran into and have seen other people mention regarding the wiring, particularly for the endstop switches. The crimped connectors are low quality and badly crimped, and the wires themselves are going to be too short to reach the control box when routed through the drag chains as instructed (at least on the 35” model).
The solution to both of these is to rewire them. I ended up running new wire for one and resoldering the aviation connector, and then just spliced an extension onto another.
Rack and pinion meshing
This isn’t really an issue with the machine, per se, but there doesn’t seem to be a good way to determine whether the rack and pinion meshing on the X and Y axes is correct, or correct enough. It’s easy with hand pressure to warp the rack slightly where it’s fastened to the aluminum extrusion, and the gantry can lock up or miss steps.
Router mount squaring
I’ve heard that the aluminum router mount is waterjet cut, instead of machined, but can’t speak to that with any confidence. What I can say is that it was somewhere in the neighborhood of 89° when I measured, causing the router to nod further forward than I could adjust anywhere else to compensate. I ended up taking apart my Z axis and shimming the mount using shim stock.
You can save time by getting this close to square to begin with before fully assembling the Z axis, and/or reshaping the mount where it bolts to the Z plate if you have the tooling.
The bulk of the assembly was pretty easy, just going slowly and carefully.