So I have been using my Shapeoko for a month or two now and I have had fairly good success with milling and engraving. I have definitely learned since I started. I’ve improved my processes and have a better understanding of how CNC milling works.
But! things can still be improved. To that end, now is the time to do some upgrades! Over the weekend I spent some time designing and implementing a few upgrades to my machine. I also have more in the works once I have the time.
Dust Containment Upgrade
One of the very first things I did when I got my shapeoko was add Y-axis rail guards. Inventables sells sheet metal guards that bolt onto the Y-axis, but for $60 I could not really justify them. My plan involved using the very box that the shapeoko came shipped in. I measured up the rails, and decided where I wanted to put in the bends. A couple minutes later, I had some cardboard guards that would help keep most of the chips and debris off of my Y-axis rails. The guards help to contain a majority of the dust within the work area, so that I am not covering my surroundings in plastic chips. Eventually I plan to upgrade the guards, perhaps to acrylic panels or something, but for now, the cardboard gets the job done.
Work holding upgrade
Work holding is a new requirement for me. My 3D printing experience did not help much here. On my 3D printer, occasionally I use hair spray to give the filament better adhesion to the glass. This helps the plastic adhere on the first layer much better. But this has no application with a mill. I don’t have first layer adhesion issues, I have to hold the stock material stationary during the whole milling operation. At first I used wood screws directly into my MDF waste board to hold things down. This worked, but I had to drill into whatever my material was first, and then sand down the hole in my waste board for the next job to ensure that it did not ruin the levelness of my bed. This was a pain. So the next step was wooden hold down tabs that were screwed into the waste board and held the edges of the stock. This worked OK, but I still had to mess around with drilling for different sized pieces of stock.
I again took inspiration from the shapeoko community for the solution. I purchased a set of 5×500 mm pieces of 20×80 mm T-slot aluminum extrusion from misumiUSA and laid them side by side to build up my bed. This allowed me to have several different slots that run the whole length of the machine. I can run hold down clamps in any of the rails to hold down whatever I am working on. It still requires me to attach a waste board first (because through cuts would ruin my bed) but it looks better, and allows for more options.
The problem posed by using five separate pieces was attaching the 5 sections together reliably so that there was no change in level from piece to piece. My dad came up with an idea here, and to that end we purchased a 4 foot length of aluminum angle extrusion from Ace Hardware. All I had to do now was cut out sections with a hacksaw, and drill holes in for a bolt. I used two sections that each spanned 3 of the extrusions underneath to bind them together, and small sections on each corner to bind the whole table to the frame of the shapeoko. These are not very consistent, but they do work for now. I may design a printed part to more consistently attach these pieces. I also found a design on thingiverse that was specifically designed to hold to pieces of extrusion next to each other tightly. My print failed halfway when one of the pieces dislodged and ran into another one, but as it turned out, they still functioned. I placed one over the threshold of two pieces and at first was confused how it would fit. I thought I had scaled it wrong and that I would need to reprint. However, add in a hammer and some force and they went right into the extrusion. The insertion process even trimmed off excess that wouldn’t fit. I used a screwdriver extension to insert some of the pieces even further in too, so as to give some more holding strength. After all the connections were inserted and pounded flush, the whole thing was connected very rigidly. Add in the aluminum angles on the underside, and I had a nice one piece extrusion bed.
It is ironic though, that I do not use hold downs at all right now. Instead, I have just been using double sided tape to hold down my sheets of plastic. This has been very effective up until recently, when all the lines drawn in my waste board from through cuts have reduced the amount of surface area to stick to. Next time I work, I will flip around my waste board and use the less cut up side.
Dust/debris from cutting is also a new thing to me. My printer only puts out what it uses. My mill has to remove the material and consequently throws chips everywhere. I use a shop vac on the side to do some removal while it is running, but I do not like to have to babysit during a 2 hour milling cycle. The chips also get thrown onto my nearby printer, which does not like little tiny particles of stuff getting caught on the linear rods or in the printed part. The solution I decided on, like many other shapeoko users, was to add a dust shoe.
Again taking inspiration from another shapeoko user, I decided to use clear ¼” acrylic for my dust shoe. The idea is to enclose the bit with a brush so that all the chips are caught within and then vacuumed out via attached vacuum. My design incorporates 3 pieces in the dust shoe portion: a rounded plate affixed to my spindle by means of the collar threads on the dremel, a vacuum side section, and a bit side section. The fixed plate has holes for both the vacuum coming in, and the bit coming out of the spindle. The hole for the bit was counter bored to allow the existing collar nut to be used to hold it in place. The counterbore allows for more thread engagement on on the nut, while still allowing thicker acrylic to be used. There is also a through hole for the vacuum attachment on the fixed portion. The vacuum side is smaller that the bit side, as it only holds a vacuum nozzle that is directed at the milling bit. The vacuum side plate is removable and attaches to the fixed plate by means of neodymium magnets. The bit side plate is also removable via magnets and shaped like a horseshoe to allow easy removal and attachment.
Both the vacuum and bit side have a craft foam brush super glued to their perimeters to catch chips as they are cut. This would work in turn with the vacuum to contain most of the debris of cutting right as it is cut. All three pieces had holes milled in to allow pressing in of the neodymium magnets. The magnets fit tightly, but still require super glue in the bottom of each hole to keep them from zooming out to catch any other magnet or magnetic material nearby. These magnets are seriously strong. I had to glue in the magnets in every other hole at first so that they wouldn’t pull each other out while they were drying. I did this in two goes, and clamped a piece of wood over all of them while they dried to ensure they were seated completely.
The magnets worked incredibly well. 5 magnets per side in a pattern meant that the removable sections snapped right onto their home (self aligning too). They could be removed by twisting the sections off, but otherwise they were very strongly connected. Because of the patterns they were in, once you had all five mated correctly, the sections wouldn’t even rotate because of their strength. It was extremely satisfying to have the magnets work like they did. I hope to find more ways to incorporate them into other projects.
Unfortunately, the dust shoe did not work as greatly as I had hoped. In my attempt to keep it on the smaller side, I overlooked the fact the the craft foam dust shoe could be pushed into the bit under the right (or perhaps wrong) conditions. When this happened, the cutting did not work so well. I also ran into the problem of melting chips back into pockets being cut. This was a result of not attaching the vacuum nozzle yet. I figured I could at least contain the the chips (which it did fairly well, probably around 90% containment) without the vacuum, but it contained them too well. Chips fell back into pockets and that got heat fused back into the pockets when the next pass began.
My next move with this is twofold: print a vacuum nozzle and barb to attach a ½” I.D. clear hose as my vacuum tube, and increase the diameter of the bit side. With these two changes, I should have a nice dust collection system fully functioning.
Emergency stop and reset bracket
Right now, the only way for me to really “emergency stop” my machine is to pull the plug on the power supply. This is not at all convenient or easy to reach in a true emergency. So, again turning to Amazon (yay prime shipping!), I ordered an emergency stop button. The one I ordered could not take mains voltage or amperage, but it could handle the voltage and amperage of my power supply. So I planned to attach a switch (because the power supply didn’t have one) and the E-stop in series with the hot line of the power supply. This way, I could turn the power off and on with a conveniently placed switch. I also had the ability to trigger the E-stop easily if I need to. And since the E-stop cuts power from my PSU, it will shut down my stepper motors (and the spindle once it’s controlled via relay). I also opted to add a momentary switch so that I can breakout the function of resetting my controller. I find that I hit the reset switch after each job for smoother operation. So now that the button is up front and easy to hit, its not nearly as much of a hassle. At work, we were throwing out a whole bunch of electronics stuff, so I managed to salvage a bunch of switches before they got thrown out.
So one morning I started designing my bracket/box to hold all this stuff. I wanted to have it mounted on one of the end plates of my machine, but also not interfere with the adjustment slots for the rails. So I traced the outline of the slots that needed clearance and located the holes needed to mount it. I used the same box construction method that I used in my head mounted display, interlocking slots. The E-Stop needed a simple hole for panel mounting, and the switches needed rectangular holes to snap into. I double checked my design (evidently not that well as I later found out) and went downstairs to start cutting it out.
The first problem started when I tried to layout all the pieces into one cut job. I put a part too deep in the Y direction, so that when it tried to reach that position, it skipped steps at it’s limit and ruined the rest of the job because it no longer knew its correct position. Some of the parts had already been cut out though, and were still able to be used. Another problem I had was plastic chips that were being caught by the dust shoe melting back to the stock. The dust shoe had no vacuum relief, so chips were just getting caught back in the cuts. I alleviated this by trimming down the brush ends of the dust shoe, but now it is only about 20% effective in containment of chips. I already have the ideas I need to remedy this for Rev 2 of my dust shoe, but I wanted to get this done first. I also had parts popping up prematurely as they were being cut out. Normally, as long as there is even some of the double sided tape within the cutout section, it will stay down until I pry it up. The problem here was that I have been using the same waste board section for almost two months now and all the parting lines and over cut from all this time built up, resulting in reduced surface area that to hold down stuff. Finally, the worst problem, was that I miscalculated my design. I did not account for enough clearance on the cutouts for the switches, so those would not fit/snap in correctly. Then I also did not account for their depth, which means that their terminals were interfering with the E-stop terminals. I have re-designed the top and front portions to allow for proper spacing. The next step was to re-cut those two pieces and then finish assembly.
UPDATE: Redesigned the top and front sections to that everything could fit. I had designed the front section to be removable so that I could access the bolts to hold it on, but the wires are so stiff and tight inside that it is really more of a permanent connection now. I shouldn’t need to get inside there anytime soon though.
I was able to wire it such that I could have the reset button, power switch and Emergency stop all in one 4 conductor wire. The Emergency stop is in series with the on-off switch and is wired such that if hit, it will disable power into the CNC controller. The reset button is a normally open switch that is soldered directly onto the legs of the reset switch on the controller board. This delicate soldering work was done for me by my dad, as I don’t quite have the soldering skills required for this yet.
The enclosure sits nicely on the front right rail mount plate and all three buttons work very well. The new button placement makes it much easier to use these functions of my machine.
Future Planned Upgrades
I have been having too much fun actually using the machine to do many of the upgrades that I should. But on the horizon, I have a couple more planned that could make life easier.
Limit/Homing switches is a big one. Limit switches would allow me to limit the motion of my machine such that it shouldn’t crash into any of the physical limits of the machine. Right now the machine can guess its own position, but there are no definite position feedback. You can mess up your coordinates if you move it by hand or if it skips steps somehow. Homing switches will allow for repeatability of positioning. Once I have homing switches, I could cut one layer with one tool, switch tools, re-home and then cut with another. This would be a very nice upgrade over the way I am doing it now.
My controller board and wires just hang off at the end of my machine right now. It gets covered with chips/dust everytime something cuts, and the wires can be pulled out by accident very easily. So, eventually I’d like to enclose it all in a box to keep it clean and avoid accidents. I have a layout already done for it, just need to actually get this done.
I would also like to control my spindle via controller, not just manually turning it on. I have purchased a beefcake relay kit from sparkfun, which is all ready to go, just needs to be soldered up and connect it to an outlet. I plan to connect both the spindle and a vacuum on the relay controlled outlet so that when my machine sends the “spindle on” command, the bit starts spinning and dust collections turns on at the same time.
Another possible project is a complete enclosure. I already have all the aluminum T-slot and corner braces to build it, but still need side panels to enclose it. Ideally, it would be enclosed in ¼” acrylic panels. This, however, would be incredibly expensive, as the sizes needed would be odd. This would prevent all dust escaping the machine (both good and bad, this point), but also might restrict access to the machine too greatly. This may also aid in hindering the noise it generates (a dremel screaming for a 2 hour job is not the most enjoyable of noises).