thibaultron

Part 015 Adjusting My Router – Making Mistakes, So You Don’t Have To! Part 1 After trying to surface a test board, I found that the 1” bit was leaving shallow concave groves on the surface. This indicates that the spindle is not perpendicular to the surface, or as I found out later, the X-Axis and Y-Axis travel. I assumed that the table surface was both flat and aligned to the axis travel. It was neither, but I’ll get to that later. I want to be able to do some moderately accurate machining on this device as well as routing. The shallow groves that were left on the test board, might also affect the hold of the tape and superglue method of attaching work pieces, though they were shallow enough to not be a big factor. So I decided to go “Full Metal Machinist” on it! The “Full Metal Machinist” is a play on the title of an Anime series “Full Metal Alchemist”. While I sold most of my large metal working machines, I still have a small metal lathe, and a good selection of measuring tools, so I decided to use those tools to improve the “Squareness” of the frame of my machine. I have a 12” X 18” granite surface plate, that I used, but a flat piece of thick glass (well supported and used with care), or board could also be used. I bought the surface plate almost 25 years ago, and never used it, so as long as I had it, I might as well use it now. Yes, in the pictures a couple of the corners are chipped, but I can’t remember if they were so when I bought it, I do remember it was on sale. First I removed the rubber feet and set the assembly on the plate, it did rock a bit, indicating that something was wrong. I loosened the screws attaching the uprights and pressed down on the base, a slight ping was heard on one side and the base then set squarely down on the plate. The upright on that side was not properly aligned with the base, now it was. Remembering that I had had to move one of the X-Axis travel supports when I first setup the machine, I removed the whole upper assembly and placed it on the plate, it also rocked, probably do to my change torquing it. To fix this I ran the spindle assembly to one side to keep the supports properly spaced, then loosened one support at a time, while pressing down on the upright and support to insure they were flat to the plate, then I tightened the screws. Moving the assembly to the other side, I repeated the process. Now it too laid down flat, insuring it was in line also. I used my machinist square to check the squareness of the uprights to the traveler supports, and they were good. Ignore the small wedge in the photo below on the X-Axis knob assembly, that will be explained later. I forgot to take this picture the first time, so this photo is from further along in the process. I reassembled the router while it was again on the plate to insure the uprights and the base were aligned. I mentioned at the beginning that the “T” Nuts supplied with the machine had a spring loaded ball that was supposed to hold the nut in place while you attached the upper assembly, but that many of them still moved or fell out while trying to assemble the parts, the first time. This time around, I cut narrow strips of cardstock and found two layers inserted in the groves held most of them firmly in place. Two still gave me trouble, but that was better than several, and three strips was too thick for the rest of them. Using a 123 block and my square the alignment looked good. The block is sitting only on the frame side rail, there is a slight lip where the back frame piece and it meet. Before I took the machine apart I was having problems with the X-Axis binding at the end with the knob, another reason I went through all this. After I had it reassembled, I ran the spindle assembly back and forth with the knob, and it was still definitely binding at that end! I removed the mounting screws for the knob bearing and the binding disappeared, so there was something wrong with the mount. I ran the spindle all the way up against the upright and looked at the position of the mount, in relation to the screw holes. If the lower screw was installed finger tight, everything was OK, but the upper mount screw hole did not aligned with its hole on the upright. I drilled that hole out a little and installed the screw, but binding still occurred. I looked at it more closely and with the upper screw loosened, I saw a slight gap between the mount and the upright. A cardstock shim took care of the problem (see the photo two pictures above this one). I moved the spindle assembly back to the servo end, and loosened the X-Axis servo mounting screws and shifted it a little to get that end of the feed screw to slide freely into the coupler. This eliminated the binding. With a milling machine, one of the first things you do is level the head of the machine. For most of the milling machines we would use, the head/spindle is fixed in relation to the Y-Axis, but can be tilted left or right in relation to the X-Axis. This allows you the drill holes or mill a side at an angle. For normal milling you need to insure that the head is perpendicular to the table surface. You do this with a Z shaped holder and a Dial Gauge. So I made a setup for my spindle to do this. I took a ¼” Allen Wrench and a 7mm Allen wrench and super glued them as shown in the picture below. I bought Sainsmart’s 15 piece collets set when I bought the router, as I knew I would at least need the ¼” collet, and I got a discount on it when I bought it with a cutter set. I’m glad I did as the largest collet was a 7mm, which fit the ¼” Allen wrench, (remember the flats on the wrench are ¼” apart the tips fit the 7mm collet). The tips of the 7mm wrench fit the 3/8” clamp on my dial gauge. I positioned the spindle at the center of the table, X and Y, and set the position of the dial gauge midway between the Y-Axis table edge and the first etched guide line closest to it (95mm from Center). I used cheap Allen Wrenches, and will keep this setup for future use. Using this setup the Y-Axis measured 0.050” higher at the front than the matching position at the back, and the X-Axis measured 0.057” higher at the left side of the table! As this did not make sense if the table is both flat and aligned with the Y-Axis slide rods, I measured the table in relation to the front and back base frame pieces, as well as the side ones. Here is a drawing of the results. As you can see the table casting is warped along a line from the upper right corner to the lower left corner, sort of like a slice of bread held at opposing corners when held in the air. The warp is about 0.010”. The Y-Axis table heights are much better, only 0.002” difference. The X-Axis is not as good. The left side is 0.014” lower than the right. Because the table is level in the Y-Axis direction, the Y-Axis error can be corrected by tilting the whole upper assembly (forward in this case), I did not worry about it now. The X-Axis is a whole other problem! There are four ways to fix the X-Axis: 1. Ignore it. If, you are doing small pieces it may not make enough difference to bother making mods to your machine, but if you are doing larger ones this will cause a lot of carving to get even depth cuts. 2. Use a thick cutting board, and mill the top surface down a little more than 0.050”(~ 1 mm). This still leaves the problem that started this, the shallow groves. 3. Raise or lower one of the uprights to make the X-Axis travelers match the angle of the table. 4. Shim the table. This means you have to make custom shims for each linear bearing holding the table, a tedious endeavor, and means you need shim stock to start with. I chose to go with the third option. Here is where I made my first mistake. I was going with getting the spindle perpendicular to the table, which I was able to do with much work. Then I measured the height of the spindle to the table, and discovered that the X-Axis travel was over 1 mm higher on one side of the table than the other! Not good for any type of work, the X-Axis travel should always be parallel to the table! I will skip all that I went through doing it the wrong way, and go to how I finally did the adjustments. Initially the left-hand measurement was 0.057”, but I was able to get it down to the 0.027” by loosening the mounting screws between the spindle travel assembly and the plate that mounts it to the X-Axis traveler linear bearings. This was all the adjustment the screw holes allowed. I did totally remove the spindle traveler, to see if I could enlarge the holes to allow more adjustment. Don’t do this!! I discovered that the assembly is held in place with four “T” nuts, all of which promptly slid all the way to the bottom of the channels in the extrusion which makes up the base. In the end there was not enough material to allow me to enlarge the holes, anyway. I ended up having to removed the spindle traveler bottom plate so that I could slide the “T” nuts (loosely screwed into the bolts) into the channels as I slid the rest of the assembly down over them. So 0.027” is the best I could do to correct this, and as it turned out, that was enough, in the end. The spindle traveler removed from the machine, showing the “T” nuts. This picture shows the plate that attaches to the linear bearings. The plate is riveted to the bearings, not screwed, so no adjustment possible here! While from this side it looks like I could slot the screw holes horizontally, there is no meat to enlarge the holes vertically, and the other side has clearance holes machined for the screw heads with quite close tolerances, So way too much would have to be removed to make any further travel possible. Also there is too much risk of cutting up or down through the thin amount of material along the top and bottom edges, if I tried. So the only adjustments left are making the table parallel to the X-Axis traveler and rotating the whole assembly forward or backward the align the spindle traveler/ spindle shaft to be perpendicular to the table surface. The X-Axis error I got out by lowering one upright and raising the other. Unfortunately, the mounting holes in the uprights did not have enough clearance to allow me to do this, fully. I chucked one of the 1/8” bits I bought for the router in my Dremel, and slotted the hole vertically to give me the clearance I needed. I used slight pressure so as not to break the bit. I got it done this way, but the bit was chewed up enough I will not use it for cutting on the machine. To help with adjusting the height of the uprights I used some various thickness brass strips I got from K&S. I put one under an upright to raise it, or under the base extrusion to lower it. I loosened all the screws and only tightened the bottom back screw to hold the upright, in place, measured the result, I then went to the other side and adjusted that. After a few tries the situation vastly improved. I was able to get the X-Axis to within 0.12 mm (0.004”) of parallel side to side. Leaving the strips in place I then slightly loosed the two back screws and rotated the whole assembly (forward in this case), to get Z-axis perpendicular front to back. I also got this within 0.004”. Not perfect, but I think within very reasonable alignment! To measure the side to side alignment I used the machine itself. I placed that much abused 1/8” bit in the spindle and used the Z-Probe that comes with the tool and the Z-Probe function in the Grblcontrol program to measure the relative height at each side of the table. I moved the table to 0,100 mm (middle of the right side). Zeroed the Z-Axis height, then moved the spindle to the left 290 mm (291 mm is the maximum X-Axis travel), and used the probe function to find the height at that end. As I was looking to get both sides the same, I didn’t need absolute height, just the difference between the two sides. The picture below shows the Z-Probe in use, at a later time. To get the spindle perpendicular to the table, I used the dial indicator. You can’t use the machine, as the table travels under it, the spindle doe not travel across it. From my previous measurements, I knew that the table was already level with the Y-Axis slides. (in the middle, at least). After that I tightened all the mounting screws, and rechecked that setup The next Part will show the results of all this work.

Welcome aboard, and thank you for your service!

Part 14 I redesigned the cabinet for the router. The first design had a notch cut out of the bottom front to allow a little more head room, for my head. The machine will be on a shelf over my work bench, and I was allowing a little more room to prevent banging my head getting up and down. The “temporary” shelf I put up is 24” wide though, and I have not been having any trouble with hitting it so I changed it so the bottom corner is square, greatly simplifying the cutting and construction. This , of course means that once built I’ll be banging into it all the time! LOL! Here is a graphic of the cabinet design. I designed it so that the sides overlap the top, back and bottom, hiding those seams, and the door and front overlap the sides. I’ve left the top, bottom, and front overly long. I’m going to assemble them with the excess hanging past the final edge, and then trim them flush with my hand router, once the cabinet is assembled. This gives me “wiggle room” for cutting mistakes. The door will also be over long, and I will mark it once installed and cut the overhang off with a saw. I also plan to cut out the two sides slightly over size, then screw them together and finish cut them as a unit, to insure they match. The graphic below shows the parts assembled, before trimming (The overhangs are exaggerated).

Welcome aboard!

Well, I finally got the routine for skimming the top surface of the spoil board to work. Unfortunately, the bit left slight groves on the surface, which means my machine is slightly out of true! I don't know if I assembled it wrong, or when I took the spindle assembly apart I got it back together out of square. I'm going to have to get my surface plate out, and go "Full Metal Machinist" on it.

Welcome!

I used to have subscriptions to Popular Science and Popular Mechanics when I was young, took about a months allowance for each one. Stopped then in my 20s when they switched from "Take your hammer, and build this house, car etc." to "This is a hammer, and this is what you do with it" type articles.

In the Winter months, all skipjacks had this protection. I'll look for my skipjack books today, and post a picture, if I can find them (my shop is a terrible mess right now). The protection was basically sheet metal rectangles, nailed along the waterline with equal height above and below the water. If I remember correctly the sheets were 2 foot X 4 foot, with the nails in a pattern like two dice placed next to each other with the "5" sides facing you. I believe the sheets were overlapped at each end. I'll try to find the info for you today.

Just as a note to us USA members: At Home Depot, at least, the 1/2" nominal MDF is 13mm thick, not 12.7mm, and the 1/4" nominal MDF is 6mm, not 6.35mm.

I've turned all that off, already, the laptop still wants attention, regularly. Its old, and I guess lonely. Seriously, yes, I've corrected all those settings, but my laptop still needs me to move the mouse now and then. Thanks though!

Part 13 Finished the first spoil board for the 3020. It was a long road with a lot of learning curve (software design and cutting, as well as hardware). I used the 30 day Easel CNC Pro trial version to design the spoil board, which I’ll detail later. The only Pro feature I used was the Ramped Cutting one. This starts the cut at the level of the finish of the last step and ramps the cut down at either 5 or 20 degrees, to get to the next cut depth. Most of the cutting bits available for these machines cannot do plunge cuts (drilling type). The ramp allows the bit to cut down at an angle, reducing the stresses over just cutting directly down to the next depth. I’ll have to redo the design in CarveCo, once I learn how to use it. I’ve found the best way for me to learn such software, is to just go ahead and design something, so that is not a problem. At least I know that all the holes are in the right spot. One thing I will have to change is the depth of the cuts for the through holes. I’m using ½” MDF, so I set the depth at 12.7mm. Unlike modern plywood in the US, which runs 1/32” thinner than the standard old dimensions, my ½” MDF seems to be slightly thicker than ½”, and the holes did not penetrate all the way through. The standard design for spoil boards for this size machine is one ¼” thick, held in place with the Blue Tape and Superglue method. That method has you put a complete layer of Blue Painter’s Tape on the bed and bottom of the work piece. The tape strips are placed as close to each other as possible, without the edges overlapping. You then run a bead of superglue down the center of each strip, on the bed, and spray glue accelerator on the tape on the work piece. Then you carefully align the work piece and press it down. The tape holds well enough to both parts to provide a good attachment for most cuts. After you are done, you pry up one corner of the workpiece with a chisel or screwdriver. This breaks the tape’s hold, and you can pull off the workpiece. Then you peal the two superglued layers of tape off of the bed. The picture below shows a quick example of a layer of tape applied to the finished spoil board. Normally I would not have that exposed section on the right, and I would have trimmed the loose ends. Next time I will buy thick superglue, not the Gel type shown. The gel type is very hard to get to the tip of the bottle. I have to bang the tip (with the cap on, of course) on the table several times to get the glue to fall into the tip area. I like this method for work pieces, but would prefer to bolt the spoil board to the bed. I’m not going to be using the machine as someone would for a business, so the spoil board will not be changed as often as if I was running it 8 or so hours each day. Thus if I used the tape method it might have to survive months between board changes, not weeks. I don’t trust the tape bond over a long time span. Thus to allow for the bolt heads, I went with the ½” MDF. Anyway, the design is for a ½” thick MDF spoil board held down with 6 recessed screws, one in each corner and two in the center, with 30 holes drilled to match those in the bed. The corner holes and the two center holes are not that useful as hold downs for the type of clamps supplied with the router, so the bolts occupying them should not greatly affect the clamping ability of the spoil board. Below is a picture of the finished spoil board to this point (milling it flat, later, will be the final step). I found that the software limit settings did play an important role in cutting this board. The Easel CNC software defaults to the origin being the lower left corner, rather than the machine’s upper right, so I ran the tool to the bottom left corner, using the limits to set that location. There may be a way to change this in Easel, but I decided to just “Go with the flow.”, for this first design. I’ll detail the design process in a later post, as well as when I redo it in CarveCo. I designed it using the 30 day trial version of Easel Pro. Unfortunately due to preparing for, going on a weeklong Thanksgiving vacation, and doing all the normal activities skipped at home during same, I was delayed in starting the board for 3 weeks. I finished cutting the boards below and making the last correction 6 hours after the trial expired! This means that the ramp function was removed from my final version of the files, thus the need to redo it in CarveCo. I started with the design cut into a piece of ¾” plywood. This gave me some added margin of error, if I accidentally set some cuts too deep, or some other problem cropped up. I ran the spindle to the lower left with a “V” bit installed to show me were that “0” location was, and placed tape along the edge of the bed to help me align the work piece. I clamped the plywood using the blue tape method (above), and started the machining. There were a number of problems I ran into: 1. My first laptop developed a severe heat problem, and eventually died. Luckily I had an older laptop, and was able to switch my hard drive over to it, and continue. 2. I have not built a case for the machine, and the laptop is sitting next to the router. This was depositing a fair amount of sawdust onto the computer! 3. Vacuuming the sawdust off the computer while you are cutting, is a pain, especially if you accidentally hit the power button, while doing so, shutting down the computer in the middle of the job! 4. Pausing the job in the middle, due to the Admiral calling you off to do something else, doesn’t work, at least with my laptop! The job was locked up both times I tried it. 5. You have to babysit the router the whole time it is cutting, as Murphy dictates that bad things will happen as soon as you step away. 6. You have to periodically vacuum up the sawdust/shavings, anyway. 7. Don’t accidentally hit the power switch on your power strip with the vacuum cleaner hose while the job is running, either! 8. Having to restart an hour or two long job, halfway through due to any of the above problems, is really aggravating!!!! 9. Enclosure needed! Cutting MDF left me covered in a layer of MDF dust, even with regular vacuuming! 10. It is safer to move the mouse every few minutes to make sure that the laptop is not turning anything off. 11. Don’t try doing anything else on the computer (at least an older one like this), as the G-Code program may crash! It apparently uses a lot of computer resources! I did find that returning the machine to the starting point (sometimes several times) was reliable, at least. I set the cutting speeds very conservatively, so the jobs probably could have been run in shorter times. I read that the router should be creating chips not sawdust, otherwise you can overheat the wood or the cutter. My settings were mostly making sawdust. I checked after the finish of each job, and the bit was not heating up, though. I cut the clearance holes first. The first solution I tried to keep the sawdust off the computer, was to place a storage tub top between the machine and the laptop. This worked, but blocked my view of the computer screen, and access to the mouse. I found an old box and cut it to use as a block. This worked well as a temporary measure. After the clearance holes were finished, I cut the through holes. I had cut the through holes in the plywood 12mm deep, planning to drill them the rest of the way through, using the existing holes, as a guide. This would have worked, if my drill press was not presently buried, trying to do this by hand allowed the holes to wonder, and many of them did not clear the bed holes. This did show some errors in the design, that I was able to correct for the final design, though. Another problem I found was that the screw heads were thicker than I had thought, and I deepened the clearance holes by another 4mm, to give more clearance for miscuts, during regular use. In the future I will grind the tops of the bolts down some, for added clearance. Presently I have about 5/32nds” (4mm) of clearance between the top of the screw and the surface of the spoil board. I’d like to get that up to ¼” (6mm). This is a shot of the first version of the finished plywood board, I later cut the clearance holes deeper, for the added clearance. The black marks indicate where the drill holes wandered off center, blocking the bed holes. This is one of the screws installed. I used a “V” bit and ran the machine by “hand” from the computer, to mark the limits of where the tool bits can cut. See the lines by the screw head above. Using the lines on the ¾” plywood, for alignment of both the “0” position, and getting the board square, I mounted a piece of ½” MDF and cut my first spoil board. As I said above, the through holes did not go all the way through, but I only had to drill out a thin skin to complete them. After sanding the top and using a large flat file to get rid of some slight existing ridges around the holes (top and bottom), I mounted it, with 12mm long 6mm screws. The final step will be to use a 24mm flat bit to mill the spoil board surface flat and parallel to the bed and spindle. I found a canned program to do this, but it runs into problems with the software limits. I’ll have to investigate this further, or design my own program.

I'll post more later, but I just finished cutting my first Spoil Board today! Has all 30 holes for the clamps with 6 of them recessed for machine screws to hold it down. I still have to create a program to mill the top flat, but have been struggling with Software Limit issues. Even a canned routine will not run. More to figure out. The board is 1/2" MDF, rather than the standard 1/4", to allow for the screws. I don't like the idea of holding the board down with blue tape for long periods, as even Blue Tape will leave deposits after an extended period.

Easel CNC assumes the 0,0 point is at the bottom left corner (like regular Cartian Coordinates). So I always have to move the 0 when using Easel. I'm just starting with CarveCo, and haven't drawn anything in it yet.

Thistle, your limit switches should have stopped the travel, before the mechanical limits were reached. My problem was only with the software limits. Any G-code software problems should not effect the hardware limit switches! Run the table by hand and test that the switches, actually activate, before the mechanical limits are reached. Also make sure that you have at least a few mm of clearance after the switch activates, to allow the whole system to physically stop, when running full out, and for the fact that switches have a tolerance for activation. The limit switches that are supplied with the 3020, have about a + 0.008" tolerance.

Part 12 OK, the single X and Y-Axis limit switches are not enough! I have to contact SainSmart again about getting a set of matching limit switches and mounts. It turns out that the Software Limits are only effective, if you never change the “0” positions of the axis. You have to do this for most jobs, as you have to set the axis “0” to some position over whatever chunk of workpiece you want to carve. Once the “0” position is changed the soft limits go out the window, and the machine will happily go past the mechanical limits, as the total available travel is no longer “X”, it is now “X- however much you moved the origin.”

Years ago I worked in a computer lab, with a raised floor with removable 2X2 foot "Tiles". You lifted the tiles to run cabling underneath the floor. The tiles had blue carpet squares glued on top. Over the years many carpet squares came loose, making the floor look bad. Our boss convinced the company to replaced them. They choose a multi colored earth tone "Where the h..l did did that part go!" pattern for the new carpet squares! After they finished we and the boss wished we had just kept quite about the floor!

Yes, I've used mine for that more times than I care to admit to!

Be careful of that heater plugged into the power strip! My friends had the cord to their's catch fire 2 weeks ago! Thankfully they caught it in time.

Part 11 Ran into a problem yesterday. The Z-Axis mechanism was locking up during the "Homing" operation. It was mechanically locking before it contacted the upper limit switch. The stepper motor was buzzing loudly as it was still trying to turn the feedscrew. I couldn’t see any debris in the slot at the top back of the feedscrew nut. Then I noticed that the coupling nut looked like it was lower than it had been. The other coupling nuts sit in a depression in the frame that puts the set screw for tightening the servo end of the coupler partially below the frame surface. Thinking that the nut for the Z-Axis was the same, I took the assembly apart. Nope! It turns out that the coupler hits the top of the seated feeds crew, before the nut gets that deep into the upper frame/motor mount! ###....####. While putting everything back together was not difficult, and the parts seem to slot nicely together, I hope that nothing is just slightly out of factory alignment. After running the feedscrew, by hand, until the slide was mostly at the bottom (being careful to hold it off the limit switch), I Then slid the coupler back onto the servo shaft as far as it would go, and tightened the clamp setscrew. The slide now cycles correctly. I also had removed the spindle motor during the disassembly, and I had noticed that the wires had sagged and were touching the frame, during cutting. So I turned the motor to put the wires from the motor to exiting from the right, as shown in the instructions. This seems to help. I have been using an old piece of Particle Board (as opposed to MDF) for my trial cuts, and I have broken 2 20 degree “V” bits in short order. This is a very old chunk of material (at least 25 years old) and may be of different composition to modern stock. In any case I would not recommend using Particle Board on this machine with “V” bits. I’ve purchased both ¼” and ½” MDF boards, and will use these from now on. I’ve also found that the “V” bits stick into the aluminum top surface of the Z-Probe when using it (deep enough that the probe will raise off the workpiece when the slide is lifted back up). This gives a false height, naturally, and mars the probe surface. It use the slip of paper method for setting the “V” bit height now. I have not used any of the other types of bits, and will report on them, when I do.

Shipping Pallet patent November 1939, so it would be correct for 1940.

Metal Lathe - 1. A device for throwing chuck keys at your face. 2. A device for throwing large semi round workpieces at your body, because you forgot to completely tighten the chuck (reason chuck key was still in the chuck). 3. A device with a lot of heavy chucks that fall on your feet, during changing.

Thanks!

As you can see in the last picture of Part 11, I was incorrect about UGS having Jog buttons! The "Jog Controller" button opens the Jog button window, unless of course, you've accidentally clicked the "X" in the Jog Controller window! In that case it disappears, and I haven't been able to figure out how to get it back! ***,,,****! jog

Clare, what is your stain formula?