_Michael's Dovetail Guide

[_Michael]

@Tpoole1988

Thank you very much.

You’re welcome.

I was able to make the slot after I mounted the LongMill by using the lower set of mount holes and a 3.5" bit. I installed the stationary 2x4 after the slot was cut so I could get it to be flush. So with the router low and a longish bit I was able to get through the 3/4 MDF and the 3/4 plywood. Just have to plan the spacing for the spoil board strips before making the slot so you can cover it.

So get your mill setup nice and square and then everything can be made to be square to the machine without too much fuss.

Not sure if you are getting the limit switches for your mill but I have found them to work well with the spoil board grid starting from the center. When I found the middle of the spoil board for making the grid I wrote down the coordinates and saved them to the last workspace which I don’t use for actual work. That way if I am working from the center I can home the machine, switch to that workspace and hit the GoTo XY button and my mill is back in the center. Then switch to the workspace I’m going to work in and set the XY to zero. I’ve been able, with the help of some dowels and holes, to take a project off the mill for sanding etc and put it back on the next day and continue, no problem.

[Tpoole1988]

Okay Michael, the day of greatness is upon us. I finally have my table built and my CNC unboxed and starting to piece it together. I have a couple of questions about the spoil board.

1. You’re clearly a pro at this. It looks like you carved the top of your spoil board first and then cut out the pieces where your t tracks lay between them. How did you calculate and perfectly cut out the distance between each t track? I zoomed in on the images you posted and it’s super precise, I assume you didn’t cut those strips out by hand?

2. Being new to this, I still understand that lines and symmetry play a large part in getting pieces squared up and cut evenly, but what is the purpose of having them etched into the spoil board? I see many set-ups with the lines carved in, I just haven’t reached the point yet where I understand their usefulness. I would assume you wouldn’t need them if you’re using the touch-plate on the corner of your piece as a starting point.

3. I plan on running my t tracks vertically as opposed to your horizontal runs. I think I’ll carve out a slot on the far right/left like you did at the front of your table for inserting pieces to cut vertically. I love the Dovetail work/diagram/explanation you started this thread with and want to take full advantage of that.

[_Michael]

A day of greatness indeed! Now more waiting game for you

@Tpoole1988

You’re clearly a pro at this.

Not really, but thanks. I got my LongMill 2+ years ago, and been reading here ever since is all.

This is how I made my setup. First I made the table, 2x4 construction as straight and square as I could. Which for me meant building it on the concrete floor and hoping that was good enough. My mill is on the left side of my table so when I figured out where the front left foot should be, I used a straight edge and square to mark an L. I attached the left front in the corner of the L and the rest of the left feet touching the line. Then with the X rail forward I mounted the right front touching the front line and as I moved the X rail back I attached each remaining right foot.

If I had to do it over I would most likely try the method where you mount the mill to strips with a slot and bolts for adjustment. Even when drilling the holes for mount screws first it was easy to mess up the placement by over tightening an angled screw. Not sure if the MK2 screws go straight down or not but some of the MK1 ones are angled.

After mounting the mill I jogged to the 4 corners and marked the bit center for each corner. The t tracks came in packages of 4 and I didn’t think 4 was enough so I bought 8 t tracks. With the number of tracks, strips, and the area to cover known it wasn’t too hard to figure out the spacing. The strips ended up being being about 4", they aren’t exact, I ripped them on a table saw. I marked lines for the t tracks and installed them first. With t tracks mounted I was able to clamp the strips so I could mill the counter sunk screw holes for attaching the strips. I made one program and ran it in the same place 7 times switching out the strips each time.

The spoil board I made actually covers more area than the center of the bit can reach. That’s okay as long as you can reach the edge of the spoil board with your surfacing bit. You want to be able to surface the whole top of the spoil board in case a work piece needs to extend off from it, for pass through work etc.

The grid was done last with my laser. Since it was done last with everything mounted it is square to the mill. I needed to make a mount so I could put the laser in the center of the router mount so the laser could reach everywhere. I took this opportunity to cut the cable on my router and put in some connectors, one each way so I can’t plug it in backwards. Then I put connectors on a salvaged cord so I can use the palm router off from the mill if I need to.

The main use for the grid lines is just to place work pieces square to the mill easily. Before I had the lines I would put a board touching the bit at the front and then while holding the front, jog the machine back and move the board to touching the bit, hoping I didn’t move the front etc. Now I just take the straightest edge and place it touching the line and clamp away. The touch plate will find a corner but it won’t make sure you mounted the stock material square if your trying to use all of it. Being mounted super square may not matter if your not using the edge of the material. But wood is expensive now a days and I try my best to use as much of it as I can.

As far as the direction of the t tracks goes, I think it’s just user preference. I knew I was building an enclosure, primarily to keep the laser fumes contained/exhausted, and that I would not have access from the sides so that dictated the t track direction for me.

Long read, I know, hope some of it’s useful!

[Tpoole1988]

Hey Michael, I’m back for more. I was wondering if you could help me with the tool path for a 60 degree 1/4 V-bit to make this spoil board. I knew the lines wouldn’t show up very well on plain MDF so I looked online for inspiration and saw where someone had painted their spoil board black and then used a V-Bit to cut in 0.1 - 0.2 mm and create clearly visible lines. I painted my spoil board black and I’m going to attempt to do the same thing. The VCarve file you sent over is only allowing me to run it in a 2D Profile Toolpath because it’s saying there are “Open Vectors”. If I use the V-Bit in the 2D Profile Toolpath is there anything I should look out for in terms of setting my start depth/cut depth, starting point, (do I use the AutoZero Touch Plate?). After squaring, tramming, painting, etc I would hate to have come this far to drive my bit straight through the bottom of my table and start over.

*Also, I only deleted the text on the bottom and top of your original file because I thought it might be adding to the “Open Vectors” error and I was attempting to clear out anything that might have been leading to the issue!

Open Vectors Error

My 2D Profile Toolpath *File below image

SpoilboardText48x30.crv (2.7 MB)

Envisioned Output

Spoil Board Setup

[_Michael]

You should be fine to use the profile tool path with a v-bit. The v-carve tool path is to clear out an area inside of closed vectors. It’s like a pocket tool path with angled sides.

You should use the profile tool path, on the line, with the z set to top of material in V-Carve. Use small cut depth to start because you can always stop the program and go a little deeper if needed but it’s hard to go the other way if you cut too deep, “Hey Joe!, I cut this board 3 times and it’s still too short”.

Your setup is looking good by the way. I like the black!

[Tpoole1988]

Michael it looks great!.. well, halfway great. Things were going fantastic until about halfway through the job. That’s when I noticed the machine was just slightly grinding on the left Y-axis rail when it was moving X&Y simultaneously, but it was within the bounds of acceptable from what I’ve researched and seen online. Then I really saw it; I noticed that the left Y-axis rod was spinning freely and the anti-backlash nut was completely loose. I paused the job, tightened the nut down by hand to a firm but not “death-grip” tightness. Also checked the other backlash nuts and everything seemed good. I clicked “Resume Job” and the machine ground to a complete halt. Paused the job again, loosened backlash nut a little bit on the left Y-axis, I confirmed it was around the same tension as all the other backlash nuts. When I pressed “Resume Job” this time, $h!t hit the fan… the Y-axis rails quickly became misaligned, there was some serious griding, the left side Y-axis jammed up, bounced, and I heard a small “pop/snap”.

I hand cranked the Y-axis rails back into alignment. They run in alignment under the “Precise” jogging, but they jump out of alignment if I speed it up to “Normal” or “Rapid”. Clearly I need to deconstruct this little fella and see what happened, but I’m guessing I broke my anti-backlash nut?

Before the Implosion

[_Michael]

That’s too bad that it messed up. I have a couple of tips that may help you.

First, if your machine’s Y rails are even you can move the machine all the way to the back with the motors and then either disable soft limits or just move the rest of the way back by hand. Moving both sides back until they hit the stops should put you back in alignment.

Second, I’ll share how I adjust the backlash blocks. I loosen the adjusting bolt until I can feel, and hear if it’s quiet, the play in the block. Do this by grabbing the plate that the backlash block is attached to and try to wiggle it back and forth. Now that you can feel the play tighten the adjusting bolt until that play goes away. You can give it a smidge more, like 1/16 turn, after the play is gone but not much more or it will cause too much resistance and bind.

Good luck and remember that we all had to go through the learning process. It’s all part of the journey and I’m confident you will get there!

EDIT: Forgot to mention in the first tip that you will need to turn soft limits back on and home again. If you moved the last bit by hand you just need to home again. Any time you move something by hand, crash or bind the machine you will need to home again.

[Tpoole1988]

Awesome advice, I’ll get the backlash nuts adjusted following your method.

I saw some advice you gave someone about finding your machines true center and followed it to a T, leading to a little discovery. I run my y-axis all the way to the front until both sides are touching. I jog it all the way to the back and the left side touches about 1-2 mm before the right side connects with the back plate. Does the right side need adjustment for that small of a misalignment? What would be the simplest way to correct that?

[_Michael]

I was thinking that you had the limit sensors but it doesn’t look like it from the photos now that I look closer. If you don’t have them then you can disregard the homing part of my first tip in my last post. Any manual move, crash or bind will still mess up your zeroes for the axis that were moved/crashed so keep that in mind.

Now on to your last question. I would start by adjusting the backlash blocks and then test again. Maybe one side bound a little bit. If after the adjustment the problem is still there I would measure each Y rail to see if they are different lengths from plate to plate. I suggest using either a yard stick or a story stick to make that measurement. You will get a more accurate measurement that way. If you don’t know what a story stick is, I didn’t when I started, it’s just a board that is too long to fit and you can mark the end and compare to the other side or you can keep trimming it on the table saw until it fits on one side and then compare to the other. You can use a tape measure it’s just a little harder to be super accurate IMHO due to reading in an inside corner or the tape bending different from side to side.

If the lengths are different then we would expect one side to hit before the other. In that case you would need to use the calibration tool in gSender which is a good thing to do in any case to check the squareness of your assembly. Start by squaring the machine to the front and run the calibration test. It will have you move the router and make some measurements that form a right triangle. You can use a v-bit and just make a small mark in the spoil board at each point of the triangle, you don’t even need to turn on the router just lower it and make the mark. Using that information it will tell you if one side needs to be moved to get square cuts. If the X axis is not perpendicular to the Y then your squares will actually be parallelograms. If the front is not right try the back. If neither is correct then pick one and either move the whole board that the Y rail is mounted to, if it’s easy, or you can make a shim of the thickness that it tells you the rail needs to be moved by. Then you would take that shim and fix it to the plate at the end of the Y rail so that when you bump it the machine is square, hitting the plate on one side and the shim on the other. One other thing, making a bigger triangle will be more accurate than a small one because any measuring mistake will be a smaller portion of the total measurement.

If both sides are the same length and you still reach one side before the other after adjusting the backlash blocks then I’m not sure how to fix that. Let’s hope that’s not the case for now and cross that bridge if we come to it.

EDIT: With the help of the fabulous @gwilki , I have a couple of more things for you to check! The first are the couplers that connect the lead screws to the motors. If there is slipping there that could lead to one side lagging behind the other. To check this you can change $1 in gSender’s firmware tool to 255. This will keep the motors energized and you can try to turn the lead screw by hand. You shouldn’t be able to turn it without making the motor skip. Change $1 back to the default of 100 when you are done or your motors might run hot.

The second is the tightness of the v-wheels. If they are too tight this could cause an axis to bind. You want all the wheels touching the rails but you should be able to turn them by hand. The wheels have kinda sharp edges so I figure if I can turn them by hand without causing too much pain then they aren’t too tight.

Thanks for having my back Grant!