Assembling Aluminium CNC Machine / MGN12 Linear Rails / XYZ Axis / Spindle Mount

In the last video I prepared some water jet cut aluminium plates by drilling, countersinking, reaming and tapping out their holes to the correct sizes. In this video I receive long MGN12 linear rails and continue with the assembly of my CNC machine.

So my longer rails have arrived today as well. I’ve got the 100, 65. This rail is the only unusual ones. Everything else is pretty consistent. What I’m going to do now is leave the blocks in some alcohol. All of these have the letter B engraved on them and this one has a W. Don’t know what that means.

This is what the methylated spirit looks like after the rail blocks have been in the solution for about three days.

I painstakingly check each rail block, replace any lost bearings and placing the block on a rail. If I don’t think it sounds or feels smooth I carefully move the block to another rail – picking the best combination.

Eventually I got everything working smoothly bar one length. To just illustrate the problem I’m having. These are really good. Good. Good. Good. Good. This one, everything is difficult.

I bought these from a seller on AliExpress and asked them what they suggested I do and they suggested trying to seat the blocks by hammering them.

So this rail is not ground properly. It’s a bit too thick so I need a replacement. Instead, I actually used a thin piece of diamond abrasive which I drew along the length of the channels the bearings would ran along, and this actually improved things.

This might be bowed as well. Once it’s pressed flat against the aluminium extrusion it should be a lot better. It’s not as good as the other rail.

I did try hammering some as well, but that didn’t seem to make much of a difference. It is clear some run better than other but they range seem passable. I’m going to put the best moving rail on the fixed section of each plate and the slightly coarser sounding rail on the section with the slotted holes.

I’m now making the spindle mounts. To save time and material I’m going to do this by cutting the outer shape first – this is a 117mm square on the table saw.

I then making a cradle to hold them on the CNC machine while I mill the clamping hole. I made the cradle slightly undersized and re-sized the hole till I got a tight fit.

I’m using 16mm acetal which is an engineering plastic, and a really nice material to work with on the CNC machine. I then cut the hole out using tabs or bridges to hold the off-cut in place. After which I take that way and make a second pass while the piece is still in the cradle to mill the bridge away.

I check the first piece on the 3d printed mock up, and happy with it I cut the second one out as well.

I then mark, drill and tap all the holes on the spindle plates. The ones at the rear will be for fixing the plates to the z axis pate, and there is also a clamping mechanism at the front. I use a bed vice to hold the material  in the right position and also use a bit of 6mm ply to stop the material squashing and changing shape while drilling.

I drill the holes out with a 4.2mm stub drill bit and thread the opening with an M5 tap. I assemble and check things as I go along.

For the clamping thread I first had to drill an openings a little wider than the shaft of the tapping bit through the clamping section. I spot marked the material with the larger drill bit. You can see I used a bit of 6mm sheet material to top the material flexing while drilling, and then swapped over to a 4.2mm bit to make the hole before tapping.

What I’m going to do now is just cut the ends of the rails. I’ve got a piece of C-beam that’s just shy of 30cm.

I don’t want to have to buy another rail or cut a larger spare piece down so this’ll have to do. If I’m careful I shouldn’t weekend the rail too much.

I then began setting the first rail the correct distance from the edge of the c-beam – and used my dial gauge jig to line up the second rail. I explain how this works in a previous video so I’ll brush over this. So it’s about 0.03 – 0.02 between the rails.

I’m using stainless steel machine screws here but later I swap over to high tensile ones. I then fixed one side of the rail blocks to the plate using 8mm machine screws, followed by 10mm machine screws with a washer on the opposite side. One side fixed securely while the other, the longer machine screws, bottoms out in the threaded holes and allows for a little flexibility in the movement.

I’m now cutting the standoffs for the nut-blocks. I had to surface the acetal down to 10mm and then cut the shapes out. I’m cutting a few more than I would need in case I build other machines in the future. I measure from the top of the material each time but I could in theory measure from the waste board upwards. But what I’ve noticed with the bCNC – just went to make a cup of tea and the thing just kicked itself off. I should have really put more tabs there. I got another two and then I had to just stop. It started to sound a bit funny – I could hear a crunching sound as it was going down and I can’t tell if that something else breaking.

That was the last job I was planning to do with this CNC machine as it is. The clamping plates are starting to crack, and I’ve got all the parts ready for the new version of the machine and these pieces here that I just cut are the stand offs for the nut blocks.

Time to die.

The last thing I need to do, is drill a hole through this nut block. I’ve actually fitted some on already. I just simply held it in place and put the machine screws through and all I need to do now is tap the nut block so I make the centre of the hole and where that lines up with the plate.

I use a parallel punch held in the bearings at 90 degrees to the plate, knocking it with a hammer to make a mark on the blank nut block. I then drill a 7mm opening and use a TR8 tap that I made some time ago to make the thread.

I decided to make my own tap because I’ve noticed there are slight variations depending where you buy the leadscrew and blocks from. Some just don’t fit at all despite being listed as compatible. This is because some threads are square and rounded. Even some sets even from more reputable places can be quite difficult to thread – so for best results I found making my own was the solution.

I also drill a smaller opening for an M5 threaded which accepts a grub screw to tension the nutblock. When it is in position, I will also add a nut to lock it off once in the correct position.

So if you going in from the bottom, make sure you put the washer then the locking collar. I’m just loosening the tension-ing grub screw. It’s a bit hard to see but make sure your tensioning grub screw is extended out a fair bit because you may need to push the tab in while you’re putting the leadscrew through the threaded hole. If it’s not in the right position it can actually pull the smaller section that has a thread on it away from the main body and tension it before you have a change to actually use the grub-screw and locking nut. So you should be able to twist it in by hand. Locking collar, washer.

Ok that feel really solid.

So I’ve just placed the hard stops in position, somehow I’ve over looked it but this is really close to that. The distance between the back of the plate and the face of the aluminium extrusion that faces the plate is 30mm and that’s correct so the error is in the acetal plates I cut… Nooo, the error is in me having reduced the thickness of the plates but not doing so in the model 3D model, and compensating with the positioning of the spindle opening. As we get closer to the end I’ll share more of my cock-ups.

So that’s even more reason to be pleased I didn’t cut these in aluminium. In future what I could do is just add a curvature along face of this curve here and at the rear, because they’re both identical and that would create enough room for the spindle to pass.

So the rails are a couple mill longer and I’m not going to by another length of aluminium extrusion, so I’m just going to cut these down with an angle grinder.

I’ve just put all the t-nuts on.

So I’ve just put the t-nuts and the M3 machine screws in, so this is in position. What I do now use the same jig on the opposite side to set the first three machine screws at the correct distance…

I’s almost like a hole in the centre would be really useful for setting these.

So one problem I think I can foresee, is how to attach these limit switches underneath the plate. I don’t particularly want to do them now because I haven’t really through about that stage. I then tried to rehearse the assembly imagining this had been screwed down. I think it’s going to be really tricky. I think what I’m going to do need is some kind of 1/4 inch ratchet that can get in there and tighten these up when I need to. I don’t think it’s worth putting these on and wiring them now, because I just don’t know how I’m going to do these. Whether they’re going to be normally open or closed. The other thing I’ve just realised as well, if I grab a straight edge from one of the squares. If I push that down on one side, you can see that these are not lining up properly.

Carpenters use winding sticks to do a similar thing on boards, and I should have used this method earlier to identify any major twists before assembling a good deal of the machine. What occurred to me, unfortunately after fixing at least three of the plate using the slotted holes, was that I could have used a different method – shims.

The plates can pinch and restrict movement when tightened to twisted rail blocks. While I can still move my plates when everything is fully tightened, it’s preferable to get things running as optimally as possible so there’s less strain on the motor and parts. What I should have done was to check the distance at each threaded hole on the rail blocks, with a feeler gauge and a straight edge – pushing the edge down against the un-slotted rail bocks as my reference – and after identifying the gap with the feeler gauge used some thin 0.2mm shim washers to pack out the gap between the plate and the block. Now the twists could be the rails or rail blocks themselves, but I also have had the experience of receiving aluminium profile that’s not been square. There’s not much you can do with the latter but the shims would help with the former.

So these sliding nuts will be for the angle bracket that adds a little more support to the y plates when they’re screwed down.

I’m going to jump ahead a little – as I’ve essentially installed all the rails using the same method as I showed on the z axis. In addition to this, I changed the 3d printed 90 degree brackets under the waste board with aluminium ones.

Ok I’ve just noticed something that I need to change in the design and it’s these plates here. The bottom section here which screws into this 20x40mm piece of ali-extrusion is a little bit shifted to one side and if this end is in line with that face of the plate the rail block actually hits into that and can’t move the full extent. So I’mg going to have to change the design for the plate but to get around it, I can just use it lined up with this side of the c-beam.

I’ve just fixed everything together, apart from this section here. The two support pieces at the front and back are a bit too wide, so I’m going to measure  it, dismantle it and cut it down.

I also had to make a simple tap to further adapt the y plates nut blocks to run more smoothly along their threaded rods, but in the end I had this for my efforts.

I’ve plugged it in, the limit switches are not connected so i can’t home the machine but I can move it around, and it does seem to be working.

I can’t home the machine yet as I still haven’t installed the limit switches but it’s moving, and it’s a little quieter than the previous version. Again I overlooked some additional things including not lining up the motor plates properly to the aluminium profile, and there not being enough space for the drag chain to fit along one of the y axis c-beams where I intended it to go, and that the x axis leadscrew is so long if I move to one end, it flexes on the other side. I didn’t notice this before in the previous machine because the leadscrew was facing away from me and this will be a problem if I hit the hard stops, or work on harder material or cut too quickly. One way I get around this is to put a locking collar on the outside of the plate – fixing the leadscrew on both ends. I could still install a drag chain where I had intended, but that would reduce the width of the cutting capacity.

So every time I get a step closer to finishing, I realise there’s a ton more things I overlooked or new ideas I’d rather follow despite being half way through realising the previous ones. But hopefully in the next video, which should be the final one, I’ll just need to install the drag chains, fit the new limit switches, and remove the old wiring for the proximity sensors – and then I can actually make something. Fancy that.

Leave a Reply