Build Big Cnc at Home 3x2m




Introduction: Build Big Cnc at Home 3x2m

So why this should be of any interest of yours?

This machine was build without any special tools, basicaly drill, clamps, saw and meter would do. Ofcourse there is more, like drill bits or, taps... but i built thisone without having even a proper table.

Machine was designed to use as few part as possible and to be as precise and strong as possible, also speed was desirable as i wanted to use it with plasma, then i also wanted to be able take the machine apart, build it from localy available components so no premade table parts are needed to make the build, and after all this this machine still needed to be nice and square. Weight of some components and force distribution was also thought over.

At the end here are some realy interesting solutions for problems i delt with, worth to check out, at least i believe so.

Step 1: Table Support (aviod This)

As i never tried anything like this and had some vision of proffesionaly looking sturdy table i decided to weld a construction. I had access to TIG so i was confident to make nice welds.

This was a mistake. If you are not proffesional welder forget about this. The metal will be bent and it just results in a need to add spacers and the welding of both suports took me just too long. I am adding here some images from the process but look at the 3D image, the left part is variant what i should have done (or some similar think), only few square tubes screwed together would have been great.

I user 80x80mm tubes for the constructon. This was more than suficient for sturdy table.

One support was welded with help of clams and magnetic squares and then put pieces of the other one on top of first to make them similar as possible.

To connect them i drilled holes in the edge of the square tubes and made a tap there (8th image). Last image shows the conencting squre tube, there are 8 holes, 4 (visible once) so i can use allen wrench to easily attach screws and on the oposite sides are another holes so the screw can go trough.

The two longest tubes on top of this support will hold the table bed. Try make them as parallel as possible but even if on one side will be closer together than the other it doesnt affect the squarenes of the table anyways. So dont wory too much.

So thats about it for the support construction.

Step 2: Table Part1

This is basicaly the start of the construction. As you can see i did it just on the ground, no table as attaching the components together will square the table in directions where it is needed.

In this step is important making the corners square as possible, but again if not, it will not affect the final machine anyway. you will see...

So the table is here up side down the 80x80 will be on the top and the 60x40 will create the bottom of the "bed"

You can see 2x2cm tube on the image held in place by squares, this was helping me to keep same distance between the 80x80 tubes.

Here you measure the squarnes, distances, diagonals and attach the two 60x40 end pieces. Be carefull here as the 80x80 do not neccesary have to be same length... i my case they wherent (poor cut from the shop).

I attached every 60x40 tube on every side by 2 holes. but for these end pieces first only by one hole, check the squareness tighten it up, check again and then finished the rest.

The attaching process in next step

Step 3: Table Step 2

image 1

First i made mysel a plastic helper with positions of holes. No special decision for position, only that i have two holes in some reasonable places. You can ofcourse make metal one which would be even better, or just printout, or measure position everytime, but this is quick. you just want to make 2 marks with puncher so you know positions for drilling.

image 2,3

When tube is placed and attached you drill hole trough the 60x40 and also trough 80x80 underneath. This is so the hole in 80x80 is nicely aligned with the holes you made trough 60x40.

image 4,5

You make the holes bigger in 80x80 so you can tap that hole and in 60x40 so troug one hole; the screw can fit and from oposite side bigger hole so the screw head can go trough. Then you can easily put screw in place from the top and tighten it up.

last image is with connected tubes

Step 4: Table Step 3

I have there 10x 60x40 tubes. i measured the distances marked the positions of tubes to be equaly spaced and by previously described method i attached them one by one.

Step 5: Table Step 4

image 1

I turned the bed up side down and put it to the support construction.

image 2

I centered the bed properly and drilled attaching holes trough every tube. Then moved the bed few centimetes to side so i can tap the holes for screw to the support 80x80 tubes and widen the holes in the bed itself for screw and screwhead (basicaly same process as before). Moved table back to place and attached it by screws to the support. On image there is only the fist hole properly done. This is as i was lazy and made only 4 attach points at the ends.

You can see the bed is wider than the support. I felt like this is desirable support point compare to having the supports all the way to sides as then the 60x40 would be unsupported on longer length. This seemd to me more balanced design.

Step 6: Table Step 5

I decided rising the bed sides in compare to making the X axis bridge construciton high. This is more sturdy solution and less material will be on X bridge less rotation forces if the X itself had been on supports. And this will have one more great benefit.

I decided to raise the bed on 4 places which allow for easy horizontal position adjustment of Y axis, and this will be the benefit, you will understand the importance in next steps.

I made myself again little plastic helper so the holes i make are properly placed. Little cube which acts as spacer to rise the last 80x80 tube which will together create sufficient distance from bed.

On the last image you see how i aligned the top and bottom square tubes. I will keep this side fixed.

Step 7: Linear Rails

I decided to attach the linear rails at the edge od the tube as it is easy to align there.

You can see on image, there are two clamps on one holding place one holding piece of metal which i used to align the rail and one holding the rail against that piece.

All is fine here but from experience i know now that i should probably used thick wall alluminum here as the square tubes have a flaw which i thought i need to avoid, but 90% it would have been just fine.

The 4th image is showing in exaduration how the metal profile actualy looked. So i decided to flatten it out using some sanding pinkish stone and angle grinder which allowed me to put the sanding stone flat on the beam.

I tried adding some flat iron piece on top of the 80x80 top tube but that also ended up as failure because that piece was bent too.

So the last image is that flattened top rail. Just that shiny metal part is of interest here.

One think, if i had used a supported round linear rails i wouldnt need to wory abbout the flatness of the 80x80cm supports as the carriages on round rails can be twisted to the sides a bit so they would automaticaly compensate for this issue, althou then still, the rack would be in angle but the motor position can be adjusted so that would have been fine too.

Step 8: Linear Rails Attaching Process

Ignore the fact here that the raili is not aligned with side of the rail in the first images. it is just to explain what i did here.


Im my rails the hole was 6mm, so when rail properly positioned i drilled a center mark with 6mm drill bit.

Image 4,5

then with 4,2mm drill bit which was a tap hole for 5mm screw i drill complete hole.

Image 6,7

then i made a thread and attached the rail at that place.

Then i moved the clamps to the next closest hole and repeated this process. Thisway i attached the whole rail to the beam.Actualy im my case every 3rd hole was used.

Step 9: Linear Rails Right

Right linear rail was also attached to the 80x80 tube and just put on the square spacers without tightening the screws. This allows for left right motion so when X axis is added it will shift the rail with the top support square tube to the right place and then the screws can be properly tightened. All the rest of the components are basicaly square to eachother so even with introduced imperfections this still can be perfect machine as all problematic parts can be set.

Step 10: X Axis Step1

The holdingplates on images are lasercut. This is only because i was testing the process not knowing if it is good way to go. You can measure holes and make proper alluminum piece that will be perfectly fine.

You could print the holes position on paper use puncher to mark holes and then drill them and you got all u need. I measured my parts and made 3D model and then autocad drawing so i could make the laser cut parts but printing would be probably better as relying on the plastic to hold the drillbit in correct place is not the best solution, printout and puncher would be better way to go.

So why i did this; I attached the holding plates to the linear slides on both sides. moved them to the edge of the machine, put on top the X axis aluminium square tube on them and hold it in place by clamps. Thisway i can mark position of holder on the top tube which will be then in right distance from each rail.

On last image, the plate have holes to attach to the linear slides. The rest are threaded holes to hold the X axis top and bottom square tube.

Step 11: X Axis Step2

image 1

The holding plate was put to marked position and holes where made to the top tube.

then the tube was attached to that plastic holder

image 2

puting in place the bottom tube and marking position on the holder

image 3

the holder is placed on bottom tube to marked position and the holes are made

image 4

note: 6 holes because the 2 in the center just allow the attaching screw from top tube to go trough it in case it is too long, well first the holes need to be sized up.

Image 5

the tubes have only holes the screws goe through, the holder have all the threads. The red are for top tube and green for bottom

Step 12: X Axis Step3

At this stage i was confident that i can make alumimum plates.

I used the plastic plates to drill holes in the alluminum, this was ok but i ended up with bit off position and needed bigger holes for screws in the tubes so i used spacer between hole and screw which could have been possible to use in any case if the attaching hole needed to be bigger due to imperfections.

the alluminum plates on both sides are mirror of eachother so you can make 2 exactly same plates and then just turn one upside down.

Also all alluminum plates were bought at those sizes i just ordedred them like that. Note that the plates will also not be perfectly straight which i was very dissapointed with but at the end after the machine was tighted up it was no deal.

Last image shows nicely the spacers (otherwise that is form different build step)

Step 13: Y Table Adjustment

Now when the X axis is assembeled i have put it ont the Y slides.

As the right 80x80 tube with attached linear rail was not tighten by screws to the spacer cubes i moved the X axis up and down on Y slides which made the left and right side same distance from each other. Then i tighten the right side so it doesnt move anymore.

Note taking the X axis of the Y rails with the linear slides attached... not realy ok, as you loose some balls from them, but they can be put back if u find them.

Trying to put X axis back with the slides on is just not possible, the balls will be all over the place, you need to detach the slides and put them on the rail one by one and then attach the X axis on it.

Step 14: X Plate and Rails

At this stage i had all components related to the sizes of holes of plates so i made the X plate.

1 image

I used again laser cuted pices but printout would have been fine.

2 image

I attached first linear rail to the top, same proces as on Y axis.

Last image

I attached the X plate and the bottom rail was positioned based on it.

The bottom rail doesnt go to bottom part of the X axis because there is needed space for some Z axis component, but this will not affect how sturdy is the machine.

Note that the top and bottom tubes are around the Y plates which makes for nice distribution of all forces during cutting, specialy compared to designs with high X axis bridges. Also the Y rails are already part of the table construction and in the center of the X bridge which makes for sturdy design.

Step 15: X,Y Axis Together

Machine is nicely together. I was very happy to see all axis are nicely moving.

Some cutting job was needed to alow for full X plate movement.

Step 16: X Plate, Mounting the Z Slides

Z axis:

I attached slides to the X plate instead of rails. Thisway i know the Z plate will be as close as possible to the bearing points, again resulting in stronger setup.

All of the small aluminum pieces are just spacers.In this setup the Z axis moves full distance between bed and bottom of X axis.

The spacers might look complicated but, basicaly i used two 10cm thick alluminum pieces, the long ones are for bottom and the small for top.

On top they are simple, they realy are just a spacers.

At the bottom, the slide is attached to one spacer and the other have function as a spacer only but with holes avoiding existing screws.

I wanted to avoid recessed angled head screws as they will try center the screw to center of the hole which might not be same as the screw hole of the component, so thats why i rather made holes in the spacers.

Step 17: X-plate Components

First i add both rails and Z screw as i thought i will be attaching kind of an L profile between the nut and the Z plate. Well that didnt work. I had second hand ballscrew and the nut was kind of non standard, normaly is better just buy on ebay for few bucks nice alluminum holder.

So at the end from a 4x4cm square tube i made the nut holder.

So i tried to make off center hole so the nut is aligned with one side of the 4x4cm square tube. First made small hole, measured the position then widen it up using the tools on last two images.

These holes also wherent 100% right place but it was correct enough to fit the nut nicely in.

Step 18: X-plate Components

Putting the nut inside the hole in the 40x40 tube and marking the nut holding screws positions, making hole added screw, then the same for the next screws. This is best way to do it as you avoid imprecise positioning of the screws.

when the nutholder was finished;

I have attached the slides and the ball screw to the X plate, then took the Z plate and attachd it to the Z rails just on 4 places so i can mark where the holes in the Z plate which should connect to the nut holder should be, and marked their position on the nut.

Step 19: X-plate Components

I made threaded hole to the nut holder and lightly reatached the components to X plate.

Attached the Z plate to the Z rails on only 4 places. The threaded holes for attaching Z rail are in the Z plate. As i used Allen screws it was relatively easy to attach the screws from the side.

I moved the plate up and down and slowly tightening the screws for all components so there is no unnecessary tension due to bad positioning. For screws unreachable from side i had mounting holes in the Z plate.

One more think, the rails and the nut must be in one plane so when attaching the Z plate they all are nicely touching eachother. My nut was bit recessed but 1mm spacer on every hole did the job.

Step 20: Rack and Pinion Decision

X,Y axis are racks and pinoin system and one of the reasons was not just speed but also i thought the screw would just be too long.

I had little manufacturing posibilities so i needed to choose the pinion precisely for motor, making bigger holes to pinions or similar magic would not have been possible for me.

I ended up with relatively small motors, this was also something i wanted to achieve so i can avoid adding weight to the X axis.

I tried bigger setup but i would have ended up with similar result as the pinion on bigger motor was also bigger so...

I choosed 3Nm Hybrid stepper motor, 1.5 Mod rack and 2cm diameter pinion.

Btw no panic from hybrid motors the are still puls/direction driven as every other standard steper motor driver uses.

So for one rotation the motor would move 2 x 3.14 = roughly 6 cm = 60mm.

This is a lot, so the machine will be very quick, but also precise enough as 60mm/1000 (1000 is native encoder resolution on that motor) = 0.06mm is hardware resolution of this setup. this is great news because this is guaranteed and i am not big believer in microsteps as i had experience that they are just not much good.

When measured with micrometer on final machine the backlash was around 0.1mm and repetability was in 1/1000 of a mm, so basicaly exact as there was very little comonents in the system.

About power, 3Nm is 30kg/cm. As radius of my pinion (or better the circle of pinion which is in contact with the rack) is 1cm then my motor can push by 30kg force. 2 motors on Y axis is 60Kg force, that is more than enough.It is good to realize that let say planning wood... the speed and force you can do it has its limits also due to material. I tried to push the X axis plate where was together only 30kg (1 motor force) and i just couldnt. It was very sturdy.

Also no need to make the system oversized in fear of loosing steps, as that cannot happend with the hybrid motors. They can change driving current so they can change force they produce to get to the position needed. Drivers also have hyperposition alarm. This happens if the motor cannot move to desired position. Fantastic if you hit material, the driver will turn off the motor and generates alarm signal, resulting in less possibility taht you destroy your machine in compare to owersized steper which will keep on pushing the axis for ever no matter if hitting an obstacle.

Step 21: Motor Holder System X Axis

I had thought ower many designs and thisone seemd to make most sense. I didnt see this approach before but for me it seemd to address many of my issues. The fact i dont have precise tools means i wanted to make it adjustable in case that it is needed.

So i decided that the motor will be pressed against the rail which will compensate for backlash as much as possible (this is ofcourse used on other systems too but the way it is attached to the axis is different) and will deal with the fact if the rack could be not exactly mounted atc..

As seen on the images the motor sits on door like holder. this is very sturdy solution. It is made from 80x80x cut square pipe. It also alowes for adjustment in case the pinion doesnt sit properly on the rack. It is enoug to move the baring housings little bit around or add spacers to change the angles and move motor to desired position.

The Y axis have even bigger adjustment range

Step 22: Y Axis Motor Holder

Again, just a metal plate this time from some scrap metal i found.

The ballbearing houses are attached to the motor holding plate, and similary as on X axis screw goes trough center of the bearings. This screw is connected to the X axis trough the hole and nut from both sides of top and bottom of the metal plates of the 80x80cm square tube.

This setup have potencial to adjust all angles, the bigger is the hole in the X axis square tube the bigger the adjustability just move left right front back to adjust the angles. For the height you can add or remove spacers from under the ballbearings.

Step 23: Z Axis Motor Holder

Z axis have a ball screw so the holder is different. It is attached on top of the X axis plate.

Again it is from scrap metal i found because it was just too convenient given its size, otherwise i would need at least 100x100mm L to attach the motor. Thisway it fit nicely over the whole X holding plate and went far back without need to extend it. The hole in the center is just because the X motor holder nut was higher than the X plate. This wouldnt happend but i got smaller sized Alluminum plate for X holder than i wanted.

Last image is the holding plate too it have spacers on the attachment holes. this is in case i mount Z motor badly and the belt conencting motor and z screw will need to be tighter, i will just remove the spacers and it will move the plate further back.

Step 24: Other Steps

Somewhere in the process i attached the racks, added springs to pres mottors against them.

There where no holes in the racks so i used linear rail as a blueprint and marked the holes on rack and then drilled them.

I attached them to the machine in same way as the linear rails.

On X axis i used spacer as i couldnt directly align it with the edge as linear rail was already mounted there. Size of spacer doesnt matter too much as you could see if the rail was few mm back it would make no difference.

Step 25: The End

After the basics where done, energy chains, cabling, limit switches, dust extraction was added.

There was some fun with electric noise from spindle, some interesting limit switch setup was made, and some circuits to take care about cases when you loose electricity or signal to spindle to avoid loss of work or damage to the machine or workpiece, but i will leave this till the next instructable.

So here you have it; big machine completly made in a "garage", with high positioning precision, high speed, very well thought force distribution on axis, high squarenss of different planes (like X axis towards bed), safety due to hybrid motors, possibility to take it apart and move, construction made from localy available materials and only few tools used.

The last 3 images show the precision of the machine.

Hope you enjoyed this instructable.

Be the First to Share


    • Stone Concrete Cement Contest

      Stone Concrete Cement Contest
    • Pets Challenge

      Pets Challenge
    • Leather Challenge

      Leather Challenge



    Question 2 years ago on Step 7

    What type of rails are they? I have fond rails that are at most 2000mm in length on amazon. I want some rails that are 3000mm. It is for wood working. Any suggestions on how to do this or where to get some cost effective rails? I was told not to butt together rails but I cant find any long enough.


    4 years ago

    This is great. I've been looking for a larger design that uses rack and pinion other than the aluminum extrusion or skate bearing specific types and your ideas have given me the start I need. It won't be as big as this (only looking for 4X4 feet work area) and I'm looking at an SBR type motion system but I think there is a lot of info I can use here.




    4 years ago

    Could you give a parts list and sources please?


    Reply 4 years ago

    just added. you will find it in one of the comment


    4 years ago

    Looks great! I'd like to build one for myself.

    Please publish your parts list and your drawings. It would make it easier for us.

    What software do you use?


    Reply 4 years ago

    I have 3d model but i will keep that, as that was to help me to decide about thinks, not much about corect dimensions.

    I have searched for known components so here are the racks, pinions, drivers, motors. the rest was second hand or from local shops so i dont have any links for that. The Z axis screw as i recall is 2cm wide and z rails where only 15mm wide not 20cm as the rest but all that was scond hand thats why i made the nut holder atc... you will need shielded cables, energy chains limitswitches but for all that i just choosed what found first but the component list holds the most important parts on which i spent lot of time reasoning why these.

    X axis:

    Y axis

    the links remind me one think i didnt added in instructable th racks for Y axis are made form one 2m long and one 1m long racks properly aligned..

    Z axis

    aluminum plates where from


    4 years ago

    Wow awesome build! I actually had the exact same idea of building a large CNC router into an shipping container! But I guess this isn't actually a shipping container? (since the width of the CNC is "only" 1.6m).

    How much did you pay for this? How expensive is rack and pinion?

    Did you look at "belt and pinion"? It supposes to be better because two belts are laid on top of each other for more support.

    I was thinking of buying a welder and learning welding to be able to manufacture something like this easier. But the gantry could be made out of relatively cheap aluminium profile as well.


    Reply 4 years ago

    It is shipping container, but yeah as i am checking the machine it is 2m wide :D i need to change this in name. the woking area is x1430, y2680. cost was around 3000e. but not sure exactly as i bought many parts and i lost track. I looked into belt and pinion but belts are stretchy and ive seen big machines with belts and i just coulnt imagine 3m belt to move 50kg (or more) bridge without stretching. Yeah even simple aluminum or better dural square tubes would be good choice.


    Reply 4 years ago

    Haha ok, just a small measuring error :P

    3000€? If you're in Europe, where did you get the rack and pinion stuff and how much did it cost?

    I'd love to build a really shoddy but cheap CNC like the one "Absorber of Light" build (you can find on youtube, not even using linear bearings, just skateboard rollers on steel beams!) but the costly things are the lead screws.


    Reply 4 years ago

    Yeah, i was fixated on fact that on of the square tubes on x axis is 1700mm. but i just checked the real width :).
    I wasnt confident enoug to use ball bearings, i did some designs but i needed the table to work at least "good enough" othervise this would be to big loss for me, plus as you could see the amount of tools i have, so it would be hard for make the rails, maybe next time..
    yup, europe, I ordered my stuff from here:


    Reply 4 years ago

    Ah thanks for the link, those are good prices. Rack and Pinion seems like a good way to reduce costs for a large scale CNC. So pressing the pinion gears into the rack is enough to reduce backlash? Can you add another spur and preload a rack and pinion?

    So what was actually so expensive? I'd love to see an approximate cost breakdown.

    Can you mill aluminium with it?


    Reply 4 years ago

    Yes that reduces the backlash but the factory precision of the rack and pinion i used is in lower maybe middle range, there are much higher quality racks if 0.1mm backlash is not enough for you. Also you still can compensate in software, but i didnt fell the need for it.
    I counted in all parts not only table + surely i did some bad decisions for unnecesary expensive components in controll box including the box itself, substract inverter, spindle... yeah maybe even 1000e less. This project took a while to finish so i can tell only roughly at the mmnt, maybe i will look into this sometimes.

    Yes i tried aluminum not a problem. but you need cooling for the bit.


    Reply 4 years ago

    One more question about the hybrid steppers. Are those stepper motor with an encoder, or true servos (DC motor) with encoders that just take a stepper signal?

    Also, where did you buy them / model number?

    Thanks you!


    4 years ago

    If anyone does try to weld their table, I would suggest a single pass E6010/E6011 weld (you NEED DC for E6010, E6011 doesn't really care about polarity much, but DC+ or AC are what it's designed for), I noticed that the root pass didn't really warp the material (single pass E6010/E6011), but additional passes using other rods did, most other welding processes and rods will have more warpage. Another trick, to keep the warpage as low as possible is to clamp the weldment to a thicker, stiffer piece of steel, like heavy wall square tubing, or thick angle iron, a good welding table works too. The weld shrinks as it cools, so position the stiffener, and clamps with this in mind.

    Oh, and E6010/E6011 are special, they are used differently than other rods, try to look up videos and instructions on how to use them properly. DO NOT purchase an AC stick welder from Harbor Freight, the DC ones are just fine, but all you get from the AC ones is frusturation.


    4 years ago

    Thanks for posting. A plasma table runs "roughly" 5k - 15K U.S. for the smaller versions (4 x 4). I keep going back and forth on if I should try to build my own or just keep saving. I don't have any real way to increase my income enough to save faster, but because of the complexity of the process I don't have them time either. So I really appreciate the time you took into posting this! Few hundred more instructables like this and I might make up my mind. Thanks.


    4 years ago

    kinda wrong about the welding aspect. Sure, if you do a 100% weld immediately without clamping or bracing it will warp. On the other hand if you tack all four sides then come back and put a 1 or 2cm weld on a side and the same on the opposite side you should be good to go.

    I welded almost 8cm thick steel missile tubes to almost 10cm thick steel submarine hull, the diameter was greater than 2m and I kept the center of the tube within a few mm of true. I used 50 kilos of welding wire each shift and it took weeks to complete.


    Reply 4 years ago

    You basicaly described the proces how i was trying to do it. I touched the oposite corners atc.. but still the owerhanging top parts of the support where pulled down. It was annoying thats why i dont recomend it. We are talking 1mm per roughly 40cm. But if you know how to make the weld properly then sure, welded construction counts like one piece. So better then more screwed together.


    4 years ago

    For having "little manufacturing capabilities", you have done a wonderful job. Just reading the description made me feel happy about your skill and confidence. Keep it up. Great job!


    Reply 4 years ago

    Thanks :)


    4 years ago

    Vlado - Good work. The world needs more people like you. Ability to make something out of almost nothing is rare. Especially in USA where I live and people would rather play computer games and complain about things that are not important..

    We all have access to good quality Chinese made sold at bargain prices as long as the delivery mail works where we live. As long as those tools are used to make things that help people live rather than things that kill our world has a good future.

    Also you show us that you are a teacher as well. An old favorite saying I learned ...... If you can measure it you can make it. You show us all how true those words are.

    Ed Weldon, California, USA