[Jermsalmond]

Hi everyone. I am new to your forum but have been working on a design for the past few nights to build my first CNC wood cutting/router table. For a bit of background, been 3D printing for a few years, just got a laser engraver and thought it would be awesome to have a bigger router table for more advanced projects. Below is a link to the fusion 360 Link to spin it around and look at it. It is NOT complete. I just got an X-Axis design I like and am starting on Y-axis and Z. Here are some specs on the machine I am trying to build. I plan to finish the design over the next few weeks then start purchasing components. I plan to cut sheet goods first, then dabble in aluminum. This will be used for hobby use but didn't want to make it too weak if I eventually want to turn it into a small business. I'm trying to keep it relatively cheap, like total build under 3K.

Specifications:
Cut area: 48" X x 48" Y x 6-8" Z travel.
Steppers: Nema 23 (I looked at going Nema 34 but couldn't stomach the price upgrade, save for future)
20MM ball screw 10MM pitch (cheap 2010 from amazon/china)
20MM linear rails (also amazon/china)

Link to a model you can rotate and look at. The axis is moveable in fusion360, can't figure out how to make that work in the shareable link.

https://myhub.autodesk360.com/ue2bd6...e63155e96fb5ab

Plan to have a similar Y-axis to the X-axis design that I will put a linear rail on top and bottom of the ball screw.
Z axis is a little TBD for linear motion, probably a short Y axis design.

Any input would be appreciated! Thanks for the great forum, I have found lots of good info on here.

Jeremy

[awerby]

I don't think that extruded aluminum is really stiff enough to do what you're asking of it here. Those skinny uprights holding up the gantry look particularly weak. The triangulation should go higher up. Even better would be to build up walls on either side so that the gantry didn't need them at all. And the gantry beam seems like it will have a lot of flex. I'd say make it much taller and wider, if you really want to use extrusion, or go with a steel box beam, which could have smaller dimensions since it's inherently stiffer.

[peteeng]

Hi Jerm - If your budget is $3K USD I suggest you buy a kit that does what you want. If your really keen on design and build you are about a year away from cutting something. You have started at the machine base and you need to start at the Z axis. Starting at the base means you will run out of space for the Z and it will consequently not be very stiff. The kit suppliers have been through all the issues and sorted them. You have a long way to go especially since you mention cutting aluminium and trying to make a business. All of these are in conflict with your other aims. But have a go. The forum is a good place to gain knowledge fast... welcome to 2022. Peter

[Jermsalmond]

Andrew,

Thank you for the input. I will make uprights shorter and stouter. I was so excited to get a X-axis I quick modeled the adapter plates and plopped on a upright. I think I will put the cutting surface below the side rails, that gets me 4" I can take out of the ganty uprights. I think I can get it where the cross member practically sits on the linear rails. That actually makes the Y axis easier as I can use the same stepper mount and have it hang over one side without the upright.

Thanks!

[Jermsalmond]

Hi Jerm - If your budget is $3K USD I suggest you buy a kit that does what you want. If your really keen on design and build you are about a year away from cutting something. You have started at the machine base and you need to start at the Z axis. Starting at the base means you will run out of space for the Z and it will consequently not be very stiff. The kit suppliers have been through all the issues and sorted them. You have a long way to go especially since you mention cutting aluminium and trying to make a business. All of these are in conflict with your other aims. But have a go. The forum is a good place to gain knowledge fast... welcome to 2022. Peter

Peter,
I looked at some kits and each one had a flaw I didn't like about them. Or they were super expensive like the AVID CNC once you got all the other components. I think Avid is charging a fair price as they do have a turn key system, just too spendy for me to stay married if I buy it. So I started down the path of designing one myself. I used to design weld jigs for skid steers and end effectors for robotic automation but that feels like a lifetime ago (okay, maybe 15 years). Then I didn't have such a wallet constraint as making it simple and work was really the only criteria, and I had a whole tooling shop at my disposal. Now all I got is my shleprock meat-hooks, a grinder, and a dream. Trying to design with off the shelf parts for the larger components is much harder!!

I'm sure I will blow through this budget once I get it all built, find a fatal flaw in the design and have to blow it all apart and start over. That is part of the reason I went with the 80/20 t-slot extrusion, I won't have so much time in the structure that I get discouraged to blow it apart.

The business thing is kind of long shot, I'm not the best at, lets call it "customer service" so it will probably never happen. I tried to 3D print some parts people found on the internet and didn't like the experience of dealing with them. I print useful stuff all the time for family and neighbors, beer is the currency though. I should have said once I get it going I want it stout enough that I won't worry about ball screw or bearing failure.

I see what you are saying on the Z axis I will get that knocked out next and see if I run into any constraints. I have found fusion 360 to be very saleable for size but that doesn't help if it is not stiff enough.

I would like to get you input on what things you feel will take me a year to complete before it will cut? Are you thinking the integration side trying to get it all programmed? Or the iterative process of finding a weak area where it doesn't have the cut quality I am looking for?

Appreciate you insight, keep it coming peter!

[peteeng]

Hi Jerm - If you look at threads of design/builds that take the time to design reasonably well they take a year or more to get from thought bubbles to sawdust. Sure there are projects that take 3 weeks but then there's a lot of rework in them after they get turned on. I design machines for a living so have resources and knowledge to shortcut a few things but even so my latest machine took over 1.5 years to go from paper to sawdust. Covid mucked up so many things in that project. Life gets in the way of projects!! Plus I'm designing for production, not a one off. So many variables and checking have to be done for production machines... but you will find out now your documenting the project!!! Maybe pick the "best" kit and "fix" it? or copy it and fix it...Even picking the extrusion seems simple but cutting it square has its issues. There are many things that are hidden in a CAD model that only come to the surface when you actually build it... My main advice is to complete a CAD build and cost it before you actually do anything. That will draw the line in the sand clearly. If you do anything in the "hope" that it will work or hope for a low cost, usually that hope gets smashed. So get busy in CAD and get the whole thing to a good stage of development & understanding. Peter

for instance your linear rails are mounted on the sides of the base rails. This means if you cut the cross members short/long the linears will have the same wonky shape. Plus its not easy to correct this issue.

Rails mounted on the "sides" are a good idea but difficult to pull off in practice. Since cost is a big issue here use round rails not square. They are much more forgiving in the build tolerance...

[routalot]

I agree with the comment about better triangulation in post#2 and would mention that it applies in both directions.Most of us like to have the same cutting capacity across the table as along it and this leads to the need for equal resistance to cutting forces in both directions.I believe the comparatively slender uprights will allow the gantry to lozenge.It isn't hard to extend the gantry beam to allow the use of a stiffening web outside the upright.

[DanpetersonUS]

Hello,
This is really good.
I suggest you try Long path tool is the very good
program for easily delete, copy & rename long path
files, error, unlock solution.
Try it and solve your problem.
Thanks!

[awerby]

That should help. But as Pete says, you don't want to run out of Z space. Think carefully about the depth you'll need for your thickest project. You need to be able to get to the bottom of it with a tool, and then rise up and clear the highest area of your workpiece without hitting it with that tool. So if your tool is sticking out 3 inches plus a little so the collet doesn't rub, to deal with material that's 3 inches thick, you'll need 6 inches plus the clearance amount accounted for in your Z axis.

[Dean448]

Many years ago I built a cnc router and used welded steel. Steel is pretty cheap compared to aluminum and the welder came in handy over the years.
I built it attached to a bumper pool table. A heavy base on any machine is very important. I don’t think anyone plays bumper pool so it was easy to get. I guess these days you could also use countertop material
It’s not as large as yours but I found that size to be large enough for my needs.
The Z axis is probably the most critical part. It has to clear what ever parts you have and be sturdy. Lots of reinforcements needed in the uprights and then more to support the Z axis.
I used acme lead screws and V track and bearings for X and Y. The V track holds up well to sawdust.
No one was selling DIY machines back then and I know this machine is heavier duty compared to eBay today. If your mechanically capable I’d recommend building it. You can fix or modify anything in the future without having to go back to some unknown company.


Sent from my iPhone using Tapatalk

[Dean448]

I just want to make a additional point about the base. You see all these machines for sale with little or no base. Is that a good design or just an easier way to ship the machine. Could you imagine the cost of shipping a heavy base like mine out of China? Luckily back then I had no idea what a CNC DIY router was suppose to look like.


Sent from my iPhone using Tapatalk

[Jermsalmond]

Peter,
Thanks for the additional insight! You are probably right, before this thing cuts, it will probably be a year. On the getting a kit and modifying it, I kind of enjoy starting for scratch. Most had V grove style bearings and wheels then didn't have the area to add a linear rail or would take major surgery to change that. I tinker alot, weither is it rebuilding the mower deck for a tractor, or fixing my parents halloween frankenstein which no longer moved after getting tipped over one to many times so working through issues in the design is really my happy place. I think I would be happy the first day parts show up and have new shinny toys to tinker on. Why work though someone else's issues when I can create all new ones! lol! I would like to get your take on the X-Y carrage issue I ran into today. So I got a Z-carriage design, mostly bought and adapted to the Y, Have 4.7" of space between the bottom and the scrap board mount. but I am deciding between two different designs for the X-Y adapter as I see that as a critical joint for rigidity. I present Blue Pill and Red Pill.

Quick update:
Z axis has the travel and clearance I am looking for and all modeled.
Gantry is beefed from 1x5 to 3x6 aluminum rail. Also lowered to reduce torque on well, everything.
Had to add 1" spacers to bearing blocks on Y axis to clear ball screw with off the shelf parts. Not happy putting spacers between bearings and load, but it fits.
One option I was kicking around was moving the ball screw on top, removes the need for spacer blocks on bearings but moves the drive system into a different plane than the carriage.

Blue Pill:
Is a 1" think aluminum plate (or 3/8 steel and I will add a fin welded on the outside so it doesn't popcan.
With this I will probably leave the Y alone and extend overall frame to fit everything inside, have a bit of dead space inside the table, but is a beefy design.

Red pill:
Use box tubing to come around the Y axis carriage, and extend it over the X carriage rails. Use precanned 80/20 gussets to stiffen it in both directions, (I drew two in in green, will add more).
Wit this I might turn it into a "U" instead of an "L" by moving the ball screw on top like I mentioned above. This also removes the 1" spacers bringing the Z axis closer to gantry reducing the moment on the X-axis arms.

Here is the updated design:
https://a360.co/3JHu2DI


Thank you all for the input, I am zipping along on this puppy with all your help!!

[Jermsalmond]

So I also worked out on the aluminum rail how to get the aluminum extrusion cut to length effectively and accurately. I found a few companies that will cut it to length to even fractions of inches and machine ends if you want to use tapped holes or those drop in style round anchors. 80/20 will do it or others like https://www.tnutz.com/ will do it too (bold website name, lol). I am not affiliated in any way, nor I have ordered anything from, I did find tnutz to be cheaper per foot than others and have the options I am looking for.
The other adapter plates if simple enough I will hand cut them and drill on the press, if more complex, there is a local metal working shop that has reasonable prices on laser cut components. I use them for work quite a bit.

[Jermsalmond]

I just want to make a additional point about the base. You see all these machines for sale with little or no base. Is that a good design or just an easier way to ship the machine. Could you imagine the cost of shipping a heavy base like mine out of China? Luckily back then I had no idea what a CNC DIY router was suppose to look like.


Sent from my iPhone using Tapatalk

10-4 on the base. That I will probably weld from steel. I do have a welder and hand plasma cutter (sooo much fun, would recommend) and a shop full of tools to make things. I agree it should be beefy and not bring in any movement. I am just not setup to do thinks like drill and tap 20+ holes in a straight line for the linear rails then do it again on the other side, or make mount plates and weld them up square enough to make the linear rails not bind. That is why I am using the 80/20 extrusion. I can create fixtures with 3D prints to locate the rails off one edge, leave one corner of the overall base a bit loose and slide the carriage back and forth while tightening up the rail then the frame. Also If I hate it, and the project dies, the 80/20 rail will be reconfigured and make a real sweet storage cabinet.

[Dean448]

Your model shows plenty of +Z axis travel but will it go low enough?

I would recommend adding the base plus spoil board to your model. Then make sure the spindle is accurate. No long cutter.
Lay it out so the collet is about 1/2” above the bed surface.
The Z axis should be very near or at its end of travel. Then determine if you have enough clearance to the stationary part of the Z axis. The X axis rail won’t be a clearance problem, it’s the base of the Z axis slide. You will be doing most of your cutting between this position and a couple inches higher. Does everything look well supported over this range? Or is it strung out?

I found that adding Z axis travel doesn’t really help unless your cutting on a mound shaped part. Again the machine proportions don’t need to look good. They need to be functional.

I’m sure you caught this but if you make this machine 48x96 for example it won’t have a cutting range near that.


Sent from my iPhone using Tapatalk

[peteeng]

Hi Jerm,
Only thing to add at the moment is that unsupported round rails for the Z is not very stiff. Plus think about using a Z motor that has a brake in case the Z axis falls which they can if the power goes down or drives fail etc etc. Keep adding detail that's where the devils are....Peter

[Jermsalmond]

Your model shows plenty of +Z axis travel but will it go low enough?

I would recommend adding the base plus spoil board to your model. Then make sure the spindle is accurate. No long cutter.
Lay it out so the collet is about 1/2” above the bed surface.
The Z axis should be very near or at its end of travel. Then determine if you have enough clearance to the stationary part of the Z axis. The X axis rail won’t be a clearance problem, it’s the base of the Z axis slide. You will be doing most of your cutting between this position and a couple inches higher. Does everything look well supported over this range? Or is it strung out?

I found that adding Z axis travel doesn’t really help unless your cutting on a mound shaped part. Again the machine proportions don’t need to look good. They need to be functional.

I’m sure you caught this but if you make this machine 48x96 for example it won’t have a cutting range near that.


Sent from my iPhone using Tapatalk

I will double check it as it is not locked in yet. I have some flexibility with the plate I put between the Y axis and Z axis to adjust it up and down, Right now it is at the highest where I get the full clearance for a 4" part plus the spoil board. I have a lower setting if I am doing more things that are 1" or less. I really hated how all the different Z' axis were designed. Has anyone every tried flipping where you would mount the spindle to the back of the linear rail and mount the front of the mount blocks to the Y axis? how you would assemble the bolts it would be a pain but it makes it so the bit never retracts up so high and you can get the full range of motion? Or are people using super long bits when they cut 8" deep so they don't have to worry about part to spindle interference?

I know in my first post I said I wanted 6-8 inches of Z travel then once I modeled it, the gantry was too tall for the 1% of the time I will cut something that tall. So I settled on 4" as I can see myself putting in something made of 4x4 posts (yea they are 3.5") into the machine and cutting on them. The Z+ travel was out of pure laziness. I picked a Z-axis that had 8" of travel, modeled it, and found I only wanted about half of it. I will try to probably end up with one with 6" of travel but didn't want to redesign the darn thing again.

[peteeng]

Hi Jerm - Better to redesign in CAD then in reality. takes quite a few iterations to get a machine right. Be patient and keep at it you will be pleased in the end. I mount the rails on the Z axis on most of my machines and the cars on the saddle. Works for me. 4" of Z is not enough for my jobs. Minimum would be 6" 8" if you use fixtures alot. If you don't want to have the Z hanging out all the time you can use a bolster or thick waste board to lift the job vs drop the Z. But if you use a vice or a fixture you will need more then 4" unless you are always only doing thin sheet work...Peter

Plus you have to decide where that 4" is. This is dictated by your std waste board height, your general tool length and the type of job your doing... fopr instance if you uise the front of the base as an apron ( to do edge work, ie the panel is vertical ) you may want the tool bottom to go under the waste board height...

[Dean448]

I think this is the difficult part of the design. I have a second machine that has 4 inches of clearance to the Z base (using a 2”spoil board/fixture) the pad of the axis (that goes up and down) is 6” and sturdy. The collet is 8” below that pad and the slide has a 12” range.
Tons of +Z travel but not a great deal of clearance. And it’s somewhat strung out.
And you need to hold all this rigidly.
This machine has a 12x12 XY range so I just need to hold it stationary.


Sent from my iPhone using Tapatalk

[peteeng]

Hi Jerm - Here's an image of my Z axis on my machine called Scoot. Shows the rails on the z axis not the saddle. I have tidied the wiring since that photo Peter