Contemplating a build and could use some input

Greetings! I apologize in advance for the long post but I like to provide full context and thought process so you can really get your head wrapped around where I'm going.

What I'm thinking is that I want to create something in the 48" x 48" range with maybe 6" or so of Z. The base will be a plywood torsion box, upon which I will mount linear rails as the foundation of the mechanics. Custom 1/2 inch aluminum plate gantry uprights, extruded aluminum 80/20 for the gantry with linear rails for the axis. 1/2 inch aluminum plate for the Z axis, again on linear rails. All pretty basic concept. I know, very lacking in details.

Here's the rub, I already have a CNC router. I have a Bob's CNC E3. She's a plywood machine with Nema 17's. 2 drive the gantry as X, 1 drives Y, 1 for Z. X and Y are belt driven Z is ball screw. All powered through an arduino uno. Limit switches are also in place. I want to rob all of this for my new build and drive the new CNC with all ball screws.

The gantry will drive via ball screw mounted, at least in my head, mounted outside of the rails. The Y axis mounted atop the gantry.

Now, I'm pretty much positive there may be a few issues that I'll acknowledge out of the gate. My steppers may be too small, they (arduino) may need to be reconfigured to run them correctly for ball screw vs. belts, and I definitely have to change the parameters when it comes to the actually running the machine as the work areas are vastly different. I'm not there yet.

What pitfalls might I be facing when it comes to structure and steppers? My little Bob's does a fairly good job but, it lacks rigidity and thus, accuracy. It also lacks size. I feel like the NEMA 17s are alright, but everywhere I turn I see NEMA 23s. Is 80/20 extrusion at 1545 profile rigid enough at ~48 inches? Any guidance at this stage would be great. I'm no CAD master, but it'll help me start to draw out what I want to do without having to scrap it a dozen times because my ideas won't work.

If you're talking about mounting linear rails directly to your plywood box, I think that might not work. Linear rails are very particular about flatness and orthogonality. But some are less particular than others - usually they'll give you specifications. You might need to lay down some aluminum and shim it carefully to provide a flat square subsurface for the rails. You might want to provide beefier uprights for the gantry, especially if it's higher than a few inches. I'm also skeptical about those NEMA 17 motors being powerful enough. It might be a better idea to leave your Bob's CNC as-is, and start from scratch with a different system for this machine. And no; that 80-20 extrusion doesn't have much stiffness over a 48 inch span; it will need support from something.

Hi Rehctub - Everything is a compromise but here's my suggestions:
1) do not make a column machine , make a high rail machine its stiffer and controls dust better. If you use the E3 to make the parts use good quality plywood and it will be a great machine
2) using the N17 motors is fine if you use 5mm pitch but it will be a slow machine compared to the belts. So if speed is important eg you do 3D carving then go to 10mm pitch or 20mm pitch but then you will need N23 motors I think. Belts have nearly zero inertia so don;t take up motor torque like screws do. 90% of the torque is taken up by the ballscrews if you do the math so that's why N23s are used... I have made a couple of belt drive machines and like them for speed say 20m/min
3) Square rails are good but I feel you want to go round rails a) they are cheaper b) they are foundation tolerant and they will suit your construction/budget better by the feel of what you are saying
4) design your next machine to make as many parts on the E3 as possible. Ive found plywood coated in epoxy or even mdf with melamine to produce very acceptable Maker grade machines
5) In the scheme of the new build the motors are a very small cost component and I'd go N23's keep the E3 together and sell it off after the new machine is built.

Frankie 1 - is a formply machine. The gantry is a laminate of 2mm aluminium and 17mm plywood. Its very stiff, note it is high rail. I've built a few column machines and I doubt I'd do another unless the demand was for a side loading the machine. I have had a couple of events on the mdf machine which have proved they are really strong! I'm about to add the ballscrews as soon as they arrive. If you search for Frankenrouter on the forum you will find its build thread. It uses 16mm round rails for the XY and square for the Z as I had them

MDF Base - is a 48" plus wide machine with MDF torsion box base and high rails. Its been running for two years plus and its worked out very well. I prefer formply or beech ply over MDF. Ply is a bit stiffer but not as flat and the ply has to be sealed. I used MDF to see how it would go as its half the price of ply. No complains so far with it.. The machine is called Scoot and it uses square rail. Scoot now has a 1.5kW ER16 collet spindle and it cuts aluminium easily...

Make your own plywood gantry it will be stiffer then the aluminium construction extrusion and damper. Plywood is 1/10th the density of steel so you can use quite a lot of it and it still doesn't weigh much. Solid and very thick parts don't vibrate like extrusions do and in timber they are stiff and strong. Seal them well to keep moisture out and it works great.

Go forth and design - Peter

I agree what was said before, over 48" 8020 is inadequate. Even 8040 as you see in many other kits is not enough for the gantry beam. However being easy to cut and assemble extrusion does simplify the build process tremendously while being straight enough for most hobby needs. When I first started I didn't have the precision tools or the experience to make a good machine and I struggled to achieve accuracy and repeatablility. Aluminum extrusion would've been a great option. If I had to choose how to start from nothing today I would pick an upgraded extrusion kit. The kind with profiled rails and ballscrew. Wood machines like peteeng's have a huge advantage, price. You can make and remake parts without a huge hit to the pocket. Built properly plywood or mdf will be plenty stiff and you have a current machine to make the parts for it. I'd definitely go with nema 23's over 17's for the machine dimensions you are building. I can stall my nema 17's on my 3d printers that have no real mass. Anyways, good luck in the design process. Very rewarding to design and build your own machine.

Hi CncB - heres a paper on plywood for machine part construction. I'm sure Bob at Bobs cnc would agree. Peter

Thanks for all of the replies!

I'm reconsidering some design considerations, or at least not sure what makes the most sense. I think I like the raised wall design. Less debris gets spit onto the rails. It will provide for shorter gantry sides, though this may also limit, or at least guide, drive design.

For clarity, as it may help regarding the extrusion, the "80/20" I was referring to is actually a brand, 80/20Extrusions. They produce a 6015-T5 extrusion that has a compression strength comparable to steel. When I calculated the deflection of their 1545 profile (1.5" x 4.5") I came up with .00002 with a 100 lb point load at the center of a 48" span.

The obvious problems are, I would need to use one of their metric profiles as most of the available rails and such are going to be metric. As mentioned, vibration dampening is an issue but, the benefit is, if it's as string as it says, it's strength to weight is a huge advantage. The big issue for me is I'm hesitant to do it without hearing someone else's experiences with it... just because a manufacturer says it's that rigid doesn't make it true.

My current machine, Bob's. It's made me a pretty competent operator and taught me a lot about using a CNC. It's also taught me that mistakes have consequences. If I were to construct a plywood gantry, it would have to be a permanent construction, as Bob's is screwed together and while the engineering works in a vaccum, forget to set Z just once and the torque place on that gantry will stretch those holes out and introduce enough slop to swing a dead cat through. So a plywood gantry has to be one, permanent fixed, and fused unit. That's very limiting. The good side is I've made more that a few torsion boxes to within a few thousandths of dead flat so the base, and if needed, a gantry, can be done strong. If possible I'd like to keep the gantry modular in case I need to replace, adjust, or simply want to rework something.

As for Bob's future. I don't have the space for both and it's a little tired. It still runs fairly well but I don't know what value it really has left in it. I would rob the board from it and maybe the switches but that's not really worth the money saving effort if I'm not using the drivers and steppers and it looks like NEMA 23s are the right move for a machine this big. Mind you, while I "get" the electronics, I really don't have the first clue about this. It's not from lack of ability but this is an area I have yet to delve into with CNC. I've been focused on the mechanics of it and not the powering mechanism thus far.

Speaking of steppers... so, I'm not sold on my drive system choice. I primarily do wood. I'd like to maybe be able to do some light aluminum but it's not a priority at all. Speed has never been a real issue, well, it HAS, but the issue has been Bob's was not rigid enough for the speed it was capable of and would produce bad results due to excessive deflection. Every drive seems to have its compromise. I'm even toying with rack and pinion now. I've only ever stalled my belt drive NEMA 17s when something really dumb has happened, like I failed to zero something or something obstructed the gantry. Granted, because of deflection, I routinely only cut at 30ish IPM in hardwood at a .06 - .1 depth with a .25 endmill. I'd certainly like to improve upon that but I'm not chasing speed as much as I'm looking to maintain precision and reliability. With that, what drive SHOULD I be considering? My BIG goal is to be able to do guitars.

So, thus far, we've settled the stepper issue (at least basic size, torque not withstanding). I still feel strongly that the torsion box base is going to be a good choice with torsion box raised side walls. Bedded linear rails, SBR20. NEMA 23 motors and drivers. The rest is TBD. Let's hear some more about gantry and propulsion. A balance between speed and precision, I don't care about noise, as I'll likely be routing with a VS wood router.

What say ye?

Hi CNCB - aluminium construction extrusions are convenient but convenience has downsides. Strength of materials rarely is a concern in machine design. You won't break or bend steel, aluminium or plywood if it is rigid enough. Plywood thread inserts (for bits that need to come apart) and glued construction is strong and reliable. You can even use std metal thread screws and taps and use PVA glue to set the screws and they work fine. Setting up my last MDF machine base I had an electronic glitch that made one wall motor go one way and the other wall go the other. The bed twisted up visibly and groaned. I thought it was going to break but it didn't and when it returned when I hit the big red button it went back to its place and hasn't been a problem since. So I'm confident using timber as a construction material in terms of strength. Your guitars will be placed under much more stress then your router structures. I've been involved with guitar parts on and off for years. Strings are tightening to very high tensions and guitars change shape over time.... timber under constant high stress can creep but router parts are not under constant high stress.

Rack and pinion on such a small machine will be a disappointment and cost more then C7 ballscrews. Go ballscrews. 30ipm is 750mm/min. Using 10mm pitch screws you will get 5000mm/min easily. (500rpmx10=5000) and heaps of force to cut timber. dia12mm or 16mm will do the job and I'd go 16mm for rigidity and transport security. The 12mm can be bent in transport and will sag a bit more then 16mm so will vibrate more in use.... I regularly cut hardwood ply at 3000mm/min.

So I find it interesting that a guitar builder would consider extrusions as better then timber. Timber allows you to make exactly what you want in terms of geometry and consolidation of parts. If its epoxy sealed its stable and damp. There are many threads here about extrusions vibrating, they have lots of thin free end bits that love to vibrate. So why go there? Plus if you look at the $/kg price for a construction extrusion vs a std square or RHS extrusion you will find its really expensive.. So if AL is the go use std SHS or RHS in heavy wall and its a much better solution then construction extrusions. But I'd go cabinet grade plywood in your case for base and gantry all glued together....... Cheers Peter

regarding the bending calculation. Gantries are under combined bending and torsion. Torsion is quite big when you do the numbers and construction extrusions are very poor in torsion due to having not much material on its perimeter (drive shafts are round for a reason) do you get that I do not like construction extrusions? :) Peter

By the way if you use SBR16 the stack height of the drive nuts and bearings for a 16mm screw are the same so you don't need spacers or steps in the gantry, makes design much easier and SBR16 easily do your job...

Quote:

---

Originally Posted by Cncbutcher

... Let's hear some more about gantry and propulsion. ...

---

I found that 10mm pitch ballscrews with low inductance nema 23 steppers (below 3.0 mH) were a sweet spot for my machine that is built along the same lines as yours just a tad smaller

It's precise enough yet gives me speed over 7m/min.

There's certainly much food for thought here. I appreciate your advice Peter. I'm not by any means married to the idea of using aluminum extensions. It might help to know a bit about me...

I spent many of my formative years in the US Army. Post service, I built houses for a living but my injuries sustained during my military career really precluded me from being able to do that forever. So, what does a 36 year old man naturally do? That's right, nursing school. I obtained a Bachelor's of Science in Nursing, later a Master of Health Leadership. I work for the Department of Veterans Affairs as a Caregiver Support Coordinator. All throughout, and even before, a constant was music. I identify as a bass player but lately I've been honing my rhythm guitar chops with my quilt top 60s profile Les Paul. I'v always gravitated towards music and woodworking to express my sense of creativity and escape the demons that sometimes chase folks like me. I also enjoy metal working but enjoy doesn't always equal good at. I don't build guitars yet, but I have a billion on them in my head, just waiting for me to get them built. This is a step in that direction.

With all of that said, my military mind demands order and some degree of precision. As I have grown as a person, now a grandfather, I have certainly learned to accept the nuances of imperfection as being part of the unique beauty in some things. I feel like I have a preconceived notion that the inherent desire of wood to move takes away from its precision, and I know it does, but I also know that I'm not designing and building a rocket surgery machine so compromises have to be made and I have to grow a little more.

Sorry for the digression. It's barely 5AM and I'm waiting on my coffee to brew.

I am intrigued by the gantry design used for Frankie. It seems very simple. How did you proportion the "T" portion? Might I ask why the Z carriage is so tall? Is that simply to aid in accuracy?

You've pretty well sold me on SBR 16 with 16MM ball screws... it makes perfect sense and your experience certainly carries weight.

Hi CNCB -
1) I make moulds and tall parts (bowls and boxes) which means by the time you add wasteboards or fixtures and long tools I need lots of Z. You will too with guitar bodies...
2) Frankenrouter by birth had some design rules. a) use parts I had b) make as many parts as I could myself. Frankie was a backburner project but looks like it will be a good production size machine. Having made several machines I now know several of the sticking points such as the solving the issue of mismatched stack hts with bearings and drives. Frankie uses round rails as I have not used these before but they solve a couple of issues like stack ht and cost for an entry level kit router. All parts of a machine are connected and change or commit to one thing and everything else gets affected. Like life.
3) Dimensions of the gantry - a) I had 2mm aluminium sheet in stock from an old project and laminated aluminium with ply or timber core solves several problems b) I scored some 36mm laminated pine at the hardware store really cheap as it was forklift damaged so the "top" of the gantry is 2mm al / 36mm pine /2mm Al so it worked out nominally 40mm thick. It is actually 41mm thick as I used a scrim in the gluing process. The depth of the T is defined by the "top" height of the Z axis so the air clearance of the gantry/z axis is maintained. So I made two webs and then glued some 2mm AL across the bottom to max the gantry depth yet maintain the air height.. Since I place rails on the top of the gantry (vs on front of the gantry, I do this to decouple the issue of having 8 bearing cars all in the same place) this means I need to go "down" to stop it being really tall! So it has worked out quite neat at the ends. My design struggle over a few machines has been to simplify assembly especially the saddle, make it easy to level and square the machine and to minimise/remove parts like using spacers to correct the difference between bearings and rail stacks... Machine design has several "rounds" and if you use CAD I find about 20 rounds is needed to boil a design right down to basics. If you have a couple of days reading in you look at the Brevis HD thread and the Frankenrouter thread. They capture a few years worth of chinese puzzles. Peter

note - laminating the timber with aluminium makes it very stable. This addresses your issue of wood wanting to move. But even coating the timber in epoxy or acrylic paint will improve this issue considerably . There is a commercial router maker in Germany that builds their very large routers from plywood. Plywood was considered an "advanced" material at one point on aircraft manufacture not too long ago.

Hi CNCB -
1) I make moulds and tall parts (bowls and boxes) which means by the time you add wasteboards or fixtures and long tools I need lots of Z. You will too with guitar bodies...
2) Frankenrouter by birth had some design rules. a) use parts I had b) make as many parts as I could myself. Frankie was a backburner project but looks like it will be a good production size machine. Having made several machines I now know several of the sticking points such as the solving the issue of mismatched stack hts with bearings and drives. Frankie uses round rails as I have not used these before but they solve a couple of issues like stack ht and cost for an entry level kit router. All parts of a machine are connected and change or commit to one thing and everything else gets affected. Like life.
3) Dimensions of the gantry - a) I had 2mm aluminium sheet in stock from an old project and laminated aluminium with ply or timber core solves several problems b) I scored some 36mm laminated pine at the hardware store really cheap as it was forklift damaged so the "top" of the gantry is 2mm al / 36mm pine /2mm Al so it worked out nominally 40mm thick. It is actually 41mm thick as I used a scrim in the gluing process. The depth of the T is defined by the "top" height of the Z axis so the air clearance of the gantry/z axis is maintained. So I made two webs and then glued some 2mm AL across the bottom to max the gantry depth yet maintain the air height.. Since I place rails on the top of the gantry (vs on front of the gantry, I do this to decouple the issue of having 8 bearing cars all in the same place) this means I need to go "down" to stop it being really tall! So it has worked out quite neat at the ends. My design struggle over a few machines has been to simplify assembly especially the saddle, make it easy to level and square the machine and to minimise/remove parts like using spacers to correct the difference between bearings and rail stacks... Machine design has several "rounds" and if you use CAD I find about 20 rounds is needed to boil a design right down to basics. If you have a couple of days reading in you look at the Brevis HD thread and the Frankenrouter thread. They capture a few years worth of chinese puzzles. Peter

note - laminating the timber with aluminium makes it very stable. This addresses your issue of wood wanting to move. But even coating the timber in epoxy or acrylic paint will improve this issue considerably . There is a commercial router maker in Germany that builds their very large routers from plywood. Plywood was considered an "advanced" material at one point on aircraft manufacture not too long ago.

edit - I'm a mechanical engineer and have been designing and building machinery and structures for over 45 years. In the last 7 years I have been concentrating on hobby level cnc machines. I have seen some trends. A machine with no purpose does not live well. A machine with a purpose is easier to design, as its purpose gives it direction and limits and it will get used well. A general ,purpose machine will have compromises and over/under designed bits to take care of its unknowns and it will be bigger or overdone in many areas. So if you are thinking about guitars then I suggest you skew your design towards that purpose. A 4x4 machine is a general purpose machine intended to accommodate a half sheet of ply. Fair enough... You however will be building bodies and necks and don't need 4x4 unless you have the free space, especially depth. One thing you will encounter is doing a neck and turning it over. Registering a 180deg turn is a bit tricky to get right. So eventually you will want to do this on a rotatory axis. But rotaries take up end space and I think you should think about a wider machine with less depth. I rarely use the back of my 4x4 machine and I can always index something if I have to... So think about say 5ft wide by 3ft deep. This will allow you room for the future rotary bits. The rotary is best aligned with the gantry as this is then a one axis move. If you align it transverse to the gantry you may find the finish is not quite as good as the two axis will be ever so slightly different. As you will be doing 3 axis "carving" this will also direct things like axis speed for finishing, you want fast finishing speeds as your stepover is going to be small (say 0.25mm) and the toolpath will be very long... So more thought balloons for you over coffee. :) or consider having an "apron" or a well at the front of the machine for this function. A gantry machine has extra space at the front of the machine due to the spindle overhang, this is where a rotary can be placed. Sometimes they are placed to the side of the machine as well. Good to think about this sort of thing and future proof the machine... The future gets here faster then we think. Maybe a gantry machine is not the go but more like a moving column mill so it easy to front load longer objects... Peter

seems my reply timed out so here it is again...

Haha! A rotary axis! You're teaching me 4D chess and I'm still playing checkers! I like it!

I have considered an open apron and I am probably going to incorporate that as I would like to try my hand at CNC dovetails and such as well. My 4x4 size is more or less arbitrary... in my head I think ' what if I wanted to build 2, neck through 34" scale bass guitars with bodies the size of... IDK, like a 335?' 4x4 seemed big enough but the specific size was yet to be truly ironed out. I'll give some thought to this and really hone in on what makes sense.

Looks like I'm closing in on breaking out Fusion 360 scratching some lines.

Hi CNCB - A gantry machine needs 4 axes/drives, if you make a narrow machine with an apron you can delete the 4th linear axis and put that budget into the rotary...Start with some concept hand sketches to develop the ideas fast then move into simple block geometry in Fusion. It would look like a mill vs a router. I'm away for 3 days so will get back to this next week. :) Peter if you look at the end of my Milli thread you'll get some ideas on moving column mills....

Quote:

---

Originally Posted by jckstrthmghty

I agree what was said before, over 48" 8020 is inadequate. Even 8040 as you see in many other kits is not enough for the gantry beam. However being easy to cut and assemble extrusion does simplify the build process tremendously while being straight enough for most hobby needs. When I first started I didn't have the precision tools or the experience to make a good machine and I struggled to achieve accuracy and repeatablility. Aluminum extrusion would've been a great option. If I had to choose how to start from nothing today I would pick an upgraded extrusion kit. The kind with profiled rails and ballscrew. Wood machines like peteeng's have a huge advantage, price. You can make and remake parts without a huge hit to the pocket. Built properly plywood or mdf will be plenty stiff and you have a current machine to make the parts for it. I'd definitely go with nema 23's over 17's for the machine dimensions you are building. I can stall my nema 17's on my 3d printers that have no real mass. Anyways, good luck in the design process. Very rewarding to design and build your own machine.

---

Both 8020 and 8040 stiffness and strength are dependent on the extrusion type, as you can find each in thin or thick wall. However, I do agree that neither of these extrusions (regardless of thick wall) is suitable for a 48" gantry. I went with an 8080 thin wall extrusion on my design and I am satisfied with the results. More importantly my Z feed (which I purchased as a kit is certainly not up to the challenge of stiffness. The assembly uses two 12 mm polished steel rods riding on two sets of nylon plain bearings. It flexes easily. I have ~40mm of unused Z-height and will trim off the excess to reduce flexing. I am using a single 425oz.in/3Nm NEMA 23 for both X/Y and the performance is adequate for medium duty DIY setup.

Hi CNCB - My suggest configuration for your machine is like the Mori M1. Each major component is a plywood torsion box or solid timber. It can be as wide as you like, with an apron and space for the rotary axis. You need to decide what the swing of the axis needs to be. You can also make the "bed" of the machine removeable so you can have different beds for different height jobs (some call this a bolster vs a bed). You may also want to swing the body. I assume you are an electric guitar builder.... Peter

Hi - heres an example of a front loader. Peter
https://youtu.be/YXD6aXb6PmQ

For the last 15+ years, I've been working on a mostly wood 4x8 machine. Table is a baltic birch torsion box, and gantry is a torsion box. To mount linear rails to wood, I epoxy phenolic plates to the wood, and machine it flat to accept the rails. An alternative that doesn't need machining would be to bolt a small extrusion to the wood, and mount the rails to the extrusion. Greatly spreads the load, and keeps the wood from crushing or deforming.
Here's a link to my "build", that hasn't been updated in years.
https://www.cnczone.com/forums/cnc-w...0-autocad.html