An onslaught of questions from someone new to CNC

[Rattanee]

Hey everyone,

been reading through these forums for a while now, trying to find information for a CNC project I'd like to start, but have had little success in finding specific info I'd need. But based on what I saw so far, I'm confident this group's the one to probe for information.

I'm more or less clear on most of the terminology but excuse me if I get anything wrong.

The Project details:
Goal:
Building a universal CNC that can work wood, plastics and sheet metal of 3-4 mm in thickness.
Work area: 2200mm x 1400mm
Work height: 200 mm
Building material: Aluminium
Moving gantry, fixed table

I'm planning on completely custom made mechanics made of all aluminum to keep the machine as light as possible.

Planning to use custom milled ball screws for all three axis, and I want to use a handheld router's motor for milling.

My questions:
1. What kind of force is required for my x and y axes for cutting through 3 mm sheet metal? (Of course I mean the forces needed for the actual cutting process, not including moving the weight of the gantry)
I'm pretty sure that the force required for this is more then enough for any kind of wood cutting.
2. Is there a spreadsheet or somesuch available for calculating what torque motors I'd need for each axis?
3. I'd need an approx. 2400 mm ball screw to move my x axis all the way along the workspace. Can this be done without the ball screw being too heavy or sagging in the middle?
4. I'd like to cut plastics like Lexan with this thing, which are best cut using lasers. Does anyone know if there's a laser cutting unit available for this purpose?

I'd like to keep this machine precise despite it's size, as I'd use it for pretty much anything from cutting small printed circuit boards, to making large custom shaped sheets of wood or mdf.

Thanks for any and all of your help!

[Fish4Fun]

Rattanee,

Welcome to CNC Zone!

A moving gantry is going to present real problems with cutting metal. Building from aluminum can be problematic if you are wanting to cut metal. Using a router is going to cause real problems cutting metal. A machine this size is going to be very difficult to DIY to cut metal, even aluminum.

You will likely be better off building a small desktop router for PCB work, though many have built large format machines that are capable of PCB work. The large format DIY builds are typically routers designed to cut wood & plastics and sacrifice some resolution and accuracy for higher feed rates. If you absolutely have to cut metal and plastic on the same machine, you should consider going with a laser or plasma cutter instead of a spindle for cutting the metals. With a laser/plasma cutter there are no rigidity issues like there will be with a spindle. If you build a very rigid aluminum machine, you might expect to make light cuts in aluminum with the right spindle and the right feed rate, but you will certainly not have much luck with cutting steel.

To answer your questions in order:

1) The forces required exceed the capabilities of any typical aluminum build by a lot. The main problem is not the torque on the motors, but the deflection of the machine itself. A typical metal cutting machine out weighs it's wood cutting counterpart by a couple orders of magnitude for any given work envelope. This weight is requisite because large steel sections are needed to prevent deflection.

2) A lot depends on the machine specifications and the type motors used. For instance steppers experience their highest torque at low RPM and the torque drops off significantly as the RPM increases. Servo motors have fairly constant torque over their entire RPM range. As a first approximation you should take the total mass that you will be moving and determine what your maximum travel rate will be then calculate how fast you need to be able to accelerate to or from that speed. When you reach a number, double it.

3) In a moving gantry you will likely have the long axis set up as your "Y-Axis". With a 1.4m X-Axis travel, your gantry will likely be at least 1.5m wide; if you decide to use ball screws, your build will almost certainly require two (one on each side). "Sagging" is not really the issue (though it can be problematic), the real issue is "whipping". For your 2.4m Y-Axis you will likely need to look at 20mm to 40mm ball screws, or consider a rotating nut design.

4. I have been researching DIY lasers, but there is a lot to learn. You can order the "CO2 Laser Tubes" fairly easily off ebay and other sites, but making that laser tube into a cutting head appears to be rather involved.

I'd like to keep this machine precise despite it's size, as I'd use it for pretty much anything from cutting small printed circuit boards, to making large custom shaped sheets of wood or mdf.

Yes, we all want 0.0001" accuracy in an 8ft x 4ft work envelope with 1000ipm rapids for $2000 that we can fold up against the garage wall when the wife gets home ;-) Sadly, like most things there are trade-offs. You can build for speed, accuracy, size, cost or weight, pick three, the other two will be out of your control.

The best advice I can give you is to start with a small desk-top router accurate enough for your PCB work and small wood/plastic parts. You will learn enough from it to make informed decisions about a larger format machine. You should be able to build something like John's (Microcarve) "A4" machine for ~$500 to $1500 (depending on steppers, controllers, software, rails, lead screws, etc, etc), and that money will be WELL SPENT. John sells his "A4" machines for ~$400 (w/o steppers, controller(s), software or power supply), so you should be able to source the parts for less than $400. The steppers/controller(s) and Power Supply can be as cheap or expensive as you want. Building a 4ft x 8ft format machine for your first build, you will make bigger mistakes than $1500 can fix several times before you finish the project, so consider a desktop build an inexpensive lesson.

You might consider reading some of the build threads in the woodworking machine forum here @ the CNCzone. I know you think you want to build a metal cutting machine, but I don't think you are quite ready yet.

Good Luck!

Fish

[Rattanee]

I'd like to stick to a moving gantry simply because I have no room to move a 2.5 meter long desk back and forth completely. (Any way I look at it, that would need 7.5 meters of space at a minimum)

The design I have in mind would indeed be very rigid, But I guess I can see your point, no matter how you look at it, aluminium is softer then steel.
I would actually consider laser in that case, as I know plasma isn't the most accurate tool in the shed. I'd need precise cuts in metal and plastic.

I plan on using steppers indeed as servos seem to be a lot more expensive and aside from a more consistent torque over the rpm range, I don't see much advantage. As I said I want this precise, not fast. Is there a formula that I could use to calculate this?

I am planning to have a crossbeam of the gauntry above the table which will have the cutting tool of course, which would be attached to linear slides along both edges of the table with beams in an upside-down V shape. Both ends of the V shapes would be connected by crossbeams underneath the table, and the ball screw and nuts would go through the middle of both beams. I hope this makes sense, I'm not a native English speaker.

I believe this would be a rigid enough design for any woodwork, I had no idea of what's involved in metalworking. I based the theory that I could use a router motor for metalworkign on the simple theory that a dremel cuts through metal fairly easy with a cutoff wheel, and it's not much different.

What do you mean by rotating nut design?

As long as there's electronics involved, I'll enjoy digging into that topic I have both a knack and a passion for electronics, be it building or repairing.

Nawh.... I don't want it to fold up against the wall... I'd like it collapsable so it can fit into a suitcase!

But in all seriousness, size isn't really an issue, neither is speed. I'd need accuracy and a fairly low cost.

Thing is, I could pull the parts for a small cnc together for around $500 locally, but I would not really have a use for such a small apparatus. To be honest, I have not purchased anything, exactly so I don1t mess up anything. I want to plan the machine to the last screw in CAD before I start anything, but of course to do that I need to know how to calculate the torque I need. Then again if the laser route is viable, There wouldn't really be a need for high torque, I could probably make a multi-head cnc with torque just enough for woodworking, which would be more then enough to move a laser...

Can anyone hook me up with a link to one of these 'DIY' laser cutters?

Thanks!

[Fish4Fun]

Rattanee,

The difference between metal work and wood work is night and day. I am a newbie, not too different than you. I have had a metal lathe and mill for a decade or two, but both are manual, and I cannot even begin to express the difference between metal work and wood work wrt the machine that cuts the parts. Wanting to work with metal does not mean you can build the tools yourself, or "cheaply". Everything you think you might know about physics and engineering gets challenged when you move from wood to metal wrt a cutting machine.

The torque on the motors is fairly trivial to calculate, and at the end of the day is reduced to available motors and drivers. The basic formula is F=MA. The Forces are a combination of sliding friction, overcoming static and dynamic intertia + spindle RPM, feed rate and tooth load; Acceleration is how fast you want any given axis to change it's rate of speed; and the Mass is the total weight of the Axis in motion; at the end of the day take all of that math and double the result and then look for motors with torque ratings in that range, likely 800ozin to 1200ozin. I don't know that you will find servos a more expensive solution in the torque range you will need to push your gantry around. When you move from the 3A NEMA 23 steppers to 8A+ Nema 34 steppers the requisite drivers and motors take a big jump in price, and servos become economically similar. I do not think the cost of the motors and drivers are going to be what give you "sticker shock". The ball screws and nuts you will require could easily more than your motors/drivers for any given axis.

As far as "speed" goes, you should not underestimate the "costs" of a slow machine in both quality of finish and time to complete a job. I would STRONGLY advise you to download the trial version of the CAM program you intend to use and simulate some parts you think you might want to cut. A good CAM program will give you an estimated cut time at given feed rates. My little desktop machine (8"x11" work envelope) tops out @ ~60ipm and I have already tasked it with jobs that have had over two hour run times. If you scale that up from 88 square inches to 4600 square inches, the implied run-time would exceed two days at 60ipm, but could drop to 4 hours @ 600ipm. But feed rate is not just about how fast a job gets done, it is about tooth load, spindle RPM and finish quality. If the spindle RPM is too high for the feed rate you can potentially "burn" the work and destroy the tool. In all materials you want to achieve "chips" not "dust", and this is best achieved by having a wide range of feed rates and spindle speeds.

What do you mean by rotating nut design?

A rotating nut design is exactly what it sounds like: "A fixed lead screw and a rotating nut." A rotating nut solves the problem of lead screw "whipping" by eliminating rotation of the screw itself and moving the rotation to the nut. A rotating nut design has its own problems and costs, but in larger format machines the trade-offs can be worth the cost/trouble; at the very least, it is an approach to consider in the design phase.

I know you do not think you have any use for a desktop size machine, but I think the experience you would gain from such a build would serve you very well in your larger format build. Your large machine, as you describe it, could easily cost $15k+, and a simple error in any single design feature could cost you considerably in $$, time and/or performance. Building a small format machine first will give you a tremendous amount of experience that could help prevent costly mistakes in your large format machine. You might also find having a small format CNC machine very helpful in the construction of your larger format machine even if you build a completely different type of machine (MDF, round rails, bushings, steppers, fixed gantry Versus Aluminum, supported linear rails, BB Trucks, servos, moving gantry). I spent 20 years "dreaming, designing and using manual machines", and I thought I had a very firm grasp on what I was getting into, but I was quite wrong. I am very glad I have made my mistakes on a small format machine where a bit of super-glue or a trip to the hardware store gets me back up and running.

One final word of advice, begin researching and learning the software now, don't wait until your machine is near completion to start on the software. CAD/CAM/Controller software/firmware can take a while to get a handle on. CAD is the easy part, and you likely already have that down-pat, but porting your CAD to CAM to get efficient g-code can be frustrating. I have found the Vectric suite of products to be easiest for me with what I do, but there are many other products that have strengths in other areas. I would suggest downloading the trial versions of several products and evaluating them. The software can be a non-trivial cost factor and can present a challenging learning curve. The two most popular PC controller applications are Mach3 and EMC2; there is also an Arduino based controller. Learning about the software/firmware as you build your machine will ensure you make a smooth transition from build to use.

Again, Good Luck with your build! (If you plan on documenting your build here @ the zone, I might suggest you do so in the woodworking machines forum, you are much more likely to get help there with the type of machine you are building.)

Fish

[restless1]

If you learn on a small machine you break small parts if you learn on a big machine you break big parts
Being silly here but hey I am so glad I made all my big mistakes on my 12" by 18" machine....
From that I learned the basics and so I can at least keep my damages down from simple easily corrected ones to oh crap that just cost me this months mortgage payment..
Starting small should never be underestimated!

[missing-mojave]

Hi Rattanee,

I just built a machine similar to what you're talking about (2000mm x 1400mm x 900mm 3 -1/2axis, mach 3, stepper motors, Kress spindle, x gantry axis chain drive, y&z axes lead screws). I need to take some pics and send them to you. So far I only use it for wood, foam, plastic, mdf, and low density pattern board but it works fine. I think alu would be possible with this m/c if you went slowly enough. I'll try to send you those pics tomorrow sometime.


Ciao,

M

[jharvey407]

A moving gantry is going to present real problems with cutting metal. Building from aluminum can be problematic if you are wanting to cut metal. Using a router is going to cause real problems cutting metal. A machine this size is going to be very difficult to DIY to cut metal, even aluminum.

I'm going to disagree with the statement.

I am just finishing an all aluminum moving gantry router with a PC 690 router as a spindle. It has a cutting area of 36" x 52" and I have had no problems cutting metal.

I used the previous version of this machine, while it still had a wood table, to cut the gantry sides for the current version. I cut 1/2" 6061 at 50 ipm with an amana 0-flute bit running at 27,000 rpm. I took 1/8" (0.125") passes and got a great finish.

Here is a shot of the machine while it still had the wooden table and chain drive.



The new version has an aluminum table and R&P drive. I am getting rapids of 750 ipm and have been cutting wood, plastic and metal.

Good luck with you design and build.

[missing-mojave]

Rattanee,

Here is a very basic method for estimating cutter forces. It's from a machine design class I took in engineering school about a hundred years ago. I don't have data for wood, but I'm sure you could find it on the web.

Also I included a spreadsheet that covers some Mach3 calcs you have to do when setting up your stepper motors...there are probably a few mistakes in there so consider it reference only. It also estimates the thrust you can expect given stepper motor torque, sprocket reductions, and thread pitches etc.

Hope this helps answer your questions.