Machinist question - CNC/CAM/G-code

A good friend runs a busy machine shop in our small town. They're an entirely manual shop, and I've become their trusted resource for any PC problems(which they only use to receive prints via email...a relatively new development!). So naturally, now that they're looking to purchase a used CNC lathe, he's asked me to help him learn how to use it.

A quick Google search lead me to G-code...it makes sense to me(at least on a basic level). He mentioned the machine he's looking at has a Yasnac controller on it, and from the sound of it most of these are actually controlled more from CAM software than actual G-code programming, correct?

If so, where should I point him for the CAM software? Are we talking Solidworks, etc.? I've seen you guys discuss how crazy-expensive it gets, are there any GRM-priced solutions?

Thanks again for helping me cram more stuff into my tiny brain!

Wow, this is a pretty fun opportunity for you to learn some new skills for sure!

Disclaimer: I'm the Multiaxis Product Manager for MasterCAM, the largest (by installed seats) CAM programming package [Clarkson]IN THE WORLD.[/Clarkson] I've used a lot of other CAM systems, though, so I'm not just a corporate shill, I'm also an enthusiast corporate shill :)

1st step in this thing: Have you done any jobs with a manual lathe, and/or is someone going to stop by with the machine and teach you guys the basics of setting it up? That's the "hardest" thing about running these beasts, you have to be on top of your game about managing your tools, where they're touched off, what type of insert, etc. This is all "paperwork," so to speak, but it's what the computer running the show (the YASNAC Control) uses to figure how where the tool tip is and where your part zero is so it doesn't plow through the part. Learning how to set these up correctly is much more important than programming it at first!

Now, you've gotten that out of the way. There's a lot of different options for programming, let's start with free: Notepad or the editor on the control. The control may also have what's called Conversational programming where you can enter things as a conversation, "choose to insert a rough line, from X0,Z0 to X3,Z0," etc.

CAM is helpful, though, because you can actually see the part, you can see the way the machine will move around the part as it cuts and you can simulate the material being removed. Some of the more intelligent CAM systems will even allow you to control your cuts so you don't waste time on air or other things.

There's a couple of cheap programs, but you have to watch out for getting the output correct on your control. See, here's the big rub between CAD and CAM. With CAD, you only have to be able to figure out how to generate the geometry you want to cut, with CAM, you need to actually convert that into the G-Code that the machine can run! That's accomplished through a Post Processor. Your CAM Software needs to have one specifically for your control, and possibly specifically for your machine. One thing that a lot of our new customers tell us is that they weren't getting support with their previous product, so even if the thing could do what they needed, they couldn't figure out how to get it cut on the machine, so make sure whoever you partner up with is good to actually help you when you need it.

Any chance you can post the kinds of things they want to machine? I can give you the skinny on a few different strengths and weaknesses that they're going to see in some software. If it's simple lathe profiling programs, and you can get a good post, you can't go too wrong with any of them. If there's going to be lots of roughing or tight tolerances, some programs can shine with certain things, like ours is very, very strong in the lathe material removal area and we also have some toolpath options specifically for lower-torque spindles (like making shallow cuts further away from center, where the tool acts as a big lever-arm against the spindle).

Also, any chance you can tell us what options the machine has? That could also help direct you to a good software to choose if it has, for example, a programmable sub-spindle with upper and lower turret, that's a lot more complicated than a simple 2 axis lathe with a single turret.

I use GibbsCam. Does everything and more. Producing code for both our lathe and mill with it.

CAM software is nice but not a necessity. I used to program long hand, line by line with G and M codes. There can be some variation between controls as to what G code has what function.

If we shut down again there's a chance I would be able to head up there and scope it out. It actually sounds like fun.

In reply to modernbeat:

We have GibbsCam on our CNC machines too, it's nice.

G00 rapid traverse

G01 incremental feed

G02 clockwise arc

G03 counter-clockwise arc

Etc., some lists go up to the 300's, I have a couple of folders full of them.

I assume it would be a horizontal 2 axis (4 if it has an end turret) lathe so you'd have X and Z coordinates along with the G and M codes.

Thank you gentlemen!

In reply to WonkoTheSane:

That's very encouraging that CAM is not a definite requirement going in. I have - zero - machinist experience, but these guys are pros and have been in business for probably 30-years, so from a manual operation standpoint I presume they know their stuff pretty well.

I don't know anything about the machine(and honestly, if he told me it wouldn't mean much anyway), but I do know that presently a large portion of their work is with stainless steel, and most of what I've seen in the shop are large rollers ~4'-6' long by 8"-10" diameter, with conical ends. I asked about them on my last visit(after seeing a well-used broken one on the shop floor) and he said they're rollers for a chain belt used in a foundry. However I don't know if they're looking to update so they can expand their business and compete in other areas, or if they're just looking to increase productivity.

In reply to: The_Jed: I really appreciate the offer and may very well take you up on it! Regarding lines of code, I program in SQL all day long, so program length or number of commands doesn't bother me much. From Googling/researching tonight it seems like getting the left-hand vs. right-hand orientation thing figured out in my head would be one thing, then as Wonko mentioned zeroing everything out.

These guys are realists and don't jump into anything hastily, so I know there would be no expectation of "Hey, can you get this setup tonight so we can use it in production tomorrow?" or anything.

I'll ask for more details tomorrow and post back with any updates!

Start by programming some simple parts by hand. This will help you learn the basics of g-code so you can debug the code that' gets spit out by your post processor. I put my way through engineering school running CNC mills and lathes at night.we used gibbs cam for mill programming but most lathes programs were pecked out by hand at the controller. 2 axis work is usually ridiculously easy to program. A little trig and maybe autocad to figure out some geometry can go a long way. Learning canned rough and finish cycles will save some serious time.

For a budget CAM package check out Bob cad/cam. It is really cheap, but is a bit quirky and not as polished as some of the more expensive options. One of their reps hangs out at the practical machinist message board and offers pretty good support.

The latest CAM software I've been using is Siemens NX. You don't even want to know what that costs!

I definitely agree with 2002maniac, do a few programs by hand just to get a feel for how the machine reacts to various codes.

Both Gibbs and Bob CAM packages are fine entry level stuff, NX is really sweet deal, but $$$! I go up against them a lot, since I normally swim in the deeper end of the pool.

When it does come time to start shopping CAM packages, what I always recommend to people that they get the reps in from whomever they're looking at, and ask for them to program one of your parts in front of you and run it on the machine. Remember that this is a business, so saving $2000 on the upfront cost of your CAM package doesn't make economic sense if there's another package out there that would save 200 hours of machine time a year (figure ~$50ish/hr of machine time is a $10,000/yr savings), or if it's easier to use, has less intrusive quirks to your parts, etc.

4-6' long and 8-10" of a cone seems like a lot of metal to remove! Should be lots of fun :) If you do go down the CAM route, look into something that can do a Dynamic style material removal (our word), also called Adaptive by some other packages. Last I knew, we were the only ones doing it on lathe, but I know that some other packages started working on their own lathe toolpaths after they saw ours.

This is a lot like jumping in to learn EFI, the most important thing is to figure out a goal (I want to figure out injector pulsewidth!), so you can have an are to focus on until you're comfortable enough to move to the next step. As an SQL guy, you're going to think G & M code is laughably simple. The only problem is that each control type uses their own "flavor" of G & M codes, think MYSQL vs. ORACLE, etc.

Tooling engineer/designer/post process programmer here... I use mostly MasterCAM here when I can, but because we're a contractor I see a lot of wierd stuff too... G-code is your friend, and where I'd recommend anyone to start, get a good working knowledge of, and become competent at using it to make most any 2 axis part. Think of it as the core of all other peripheral assistance programs. Everything else helps, but it all comes down to the code that the machine reads and interprets. Get the hang of 2 axis, and add the 3rd.

Concept part -> Solidworks/etc designed and identified part-> CAM for machining protocol -> post process/part efficiency programs/other newfangled software to help operators and companies from wasting stuff-> G-code-> machined part... That's if all goes well. The best part about the above process though is that you only need the furthest downstream. If you have G-code, you can make a part. Might not be what you want or designed, but it'll be a cut part. You don't need any post process or other software if you're not really utilizing it for it's desired benefit- saving time, money, and material.

I'm stopping by tonight after work, so I'll get more details then & post back here with updates. Thanks everyone!

I programmed a LOT of stuff just using a word processor and straight G-Code. Also - knowing how to do that can let you do stuff that isn't in the rulebook. For instance, I doubt there is a cam package that will let you turn parts on a mill pretending it's a lathe. But I programmed it manually to do just that...

I haven't messed with it in a while, I was hired as a freelance IT/CAD guy for a company that eventually turned into a real job. I had CAD experience, some machine shop experience (grandad owned/ran a screw machine shop, WWII era New Brittan screw machines, I was child labor in), and I came recommended from their Airgas rep, who is a good friend from college. I taught myself, from set-up, programing, nesting, CAM, trouble shooting, etc. It was trial by fire, learn by doing, I persevered, and the shop eventually started making money.

I installed/set-up a one-off 5x12', CNC, downdraft, Plasma table, which used an ancient Burny 2.8 controller, and a High Frequency Hypertherm machine torch (liquid cooled with bottled Oxygen & Nitrogen "air" source). I wasted some material, and a lot of consumables on the torch early on, but in a few months time, there wasn't another person that could program the machine, and apparently I wasn't a good teacher.

I will say this, BobCam was mentioned earlier, it was okay, and the folks at BobCAD did help write a custom "post processor" for our ancient Burny controller, it worked, but was far from ideal, I often had to tweak things. BobNest did a half-assed job of setting up the most efficient use of materials, and BobCAD is/was (we had V. 22 I think) utterly laughable as a CAD program, think a fancy version of MS Paint that could export a .dxf, I'd avoid it like the plague. I'm Jaded though, I came from formal training in Autocad, to BobCad and eventually wound up with Solidworks, so BobCad was like comparing a Cavalier to a Corvette.

All in all I'm glad I did it, I learned A LOT, but I don't miss it. Someday I intend to have my own CNC equipment, and hopefully remember some of the stuff I learned, someday.

Update - the machine they're looking at is a Mori Seiki TL-5A2000 lathe, and the controller is a Yasnac 2000G II.

The price is low enough that their thought Is they can buy this to learn/practice CNC on without fear of trashing a newer, more expensive, machine.

Are they on the right track?

Mori Sieki makes a good lathe. We have one at work, it's the go to CNC lathe, the other one is a small HAAS and it can't take much more than .050" DOC.

Then we have a FADAL VMC and tons of manual lathes, mills, grinders, presses and gear hobs.

Does the lathe they're buying come with a steady rest attachment or live tooling?

In reply to The_Jed:

No idea...and no idea what either of those are. Off to Google...

Edit: It has a steady rest attachment, but I don't think it has live tooling.

Double-edit: It has an 8-position turret, does that mean it has live tooling?

They would need a steady rest for long workpieces, as it's a support(usually a triple roller arrangement that clamps over the part to support it somewhere along its length). Another necessity that some(including the shop I work at) don't think is needed is a live center on a tailstock. If a part is in a typical CNC lathe, the software programs for a specific surface speed at the cutter tip, called sfm. As the part diameter decreases, especially so when you are cutting it off or tapering an end to a point, the machine rpms increase to keep the surface a speed the same. On a manual lathe, the operator usually sets an rpm they feel confident the machine can turn and keep the workpiece from whipping around and possibly breaking free of the chuck. The first time with a large workpiece in a CNC lathe will be a brown pants moment as those revs shoot up while its cutting. Without adequate support on a piece the size you described, its quite possible to do damage to the workpiece, the tools, and even the spindle bearings as the part deforms..

Having just looked up that model #, another feature they may want to find out about, if they do complex parts(more often a concern with live tooling and 3d operations) is how well any software they install supports DNC control. Generally if I see a machine with a CRT monitor, it has some age on it, and the memory in the machine may not be up to the quantity of code output by a higher efficiency CAD/CAM package. If the machine can only store limited lines of code, you have to drip feed the program from the PC where it was written.

Live tooling is tooling that is rotating on its own, usually driven by a sub spindle in the tool carousel. Just having places for eight different cutters does not mean you have live tooling. Live tooling functions like a miniature mill within the lathe. The main spindle can orient to a certain position, the the live tooling can drill holes or mill pockets away from the centerline of the part. You would need live tooling to cut a hex for a wrench partway along a part, or to cut pockets on the parts surface, like a slotmag wheel for example..

In reply to travellering:

Ah, ok. Makes sense, thanks!

He beat me to it.

It just basically means there are milling tools that can be used by the lathe. It opens up all kinds of possibilities!

My all time favorite machine was a Toshulin SKIQ-12 with a Fanuc 0-T control. I called it a 3 1/2 axis lathe. Machining capabilities with that old beast were limited only by the table size (48" or 49", it's been a few years) and the machinist's imagination.

It was a VTL with live tooling and if you fed it a G113 you could program it like a VMC; It used it's C axis (the table/spindle) and X axis (lateral feed, right to left, across the table) to simulate a Y axis (lateral feed fore to aft across the table), hence 3 1/2 axes.

Man, I miss that machine! It had it's flaws but it was lots of fun!

Travellering seemed to cover everything really well, the only thing I'd say to your friend is to make sure you get a really thorough PPI on that machine.. I haven't seen one of those in the wild in a while, because they're getting older, so you have to be careful of buying a lemon. I've seen lots of people in this industry who got a "smoking deal!" on a mill or lathe for only $5000, then they ended up dumping 25k into it over the next 5 years, when they could have just bought something that didn't have tolerance or reliability problems and would have actually turned a profit.

It's always hard to remember that work is different than home and it has to be reliable and dependable at work, time is certainly not "free" like it is at home. At work, every minute lost is actually two minutes, you've lost the one minute that you're not done, and you're now one minute behind on the next part...

Definitely that^.

All it takes is a few ham-fisted operators to wreck the piss out of a machine and basically ruin it. Won't hold size, cuts tons of taper, won't repeat, alarms out all of the time, etc...

Especially if he's going to hog stainless with it.

Yup, where are you located? I can recommend a couple independent service companies that will save you a ton of $$ over your local Mori Dealer, and that's if they even want to look at it.

Mori Seki equipment from the 70s/80s is high quality stuff usually, but as others said- you have no idea how it was operated or used/abused. Check the ways and make sure the lube system is functional and clean first and foremost. Lots of old Iron being sold for cheap hasn't seen waylube since the old Bush was in office. It's a very common problem.

As I type this, I'm staring at a customer's Froriep 25KE VTL. 24' table, 36' Vertical travel. Made in 1942. My company retrofitted and adapted a full working 5 axis Fanuc Alpha I controls to it, as well as a full mechanical refresh and upgrade. It took a full year.