[createthis]

Howdy folks,

New guy on the forum. Software developer by trade. Bit of an artist at heart. I have a 3d printer (a Printrbot, so I'm familiar with Kit style machines), and I'm ok with CAD. I have zero CAM experience, and beyond using basic hand tools and some power woodworking tools, zero machining experience. Still cursing my high school for not allowing me to take CAD and machine shop, back in 98. I opted for CAD at the time. Probably a smart move, but I've spent the intervening years trying to get my ideas out of software and into the real world.

I want to manufacture some of my models in high quality materials, like aluminum, stainless steel, and even blocks of solid plastic. Desktop 3d printers (for now) can only take me so far. Great for prototyping, but I want to be able to make one-offs in other materials. Here's a great example of a part I'd like to make in aluminum or stainless: Impala Knob Shapeways wants $100 for that knob in stainless steel. That just seems too high to me. I see a 12" by 2" square bar of T-304 for $88 at Online Metals. I could make half a dozen (maybe more) knobs for less than the price of a single Shapeways knob, assuming my tooling didn't break.

I'd also like to cut custom gears. I print plastic gears currently, but the tolerances are poor: DIY Servo with Plastic Gears
Oh, and I want to cut cool things like Lithophanes from materials like corian, natural stone, metals, glass, etc. Click for Lithophane example.
I also just want to get some CAM experience, because I don't think CNC is going away anytime soon. Sometimes I get really frustrated with software development. Maybe I'd enjoy being a CNC operator more. Who knows.

So here's the question...

Do I start small with $2602 for an X2 mini mill? I did some research today, and here's the Mini Mill BOM I came up with. I know the X2 is going to be excruciatingly slow milling aluminum and stainless. I'll also probably be annoyed by the speed of tool changes. If I decide I need production runs of a part, I'll have to farm out the work or buy a bigger machine. But at least it would give me CAM time, and let me get to know the basics of CNC and gain some confidence without getting into a significant debt.

Or do I go for broke and find someone willing to let me finance a $15-20k Tormach 1100? The BOM is probably 5-7k too low, because it doesn't include all shipping costs, and I'll probably want the ATC.

I bought a car this year in that price range, so I think financing is possible, if a bit daunting. The unknown for me is whether or not I'll be able to make the Tormach pay for itself and earn it's keep. I've read a lot of threads in the practical machinist forum where people were paying for the machine in the first month or two with contract work. Believe me, I'd be thrilled to be making enough to get it out of the basement and rent some real shop space, but I don't think it's reasonable to expect the machine to make ANY money right away, considering I have zero CAM experience and zero industry contacts. I don't even know if there is a shortage or overage of CNC work in my area. I'm just completely blind to the industry.

Worst case, I decide it's not the right tool for the job. I'd hate to make the investment and not use it.

I'm curious what other people think. Is the typical path, mini mill, then Tormach or a professional machine? Is there value in that approach? Or am I just the poor man paying twice? Also, where would I go to sell one of these machines if I decided CNC wasn't for me? eBay?

[doorknob]

Where are you located? Maybe there is a TechShop location nearby where you could use their machines instead of buying your own to get started (TechShop is America's 1st Nationwide Open-Access Public Workshop -- TechShop Locations). Or maybe there is a nearby Makerspace or Hackerspace that has machines that you could use.

Although you could learn a lot by using an X2, it would probably be a good idea for you to get some hands-on machining instruction (for example, a couple of Machine Tool Technology courses at a local community college or vocational school). Understanding manual machining will help prepare you for CNC machining.

Buying a Tormach without knowing in advance whether you have a market for the parts that you want to manufacture is a risky proposition. Have you looked at some alternatives to Shapeways (such as CNC Machine Shop | Custom Waterjet, Plasma, Laser Cutting | eMachineShop.com)? Setup charges are always going to be a problem for small production runs at a shop like that. Maybe consider using the RFQ section of this site to get a quote on machining a part for you, and see if you can sell them, before deciding to try in-house production.

If you do decide to jump into the deep end anyways and buy a Tormach, and then decide that it's not for you, you could probably easily sell it via the cnczone classifieds - you don't see lightly-used Tormachs for sale that often, but people are certainly looking for them.

[ger21]

I also just want to get some CAM experience,

Be aware that learning a CAM program does not make you a machinist. CAM is basically about the user setting some parameters, and letting the software do all the work. I would say that the CAM part is typically much easier than the CAD part, and shouldn't be an issue. What you really need to learn is the different techniques of machining different materials, which comes with lots of trial and error. Lots and lots of reading can really help you out here.
Vectric software can create g-code for Lithophanes, and it's really quite simple to do, once you do it the first time.

The unknown for me is whether or not I'll be able to make the Tormach pay for itself and earn it's keep. I've read a lot of threads in the practical machinist forum where people were paying for the machine in the first month or two with contract work.

CNC machines do not print money. They are just tools that can increase productivity over manual machines. In some cases, they can do things that are nearly impossible with manual machines.
If you don't already have products and a market for them, then a CNC isn't going to make you any money. And even if you do have a product that will sell, can you actually sell it for enough money and at enough volume to actually make any money?

Here's a great example of a part I'd like to make in aluminum or stainless: Impala Knob Shapeways wants $100 for that knob in stainless steel. That just seems too high to me. I see a 12" by 2" square bar of T-304 for $88 at Online Metals. I could make half a dozen (maybe more) knobs for less than the price of a single Shapeways knob, assuming my tooling didn't break.

How much is your time worth? Stainless steel can be difficult to machine. regardless of whether or not your tooling breaks, you still need to by the tooling. And there's a good chance you'll need to fabricate at least one fixture to hold the parts while you machine them. I wouldn't be surprised if it cost more than $100 each, even if you did make 6. I would expect someone to expect to make at least $50/hour to make a one off part. At $100, that part would be a bargain.

[createthis]

Hey, thanks for the reply. No, we don't have a Techshop nearby. I live in TN. We just started a Makerspace this year in town, but all we have right now are a couple of drill presses and 3d printers. I'd say the fate of the Makerspace is as yet undetermined. We've got a huge custom built CNC wood router too, but I'm interested more in metal parts at the moment.

I've not tried eMachineShop. Maybe I'll try that out later.

I checked the classifieds section here. I didn't see any X2 machines, and the only Tormech was in Chicago.

I think I've decided the X2 is the way to go for now. This guy's setup (power draw bar, flood coolant, etc) seems ideal: https://www.youtube.com/watch?v=u-Cuk5Tdios
The price is right too. Too bad it's already sold and in Canada. If I could get my hands on a machine like that for a meager $3500, I'd be a super happy camper. As is, I guess I'm looking at $2600 for a CNC X2 without the power draw bar.

I know I'll waste a lot of time fiddling and upgrading the machine, but I don't think that's necessarily a bad thing. I like to be able to maintain my own equipment. You can't get that intimate with a machine without conflict.

[doorknob]

You can indeed learn a lot with an X2, and even perform short-run manufacturing (as long as you understand its limitations). There are cnczone members who have done exactly that (and later on they typically move to bigger, more capable machines, after proving that they can make parts and make money with the X2). I have an X2 that I have not yet converted, but once I do the conversion I have some parts in mind that I may manufacture with it. Some people will ask you whether your interest is in building the machine or building parts with it (in which case you may be better off with a turnkey system), but from your description it seems like your answer would be "both".

BTW, I found your (?) DIY Servo and createthis.com store and blog to be quite interesting - I passed along those links to a friend who would find them interesting as well (he has read the same DIY gear cutting book, is working on how to build bigger robots than the typical mini mobile ones, has a Makerbot, uses OpenSCAD, and is investigating ways to generate an income from 3D printing and related techniques).

[createthis]

Yes, I'm ok with both. My motto is "knowledge is power". I'll learn whatever I can about a subject.

Your friend sounds a lot like me. I know one other guy interested in larger format servos. He runs project biped: Project Biped

Have your friend contact me, or join the site and start a discussion, or just lurk and see what happens. The DIY Servo project is all very theoretical and research oriented at the moment, but I think it has a lot of potential to become something really interesting and relevant. One of the potential applications is DIY CNC servo motors. I haven't finished my research into higher voltage system yet though, so I can't really say if it's practical yet.

[hanermo]

Your questions are good and assumptions way off base.
An x2, or any manual hobby mill, will not "make" parts in quantity.

The thormac will, slowly, compared to a HAAS mill, for example. But potentially just as accurate or more than good enough.

Making (some) parts is not hard. With X2 or something else.
Making many parts profitably cheaper than anyone (jobshop) near is very very hard and needs heavy investment.

IF you choose to build your own manufacturing capability, you will need at least 2x the machine cost in tooling.
Independent of machine cost !
This has been the avg. for all machine shops, in general, for a long time.

At a minimum, you will spend thousands on tooling, mics gauges, calipers, taps, mills, thread mills, go-no gauges etc.

There are 2 different cases here:
-business (making bits or thingies)
-manufacturing small scale metal bits

Manufacturing small scale metal bits can be fun, and may be profitable.
It needs a lot of learning, easily acquired, and lots of "stuff", most of it quite cheap, in the 30-100$ range. But remember lots.

If you wanna make a business, and make some "small" money, go to all local jobshops with a sample (3d printed is fine) and some donuts/6 pack/whatever.
+Offer to pay several at least 50-100$ for their time, and ask them "how to make this type of thing cheaply".
You will find production costs are 1-10$/piece in any reasonable quantity (+tools/material/time).

NOW you can, or not, decide to make the stuff yourself, or not.
A local jobshop can make them for you, if you are clever, and deal with them right, for a few bucks.
Tell them "anytime, offshift", in 10-15 days, when you are free, etc..

[createthis]

First of all, thanks to everyone who is reading this and replying. I appreciate the feedback. It's helping my decision making process.

An x2, or any manual hobby mill, will not "make" parts in quantity.

I know an X2 will "make" (why are we air quoting that?) parts very very slowly. That was already understood.
The Tormech will make parts less slowly than an X2, but still slower than a Haas. That was also understood. So I don't think we have a problem with assumptions there.

Your questions are good and assumptions way off base.

Erm. I don't understand why you're saying that. I wasn't talking about mass producing anything. I know about drop forging, casting, stamping, etc. Most of these techniques are better than CNC for large scale manufacturing anyway. I'm mostly interested in something that will allow me to mill parts I design in house for prototyping purposes. If I make a few and there is demand, I'll be happy to either buy a Tormech, hire out production runs, or build my own drop forge (heh).

regardless of whether or not your tooling breaks, you still need to by the tooling. And there's a good chance you'll need to fabricate at least one fixture to hold the parts while you machine them. I wouldn't be surprised if it cost more than $100 each, even if you did make 6. I would expect someone to expect to make at least $50/hour to make a one off part. At $100, that part would be a bargain.

Ok, let's take the fixture out of the equation for a second. I admittedly know very little about fixture design (though I've seen them in use). Clearly there is a cost associated with fixtures, but let's pretend for a minute that they don't exist to simplify the discussion. Also, let's assume I have spent some time nickle and diming myself to death and I have a fairly complete tooling collection for the sort of work I intend to do on a regular basis. I've learned a couple of trades over the years. I know the small tools grow like weeds. I don't intend to buy them all at once. I expect them to grow naturally over time with need.

Ok, those assumptions out of the way...

Why would it cost $100 each to manufacture 6 pieces from an $80 block of metal? Are you assuming the tooling will break immediately after 6 pieces have been cut? Or are you lumping the total cost of tooling for the 6 parts into the total cost of the 6 parts? If the latter is the case, I don't think that makes sense. I'm not going to throw away my tooling after I make 6 parts. I'm going to keep it and use it for other parts later (maybe even more of the same part, but not necessarily). I think there must be some sort of depreciation to consider, as the tool wears with time. And certainly there will be a cost associated with R&D and actually making the CAM program, but that's my time so I only have to eat it, not pay it. Of course you break a few bits in the process. But at the end of the day you've got a reproducible way to manufacture the same six parts (albiet slowly) time and time again.

I admittedly don't know how much tooling costs or how often it breaks. Maybe I'll be thoroughly shocked by the breakage rate. But with flood cooling and care, I am assuming the tools last for more than one job. I guess I expect them to last four or five jobs after the bugs have been worked out of the CAM process. Is this a false assumption?

Going back to the original argument of mass production... this isn't mass production. It's small scale prototyping for a cottage/niche market. If demand exceeds production capability, you scale the operation up. Maybe this means outsourcing, maybe it means buying faster equipment. That's a decision for another day.

Also, if I'm building the part myself, I can sell the part for $100. I cut out the labor cost associated with the machine shop because I am the machine shop. I don't need to mark it up again.

Does this make sense? Tell me where my assumptions are wrong, now. I'm curious. It's valuable info.

[awerby]

Okay, so say you've spent $2602 on a retrofitted Chinese mill and $80 on a hunk of stainless steel. You've spent $9.95 on the STL file. You still have to buy some tools. A vise, to start with - that's $100 Some parallels to go with it ($30). A small bandsaw to cut the steel into manageable sized pieces ($250). Remember to get some blades for it ($30 ea). You've got the file, so you don't need a CAD program, but you still have to buy CAM software, which isn't free. Say you get a cheap package, though, that only costs $300. Now you're up around $3300, but you're just getting started.

That knob, it turns out, isn't very easy to make on a CNC machine. The first thing you need to do is cut the half-moon hole that the flattened shaft fits into. How sharp does that inside corner need to be? If you're cutting it with a round tool, the radius of the tool determines that. So say you're looking at a tool that's 1/64" in diameter. How deep does that hole need to be? Half an inch? Deeper?

So you get online and start to look for tools like that. Okay, great, you found one - but wait; that's a 2-flute cutter made for aluminum and softer materials. You need a 4-flute model made for steel and harder materials. You go through a lot of sites. Here's one: 1/64" 4F LONG SQ Miniature Solid Carbide End Mill $44 and change, plus shipping. But wait - the length of cut is only 1/8". That's not going to work. Back to Google. If you value your time at all, the meter's ticking.

You start calling tool suppliers. They look in their computers and tell you there's no such tool as a 1/64" 4-flute endmill with a 1/2" LOC (length of cut) or at least they've never heard of it. But if you really want one, you can have them made custom - for $100 each. Better buy a few while you're at it, since if it breaks, you'll have to wait another few weeks to get another one. So fine, you're out another $400, but you finally get that precious little packet of tools. You open it. Wow, those things sure are skinny! But you stick it into the collet of your X2 and start trying to zero the axis by touching off on your part. Jog down a little, a little more - then "Plink". That's the sound of your new endmill suddenly getting a lot shorter. Okay, try again, more carefully this time. Say you succeed this time, and have an accurate idea of where the tip of the tool is in relationship to your piece of metal.

So you start the spindle, press the start button, the tool starts to plunge into the material, and "Plink" - that sound again. Turns out your cheap CAM program doesn't have a "ramping" feature that lets the tool slide into the metal at an angle, it just plunges straight in. Oh well, live and learn. You get some advice from a machinist, and drill a pilot hole for the tool to start in, so the plunge won't kill it. You've looked up the feeds and speeds on a chart, and although it didn't talk about 1/64" cutters, it seems that if you go half as fast as a 1/32" cutter you should be safe, right? Of course, the chart told you that with a tool that size you want to be spinning at 57,000 RPM, but since the X2 tops out at 2500 RPM, you figure you'll just feed slower, down to .25 ipm. This means that just making this one hole is going to take all night, but what the heck...

You start it up again, the coolant starts flooding, the tool starts cutting, tiny chips start forming, then "Plink" - the tool breaks. What now? You're starting to get frustrated, but you only have one of these tools left, so you go back to doing research. Is it the mill? Maybe. The expected TIR (runout) on the X2 is .02mm: not bad, but it's a considerable proportion of the diameter of your tiny fragile tool. Normally, the smaller the tool you use, the faster it has to spin, but your spindle is pretty slow. Also, it's not the most rigid mill in the world, so that might be a factor too.

Or is it the material? You'd heard that 304 stainless was good stuff, but maybe you were talking to welders - machinists don't like it as much, preferring 303. But reading further, you hear that any stainless steel has a problem with work-hardening - if it isn't cut with a certain amount of force, enough to get through the "skin" of material hardened from previous cutting operations, the tool can't get a bite but just tends to rub, further hardening it.

Or is it the tool? You're starting to get an idea why this isn't a common tool configuration. It's just a hair-thin piece of brittle carbide, sticking way out in space, that has grooves to further weaken it, up against a really obdurate material. That Shapeways deal isn't looking too bad now...

I suppose I could go on through another attempt, but I don't want to hear your last tool break. If you really want to make these knobs, machine a piece of graphite that would fit that half-moon hole perfectly, take it to an EDM shop, and burn it into your stainless steel blanks. Of course, machining graphite has issues of its own, like conductive dust that seeks and destroys any electronic equipment in its vicinity, but we can talk about that another time...

Andrew Werby
www.computersculpture.com

[createthis]

Very dramatic. I'm not sure if you're trying to convince me I need 4 years of machinist classes or if you just want me to give up and go home.

That's not really how I go about doing things, but you paint a pretty scary picture.

Also, the $9 model is mine. I'm the author. That's my site.

[awerby]

I'm not trying to talk you out of anything (actually, I'm in the business of helping people get into this stuff, not discouraging them), but you asked what the potential problems would be in making a part like that, so I thought I'd explain some of them. CNC milling is useful for many things, but there are some that just don't work too well. Stainless steel is one of the most difficult materials to machine, and deep holes with sharp inside corners are among the most difficult tasks to attempt. The rest of that part isn't so bad, although you'd need to figure out how to hold onto it as you're cutting it, hence the discussion of fixtures above.

While 4 years of machinist classes would certainly be a good thing, it's perfectly possible to learn a lot about CNC machining at home, working through software manuals and doing some trial and error experimentation. But it's helpful to start with projects that are simpler, in materials that are less obdurate. If you really wanted to produce those knobs by CNC machining, I'm sure there's a way, but designing the part with the process in mind would be advisable. For instance, you might be able to make that D-shaped hole in two parts, with a plate that screws onto one side to create the flat.

Congratulations on the site, by the way. Obviously, the items on it were designed to be produced by 3D printers. Some of the things on it, like those Arduino holders, would be a lot more feasible to create on a CNC mill than that knob. Others, like the lotus or the lampshades, would be even more difficult. Once you get a little experience in machining, you'll be better able to tell the differences yourself.

Andrew Werby
www.computersculpture.com

[createthis]

I'm actually really interested in the problems associated with milling various parts. That's one of the reasons why I want to get my own machine. I want to understand the problems first hand. I'm not sure I can get that information from youtube and books. I appreciate the fact that the D section is difficult to machine. I'll be the first to admit that I have no idea what is difficult to machine and what isn't at this point. I just get a little frustrated reading through the account of how you (tongue in cheek, I'm sure), portrayed me handling the situation. If anything, I think I'd probably have the opposite problem. I tend to over analyze everything way before I actually pick up a tool and do anything. I'm one of those people who suffers from writer's block and I have trouble getting started. I guess that's why I'm here. I mean, I'm sure I'll break stuff, and some of it will be expensive, but I'm really a pretty cautious person.

So how would I handle the problem?

I'd just change the design. The purpose of that D section on the knob is to key the shaft so it turns as the knob is turned by hand. If it's difficult to cut the D section, I would change the design to be more CNC friendly way before I seriously considered buying a fragile $100 end mill. I think the easiest fix would be to just overshoot the corners. I don't think the corners are critical. It's an AC knob for pete's sake. The original part is cheap ABS plastic. It's not a rocket engine. It doesn't have to be exactly like the original. In fact, it's not, and that's sort of the point. It's a custom part from my brain to reality.

I'm not saying I wouldn't buy and break tiny delicate end mills. In particular, I think I'd be really tempted to do something incredibly stupid, like build my own bearings for no other reason than because I don't like everybody else's:

Knifemaking Tuesdays Week 68 Bearing Cages - YouTube

( I love Grimsmo's videos. I think it's totally awesome that he pulled this off, despite the fact that it has questionable practical value. I could totally see myself doing the same thing. )

Anyway, back to CNC research land...

Sigh. I've been looking at Lathes lately. That Tormech CNC Lathe that has forever been in R&D looks pretty awesome. I'm not sure I need a lathe, but I know I'd like to have one. The "serious" machinists all seem to scoff at the little 7x12 hobby lathes, instead favoring multi-ton beasts from the 30s. I'm sure all that cast iron is great for accuracy. I just don't want it in my garage. I'd probably only ever use a lathe to prep blanks for the mill, or maybe cut my own end mills? But... who knows... Research continues...