New to CNC, Need Help/Opinions

[zild1221]

Hi everyone,

I introduced myself in another thread and asked a few questions. No one has bothered to answer there though. Hopefully someone around here can help me out.

Like the title says, I am new to CNC. I do a lot of DIY projects, and I mean a lot. I work on RC Helicopters, make parts for my DIY electric drum sets, and make a ton of hardware for sim racing. Usually I have to use less than ideal materials and rig something together to make it work properly. I have gotten a quote to have a few parts made at local CNC shops and it is outrageous. This has persuaded me to take the leap into doing CNC work myself. I have 3+ years of CAD experience for my schooling, so I have AutoCAD to do the drawings on.

I figure for you guys to accurately help me decide what I need/ what I should do, you need to know what I am hoping to be able to acheive. Materials wise, I would mostly be working with plastics. HDPE and acrylic. Also, I plan to work on a wood from time to time, as well as thin aluminum (1/4" or less). Ideal actual workable area for me would be 20"x20"x5". I could settle for 16"x16"x4" however.

Now, I am stuck here. I was first considering going the DIY route and building my own CNC router. However, upon further reading of these forums, some people are saying that you do not save much money DIY. Then, other posts, people are claiming that all the pre-made systems like the "FireBall V90", Zen, and the alike are overpriced outrageously. This is where I am torn.

If I were to go the DIY route, I know the basics of what I would need. However, I know nothing of specifics on how each part will affect the outcome of the machine. For example, how much of a difference X strength motor will really make verse XX strength motor for my machine.

I am in college currently so I am on a tight budget. I am looking to go no more than $1500, but even that is a little high. There would really really have to be an advantage of spending that much over a lesser amount to justify it.

Now, some more info that may help you guys out in helping me out. For one, my family owns a hardware store. This means that I can get any hardware at cost. (Fasteners of all types, power tools, pretty much anything you can name) Right there, if I go DIY, that should save me some money I would assume. Another thing is, I do not really have that much access to any precision tools to fabricate parts for this. I can make simple cuts and do simple drilling here and there. I have worked with materials like HDPE before for simple things.

Anyway, hopefully you guys can help me go from here. I am really eager to get something going here. Either ordering a prebuilt machine, or building my own. Also, I am not afraid of preowned parts or even whole machines as long as they are in decent condition. If there is anything else you need to know to most effectively help me out, let me know.

Thanks in advance!

[microcarve]

Well, the first question should always be....what do you
want to make?

(Yep, you mentioned that...

That'll decide how much $$$ will be needed to accomplish that
goal.

How much $$$ is available will determine how 'Good' of components,
materials, quality of parts, etc can be used. It matters.

But....there are a lot of ifs, ands, and buts....

IF...someone is very good with mechanical things then they can
cobble something together from complete junk. How well that works
is decided by how 'Good' someone is with the mechanical aspects, and
it also depends a *Huge* amount of what sort of tools they have
access to. And how well they know how to use them.

Making a design in CAD can often be a bit of a problem. There
are very few real-world materials that are anywhere near as flat,
round and square as the *Perfect* lines and elements that CAD will
draw out for you. So those parts need to be specially machined
very, very often to work correctly....or some type of mechanical
compensation accounted for to make corrections. Experience goes
a long way. But you can't get experience without trying. Failures
should be seen as finding out what Doesn't work....and knowing
what doesn't work can often be as...if not more...important as knowing
what 'Does' work.

(Knowing what does Not work can be directly attributed to the fact
that many of us still have fingers and hands to try something else with..

Getting back to your question....

If you want to actually *Use* a machine to Make Things, you're very likely
much better off to buy one.

It's not likely at all your "first design" will be good enough for you
and it's easy to get to a point of still working on design #2 years
later. You'll see tons of things that should'a-could'a been different
from design #1. You may end up with piles of hardware from Ebay that
are for design #3....when/if you get time.

The "bought" machine has been time tested and is known to work to the
standards it's built to. There are different levels. Basically, what may
be easily recognized as "hobby", "commercial" or "industrial"...just to
simplify for the sake of the discussion. Those levels should obviously
imply very different levels of *Cost* that are involved.

You may not know for sure if your DIY machine even works right with
nothing to make a comparison with. The bought-machine should come with
the very important and valuable....support you'll want...and need.

Outrageously overpriced?...

Well, someone has to make an initial design that works correctly...and
reliably enough to reproduce. And support it. That takes a lot of time
and money. Actually *Much* more than a lot. It's not easy...

DIY is a lot of fun, but it can get as addictive as any other thing.

The quality of whatever it is someone wants to end up making will depend
greatly on how good they are with the mechanics of things, how good
their tools are, their material choices, their design, their funds...($$$!).

A cnc machine implies a thing that's capable of doing very accurate
and precise work, so how accurate and precise it can do the work depends
on how well it's initially built.

If I needed to produce parts, and time was any factor at all, I'd join up
on the support groups the machine makers maintain and ask questions and
look at what others have done with that particular machine.

Many of us here love the challenges involved, though. Perfecting, adjusting,
tuning, upgrading. Those are all things that can take away from actually
'making' things.

Personally, I'm in the middle. I like making machines, but it's much more
sensible to just buy one if someone needs their parts/product anytime soon.

It's good to keep in mind, that it'll also take time to learn to *use* the
machine and make your parts to your desired standard. That's even more time
to consider.

If you need aluminum parts, You'll need a "better built" machine. You'll
need a good design and some Good tools if you go with DIY. Access to a lathe
and mill and the know-how to use them can't be underestimated. But, then again,
someone who's really good with mechanical things can make something that works
great, from trash bins.


DIY or Buy...?

Tough question....


John

[ssutton]

Well I went the total DIY route BUT I had access to a Bridgeport mill and a large gear head lathe and a host of other shop equipment. I designed the entire machine in 2D cad, not the best choice since I have now learned 3D cad. However, looking back on my build I think that I would look very hard at a turn-key machine if I were to need another CNC because my time investment has been huge!!!!! Hundreds of hours spent manufacturing precision parts. Without such tools at hand I would not even consider another DIY. Having said that, YOU may be much better at using what is available to you and building an awesome machine with very little investment in money/time. Every man needs to honestly access his/her skills before going the DIY route. What ever time that you think you will need to build your ideal machine take that time and triple it. Then decide if you would be better served to build it yourself.

"A determined soul with a monkey wrench can do more than a lazy man with an entire shop of tools"

I heard that somewhere, not my words but very true indeed.

Scott

[zild1221]

Well, after reading what you guys have to say, I AM STILL TORN! =]

With little ability to precisely fab parts to build my own, I am kind of forced to purchase a kit or a pre assembled machine.

Now, this is just for hobby really. It does not have to be ultra precise, I kind of just want to try something new.

Also, I should have mentioned this, but I have experience in Inventor and Solidworks. Not a ton, but enough to do basic drawings of things I need.

Really, the ideal thing for me would be a kit of just the assembly, and be able to buy the motors and electronics separate.

The closest thing I have found is the ZenBot. I have read pretty good things about it. For $799, you get the full assembly with motors. Then, I would have to purchase a stepper motor controller and power supply. Could I purchase this controller if I went this route? PBX-USB CNC Controller - PBX-USB CNC Controller

I would rather have the ability to USB than parallel port. I couldn't even find a PC with a parallel port if I wanted to.

I know it isn't the best quality, but I would have to imagine it would be sufficient for my needs? Also, I can cheaply get all of the rotary tools it supports, rather than expensive spindle motors and spindles.

Really, my plans are to learn how to use this to learn the ways. Eventually, I would love to design and fab my own with this.

I was looking at this company, m3-kit | Lumenlab , but starting at $899 for the basic package, then needing to buy motors, controller, and a spindle motor and spindle, that could easily cost much more.

I really don't know. There is just so much all at once. I can see it is difficult to get into this without knowing someone who already does this. Let me know what you guys think.

[Walky]

The problem with the USB controller is that with it you're stuck with one software, which might be a problem since you seem to be the DIY kind of guy, and might want to try different software.

Parallel port is a standard and as such is supported by most of the available software; you can get a pci parallel port quite cheap, or even an old computer as a dedicated controller pc (TurboCNC runs under DOS and runs smooth even on really old pcs like the first Pentiums). If you plan to use your machine with your everyday's computer, I really don't recommend it, since the slightest latency problem, resident program of unexpected popup window can make you lose your work progress, an expensive bit or a finger (and crash your machine if you don't have limit switches).

The Fireball isn't really overpriced, believe me; I'm trying to design an easy to build and reliably machine to sell as kits (or assembled) locally and I'm having a hard time trying to keep a low cost (it would actually be cheaper for me to just buy Fireballs or other kit and then assemble, align, test them and sell them, but even so shipping costs makes that impossible, and half the fun is about designing-building the machines anyway.

I started as you, trying to do almost everything by hand, and I've already spent more than what I'd have spent if I had gotten linear bearings and drivers from the very beginning.

Maybe a good approach would be to get linear rails and bearings, drivers, motors, and with the rails measurements (by datasheets and confirmed by hand measurement) make your plans and then send them to an online service like Ponoko to get your parts done, so most of the work left to do is assembly. I think you can get this done with less than $1000 depending on the machine's size.

There are way too many things to care about while building a cnc machine; for example, if you have the best driver/motors and linear rails mounted in a badly designed structure with a lot of flex and screw whipping, it might be a waste of time and money.

[zild1221]

^ Thanks for the info Walky. I did not know this is why there are not many USB controllers out there. I would much rather build a small system or recycle an old system with a parallel port. I love Linux distros so I will probably run that. From what I have seen, there are a few pieces of free open source software for Linux.

I think I have decided to go with one of these prebuilt kits type situations. Would you recommend the V90 over the ZenBot? Is there something else like this you would recommend? They both come out in the end to be about the same price.

If I get the hang of it, and like it, I will probably end up designing and fabbing (with the new cnc router) my own machine.

I am really really serious about doing this though. I have put in a good 48 hours of reading and searching just in the last week.

[adprinter]

Check out my build on the Machsupport Forum under Buying or Building a Routing table. Building or Buying a Wood routing table.. Beginnners guide..
Just scroll down to the MYCNC build topic, and click on the MYCNC.PDF file to download the file.
I should tell you, that I have since made MANY changes to the original build, which includes the addition of a 4th axis, and that to date, I have invested over $3500 in building the machine, buying the driver boards, motors, lead screws, couplings, energy chain for the cabling, cables, wire, nuts, bolts screws, washers, bearings, aluminum, steel, etc. with Software alone representing only about $250 of that figure.
But as others have stated here, the experience of designing, and building my own machine has no price tag. The one word of advice that I will say, is that no matter WHAT you plan to cut with your machine, you WILL WANT absolute precision in repeatability. So, before you decide to buy a kit, do a LOT of research on the kit you are considering. Buy Right, Buy ONCE, is good advice here. The advantage of building your own, is that you can incorporate adjustability into the design, which might not be available in a kit form.
Trust me on this, no matter how precise the build, there will be something that needs to be tweaked to get it just right. And if the machine is not designed to allow for this adjustability, you won't be happy with it. Practically everyone I know, who has built a CNC machine, has ideas and suggestions for an improved design. Often before they even complete construction of their first design.
I know that like me, you're as anxious to "Start Playing" as a kid in a candy store. But, like me, it sounds like you don't have the funds to waste. If you do decide to build your own, I should tell you that you will need access to basic tools which include: dial or digital calipers, drill press, metal cutting band saw, a dial or digital indicator, and taps. Also, MANY C Clamps and or bar Clamps will come in handy for the assembly, especially if you are working alone. This only covers the basics needed for the mechanical side of the build. For the electronics, you will need screwdrivers, allen keys, wire cutters, wire strippers, soldering iron, a volt/ohm meter, and an understanding of electrical schematic diagrams.
As I said, the experience gained in building your own has no price tag. And once it is complete, you will understand what to do when things go wrong. (Or at least you will have some idea of where to look). And the user forums on CNCZone, MachSupport, and others are where you will learn how to diagnose any problems you may encounter. Welcome to the crazy world of robotics, for that is basically what CNC actually is.....

[CNC Lurker]

The 2 main constraints that I've found when researching which plan to make my CNC was what is the budget, what size are the parts you want to cut, and what materials do you plan to cut. Though you can have grand plans, most people are limited by a budget.

You are on a budget, but what is your budget? $600, $1,000, or $1,500, etc.

The Zenbot is below $600, but the cutting area didn't meet the dimensions you wanted in the first post (I could settle for 16"x16"x4").

The Fireball V90 is $1334.80 when you throw in the 3-axis motor and other stuff.

If you have time on your hands, you could use that $1,330 and make a CNC out of 80/20 extrusion and have a bigger cutting area. Since you are familiar with Solidworks, take a look at the FLA-100 plans here: FineLineAutomation LINK

I substituted 1530 (1 1/2" x 3") 80/20 extrusion for the 1545 extrusion (1 1/2" x 4 1/2") and all extrusions including shipping to Texas was $370. I did buy the 3 axis electronics from that site, but I bought it on Black Thursday and it was around $500. You could go with a Probotix or Xylotex kit and save money there. I did buy all the mechanical parts from CNCRouterParts LINK

I have roughly a 24" x 34" cutting area, but you could shorten the length to get a 24" x 24" or smaller cutting area to save on extrusion cost. Remember, McMaster Carr sells the ACME threads in 3 ft and 6 ft lengths. I used the 3 ft on my Y axis, and cut the 6 ft one to get my Z and X axis. I bought the black oxide one 1/2-10 5-start, but you could get the plain steel and save some money there too. Besides, some have said that the black oxide with the delrin anti-backlush nuts on CNCRP isn't as smooth as using a plain steel, so that might be a plus. (see attached pic)

If I had to do it over again, would I buy a pre-built or make my own? Well, I haven't completed mine yet, but when I do, I believe it will be better than the CNC Shark that Rockler offers, and will be about half the price, AND I am using a full size router (Porter Cable 690). CNC Shark LINK

If you want to get cutting right away, buy a pre-made one, but if you want to save money, or get a a better machine for your money, build it yourself.

The tools I used were a Ryobi miter saw and a Ryobi 10" benchtop drill (both over 5 years old).

[zild1221]

Well, with some more thinking, I am going to go the premade route. Any other time I would build something like this myself. Really though, I would like to get going as quickly as possible and start learning. Then, if all goes well and I do not lose my hands while learning... =] , I will start planning my own build. Then, resell this one if need be.

So, for premade desktop CNC machines under $1500 that are ready to go, what is there. From what I have found, the ZenBot 1216 and the FireBall V90. Both of these come out closer to $1000 with everything I need. I would not buy the Zenbot control box. For now, I would just get the HobbyCNC board and a power supply.

Are there any others I am forgetting? The FineLineAutomation kit is out of my price range.

[zild1221]

Well, I am going crazy here! So many different possible routes.

Being a DIY'er, I cannot stand not building my own. Every night while I lay in bed I am coming up with possible designs in my mind and sketching them down on scrap paper.

I am DIY'ing! It is final. I would rather attempt and fail a whole bunch of times than not attempt.

I have decided I am going to use HDPE for the structure. Easy to work with and strong. Now to start looking at a whole lot of plans and go from there.

Are there any free plans you guys recommend over others?

Once I start designing, I will have a million questions for you guys.