[Cavnar]

I'm trying to decide on a CNC to build or buy.
I've been thinking about this for years, reading these posts for years, etc. And I'm not any closer to actually getting a CNC machine than I was before.

I NEED to be able to handle 2ft across. I have one project that I just can't do by hand, and it has to be 2ft by 6ft. I can work the 6ft in sections, but there's no way I can join panels or anything to build up the 2ft after carving.
I WANT to replace all of the other tools in my half-car-garage-sized home shop. To do this, I have to be able to do all of my other projects on the CNC, which means that it basically has to be able to handle 4x8. It may even end up, depending on the size of the footprint, that the table saw will have to go just to fit even a 2x machine in the shop. If that's the case, all of my panel work will get real sloppy with my circular saw.

My budget is around 1900. First off, is it even possible?
I looked at Joes kits. They look nice, but the 4x4 is a lot of 8020 that isn't in the kit. So I can't even price it unless I pay for the plans first, because I don't know how many other parts I'm going to have to buy. And I don't trust that I can cut the 8020 to the exact right length or drill the holes in the exact right places. I thought the 2006, although small, would at least be better in the completeness regard, but the site says it is a mixture of MDF, HDPE, and aluminum and steel parts, but only MDF and HDPE come in the kit. So I'm still left having to pay for the kit before I can figure out what other parts I would have to buy. Someone give me a link to the rest if I'm wrong.
I started to buy the 3x3 machine (fla-400, I think) from frontlineautomation. Until I realized when looking at the 2x3 (fla-100) that electronics weren't included and would push it passed my budget.
In theory, I could afford the blue chick from BuildYourOwnCNC.com, but not once I take into account that I would only be buying it to cut parts for a bigger one.

Second, the problem I have with building from scratch is that I don't have precision tools. I have an old craftsman table saw that MIGHT be able to be calibrated square. But even then, I can only cut perfect 90's on short pieces. If anything's very long, the slop in my miter guage slots and such will prevent square cuts. The fence has been pretty beat up. I think I had it square once, and it might have stayed that way for one cut, but that's it. I have a Skill brand drill press. I can line up a hole on it, but it does have slop in the spindle. I don't know if it is enough to matter when building a CNC or not. And I can't currently get the table square. I don't have a machinists square, and the various other squares I have aren't perfect enough for me. I'll get a machinists square for setting up the CNC, but I haven't had to drill anything that takes that precision yet.
Basically, I may be able to upgrade my shop to give me enough precision to build a precision CNC machine. But that in itself will eat up most of my budget, what with better guages, better fences, etc. So if I build from scratch, I either need a set of plans designed in such a way as to have enough adjustments (basically, everywhere) that I can make up for bad cuts, or reassurance from the community that the cuts aren't as hard as I think they are in the first place.

If I had any confidence that I could actually start a business with this, I might be better able to justify an increase in the budget. But as it is, I'm getting it for hobby stuff, a couple of projects that MIGHT be able to be sold if they even work out, and to help me stop spending so much frustration fighting with my crappy tools every time I want to make a box. Not to mention that I would like to get down to just this one machine that will have to be moved when the kids get out of the house and I sell it off in 3 years, instead of having to move all of the tools I have now.

So basically, with all of that in mind, am I just screwed? Either I don't get one, or I settle for crappy precision because I made a crappy build?

[louieatienza]

I'm trying to decide on a CNC to build or buy.
I've been thinking about this for years, reading these posts for years, etc. And I'm not any closer to actually getting a CNC machine than I was before.

I NEED to be able to handle 2ft across. I have one project that I just can't do by hand, and it has to be 2ft by 6ft. I can work the 6ft in sections, but there's no way I can join panels or anything to build up the 2ft after carving.
I WANT to replace all of the other tools in my half-car-garage-sized home shop. To do this, I have to be able to do all of my other projects on the CNC, which means that it basically has to be able to handle 4x8. It may even end up, depending on the size of the footprint, that the table saw will have to go just to fit even a 2x machine in the shop. If that's the case, all of my panel work will get real sloppy with my circular saw.

My budget is around 1900. First off, is it even possible?
I looked at Joes kits. They look nice, but the 4x4 is a lot of 8020 that isn't in the kit. So I can't even price it unless I pay for the plans first, because I don't know how many other parts I'm going to have to buy. And I don't trust that I can cut the 8020 to the exact right length or drill the holes in the exact right places. I thought the 2006, although small, would at least be better in the completeness regard, but the site says it is a mixture of MDF, HDPE, and aluminum and steel parts, but only MDF and HDPE come in the kit. So I'm still left having to pay for the kit before I can figure out what other parts I would have to buy. Someone give me a link to the rest if I'm wrong.
I started to buy the 3x3 machine (fla-400, I think) from frontlineautomation. Until I realized when looking at the 2x3 (fla-100) that electronics weren't included and would push it passed my budget.
In theory, I could afford the blue chick from BuildYourOwnCNC.com, but not once I take into account that I would only be buying it to cut parts for a bigger one.

Second, the problem I have with building from scratch is that I don't have precision tools. I have an old craftsman table saw that MIGHT be able to be calibrated square. But even then, I can only cut perfect 90's on short pieces. If anything's very long, the slop in my miter guage slots and such will prevent square cuts. The fence has been pretty beat up. I think I had it square once, and it might have stayed that way for one cut, but that's it. I have a Skill brand drill press. I can line up a hole on it, but it does have slop in the spindle. I don't know if it is enough to matter when building a CNC or not. And I can't currently get the table square. I don't have a machinists square, and the various other squares I have aren't perfect enough for me. I'll get a machinists square for setting up the CNC, but I haven't had to drill anything that takes that precision yet.
Basically, I may be able to upgrade my shop to give me enough precision to build a precision CNC machine. But that in itself will eat up most of my budget, what with better guages, better fences, etc. So if I build from scratch, I either need a set of plans designed in such a way as to have enough adjustments (basically, everywhere) that I can make up for bad cuts, or reassurance from the community that the cuts aren't as hard as I think they are in the first place.

If I had any confidence that I could actually start a business with this, I might be better able to justify an increase in the budget. But as it is, I'm getting it for hobby stuff, a couple of projects that MIGHT be able to be sold if they even work out, and to help me stop spending so much frustration fighting with my crappy tools every time I want to make a box. Not to mention that I would like to get down to just this one machine that will have to be moved when the kids get out of the house and I sell it off in 3 years, instead of having to move all of the tools I have now.

So basically, with all of that in mind, am I just screwed? Either I don't get one, or I settle for crappy precision because I made a crappy build?

It would be difficult to make a 2 x 6 for your budget, let alone a 4 x 8!

I think the question is, how do yo know your cuts are out of square if you don't have a precision square to reference against? You can tighten the slop on a miter slide by piening a couple places on the edge with a nailset. Or you could just mark your cutline with a good square and file, sand, or grind to the line.

To overcome the inprecision of your tools you need to take that into account building your machine. Leave clearances in you holes to adjust out any errors. Build everything in fact so that they're adjustable to an extent.

I think you'd be well over $2000, just for the electronics and some of the linear components. Heck you'll have at least a couple hundred dollars tied up in screws, nuts and bolts... That doesn't include the frame, computer... And then you need software. Who knows, an afternoon at a scrap yard could yield some key parts for cheap.

You might be better off building a small 2 x 2 machine, ad getting accustomed to using it, and then using it to build parts for your 4 x 8.

[ger21]

It's very easy to buy things twice when getting into CNC, and it happens to people every day.

The additional parts needed for the Joe's 2006 are leadscrews, bearings, gas pipe and fasteners. You can expect to spend between $300-$600 for these parts. Then add another $300 for electronics, add a PC and control software ($170 for Mach3, free for LinuxCNC)

I think the best bang for the buck is the CNC Router Parts 4x4 kit. It's a bolt together jit, so no special tools are required.
By the time you're up and running, you're looking at ±$3500 + CAD and CAM software. If you really need something this size, I don't think you'll be happy with anything cheaper.

[harryn]

I suspect that your abilities are higher than you think on building one, but you will have to be willing to job out some work to get a really nice unit.

If you lack the space, which most people actually do, then perhaps there is an under used barn or building in the area that you can rent out for a low cost. I would not be suprised if your wife is like mine, and wishes I would move some things out of the garage.

It is easier than you think to end up at $ 5K for a 4 x 8 ft router. Surprisingly easy actually, and that is even being careful with what you buy.

The reality is that very few people use a router even in a home based business with even 10% utilization. The one I am attempting to build will, in reality, probably only have 2% utilization, which makes me cringe to think about sometimes. If it weren't for my goal of using it as a father / son / daughter project, I could not justify it.

If I could find someone else locally that wanted to go in on building up a router + all of the related software, I would do it. Perhaps you can find someone within 25 - 50 miles of you that either has a router, or is interested in working with you on projects. There is a good chance that if a person has built up one of these things, they would consider to lease out time on it, or run the parts for you if you have created the code.

One approach you could try, is to do some searching here on cnczone and try to find some people within your state. Another is to place some adv. in craigslist looking for cnc wood work locally. If / when I get my router built, I would welcome ways to increase its usefullness and help break even on it.

It is rare for one person to have the time, energy, and skills to:
- Build up a router
- Know the software well
- Run the machine
- Understand wood
- Sell their services and products
- Run the business
- Keep your home life together

If you really want to do this as a "business" then think of it like a "business" and which parts of it you really want to be doing.

Just some thoughts.

[harryn]

I would also like to add that there is a BIG difference between a machine built for hobby use, and one built for commercial use.

T slot / 8020 type materials are fine for hobby use, but I can honestly say I have never seen a machine built from T slot materials on this forum that is really "commercial level reliable".

I am not saying it cannot be done, and someone is certain to point out a router that is up to commercial reliability use now that I have said it. Nonetheless, a hobby machine for occassional use and subject to frequent "adjustments" is not going to be sufficient to make your living from.

[shipbldr2000]

I understand where you are coming from. I am on my third generation machine.

You study, plan, buy, build, and you use. You learn the shortcomings of your machine and then build "the next one." Almost everyone on this site does (and that is not a bad thing if you are into building machines).

Read carefully the comments listed above about machines made from aluminum components. Then read and google.

Wood router tables suffer a lot of stress. At the risk of incurring flames I will remind you of an old expression: "The only thing stronger than steel is more steel"

If I was really serious about building a router table and I wanted it to me my "last machine" or at least "my last machine before I had so much cash I would cheerfully pay for an Onsrud or a Thermwood" I would suggest you take a very serious look at building a Mechmate. The mechmate is a steel machine, rides on steel rails, and is geared towards the home builder with a fantastic support community.

Steel is cheap, can be purchased, cut to the exact length you need, can be oversized for a few dollars, and welded up at home with a stick welder you bought on craigslist for $150 and a grinder you bought at Harbor freight for $20. Fit it out with angle iron rails and v groove bearings.

You won't get it done for $1900... but you might come closer than you think with some scrounging and time. It will be strong, it will be solid, and it might be your last machine. You'll tinker with it and upgrade over time, but you won't be tossing it out. (Just read the posts on the mechmate forum, guys seem to rarely replace them unless they are building a bigger machine)

Before you consider an aluminum railed machine... if you are considering serious heavy use... go into google, dig a bit, and find out what happens when hardened steel v groove bearings (you best lost cost rail options) roll over alumium rails while being subjected to heavy pressure. I'll give you a hint... the aluminum deforms pretty fast.

On the other hand if you are building a big table that will take 1/2 to 2/3 of your space... and you have no plans to drive it hard... using it only at hobby speeds with the resulting (expensive) short tool life that comes from an improper chip load... then go ahead and build a hobby grade machine.

Sorry to play the heavy here... but I feel this is the straight talk. You will get from your machine what you put into it. Aluminum is a great material for light & medium use. Steel is a fantastic material for HEAVY use and with some serious sweat equity steel can be very cheap. Steel will last you a lifetime.

Consider the mechmate or any other machine riding on steel rails. You can do it. Folks on the forum will reach out to you. the forums will support you through the challenges of making it through the build. Just reach out and ask for help when you need it. (I already figured out how to grind near perfect steel vgroove rails and would be happy to help you through the process... you can DIY for about $100 and a couple evenings of tinkering)

I am now going to don an asbestos suit and helmet... for a fear the flames I might be about to get from the aluminium table crowd.

Peace to all... I did not say Al tables were bad...

[RossMosh]

I would also like to add that there is a BIG difference between a machine built for hobby use, and one built for commercial use.

T slot / 8020 type materials are fine for hobby use, but I can honestly say I have never seen a machine built from T slot materials on this forum that is really "commercial level reliable".

I am not saying it cannot be done, and someone is certain to point out a router that is up to commercial reliability use now that I have said it. Nonetheless, a hobby machine for occassional use and subject to frequent "adjustments" is not going to be sufficient to make your living from.

I think ShopBot would disagree with you. Their machines are directed towards small cabinet shops and sign shops. Plenty of people use these machines to make a living from. That by definition is a commercial machine.

If you're talking about cranking out cabinet carcasses all day, then no, an extrusion build is not what you want. You want something that probably weighs a couple of tons, has servo motors, ATC, 10hp spindle, and a fantastic vacuum table. The spindle on that kind of machine is probably as expensive, if not more expensive, than most light industrial CNC machines though.

As for the original question: Your budget does not equal your demands. You could build something like a Joes 2006 and it will get you started and from there you can decide if you want to go further with this. I'm a DIYer and I've learned from past projects that you need to figure out if the project is building the CNC or learning and using the CNC. I learned this from tinkering with cars. I liked driving the car but I didn't like tinkering with the cars. Tinkering lead to driving which lead back to tinkering. I'd stop driving because I didn't want to go back to tinkering. It was the wrong way to go about things. Now I know to buy kits and so I can enjoy driving and not have to constantly worry about tinkering.

If you want to learn how to use the CNC, spend the extra money and get a machine that you can assemble in a couple of weekends. If the project is building the CNC machine, do whatever you want. Learning how to use a CNC machine, CAD, and CAM is a huge project in itself. If that's your end goal, don't drive yourself crazy trying to build and design a CNC machine from scratch. You might burn out before you even get the chance to make some chips.

shipbldr2000: Most people on here are using CNCRouterParts kit which is extrusion + steel rails. A lot of people have learned that riding on aluminum is not the way to go. That's why people use steel pipe or angle iron also. Very few people on here use aluminum anymore.

I also have some experience working with steel and welding. It's much more complicated then buying a $100 chopsaw and a $150 stick welder. If you do that, you will get a crappy product. Properly prepping surfaces, getting everything square and true, not having any warping issues, ect ect ect is not easy. Not to mention drilling steel is not exactly a fast process. There are lots of benefits to using the MechMate machine. There is no denying that. I think you're painting a picture which is not accurate though. It's not as easy as you're suggesting. Not even a little bit. Extrusion + bolt on components is worth the extra money 99% of the time. If you're comfortable working with steel and welding, you can save major bucks using all steel but let's not pretend systems like the CNCRouterParts or Joes CNC aren't a fantastic middle ground. ShopBot has sold thousands of machines on the same principles.

[shipbldr2000]

Actually, We are agreeing. I never said anything bad about aluminum frames... I said bad things about V-groove bearings on Al rails...

There are lots of great Al frame machines, no doubting that. Shopbot is a great example. But also a 12k example for a PRS or 20 kfor an alpha. 5 to 10x the OP's budget.

The OP was looking at doing a buildyourcnc system. Build your CNC systems are aluminum rails...

Given a larger budget, Shopbot is at the top of the list I would suggest to the OP... but shopbot is also NOT an aluminum rail machine. Shopbot has a robust aluminum frame with steel rails. Same with CNCRouterParts or Joes CNC. ALL excellent machines, just not in the OP's budget range.

If this guy seriously wants a machine for close to 2k... it is not going to happen with aluminum prices where they are now. It will take sweat equity from hell to come anywhere near the OP's budget.

The OP has space, time, and skills. The OP is a cabinet maker. He knows how to measure and check square. I am sure he probably has a drill press. He has clamps, and some amount of tooling. He can make a good mechmate pretty cheaply if he scrounges and takes his time.

I once again agree with you that working with steel is not a cakewalk... but if you have time and no cash, steel is great. My steel distributor gives me huge scraps for free or literally for a couple bucks. In balance it should also be noted that the $$$ you save with steel will be paid out in in sweat.

But if what the OP has is sweat and not cash... The OP can still have a great machine. He's just going to have to pay for it with hard work.

I have welded and worked with steel for years. There's a learning curve, and the OP will destroy a few parts along the way. He will learn about putting too much heat into a part, warp some things, etc... but he will figure out. Others will help him on the way.

A small budget doesn't have to mean out of reach. It just means you are really going to have to work to make it happen.

[Cavnar]

It would be difficult to make a 2 x 6 for your budget, let alone a 4 x 8!

I think the question is, how do yo know your cuts are out of square if you don't have a precision square to reference against?

I don't need 2x6. I'm fine with 2x2 (or even 2x1 for that matter) for the project I have to have it for, as I can always carve the first 2 feet, leaving markers to let me move it down precisely, and then move it down and carve the next 2 feet. I only NEED bigger if I am going to replace the rest of my tools with it.

And I know the table saw is out of square because my boxes end up slightly twisted. Opposite sides are slightly out of square in opposite directions.

The OP is a cabinet maker

I've built cabinets, but not professionally. I wouldn't say I was a cabinet maker. More of a hobby woodworker. But that's one of the projects I need to get done, that will be harder if I can't cut 4x8 panels. I'm building drawers under my workbench, and I really just don't want to have to cut a 4x8 sheet into small sections before putting it on the CNC. I want to lay out the entire sheet and cut it all in one go, not even considering that the drawer sides are 26 inches long, and that one is 4ft wide (I've already built the drawer, but I will still need to cut the face panel).

I've considered steel, and seen the mechmates, although I haven't actually read the plans. I have a mig welder I could do the welding with, but I can't see how it could not be expensive. Although I also have a poor grasp of the cost of things I don't buy regularly. And I'm much less confident about my ability to eventually get anything square there. (The most complex welding I've done is this drafting table if anyone is interested.)

But for the work I plan on doing, I think a hobby grade machine made out of MDF will be accurate enough. At the same time, I don't know what accuracy I can really expect out of any machine I build. If I carve traceries in a piece of wood, mirror it in another, and put them back to back, can I expect misalignment of 1/64, 1/32, or 1/8th? I'd like them to line up perfectly, but if I have to live with 1/64th, or 1/32nd I will. I couldn't live with 1/8th.

I'm not working toward starting a business. Not honestly. I can put an ad in craigslist or something to cut or carve stuff for small shops in the area if they have the files. Maybe do the design for something occasionally. And if that made me enough money, I would upgrade. But I'm can't afford to take time away from my current job to try and actually run a custom woodworking business or anything. The best I can hope for is to come up with some kitchy stuff (like plywood snowmen) that I can sell online, or some sort of furniture kits or something.

What I really hope to accomplish with a CNC machine eventually is 1) to be able to get this one project out of my head just to see what it looks like, whether it sells or not, 2) cut a couple of aluminum diamond-plate sheets into shape for another project (I can weld it, maybe, but I know my angle grinder won't let me cut 4 pieces out identically), and then replace all of my other tools for home projects with it. I don't mind tinkering with the software or the hardware. I like the electronics, and I program for a living. But I don't like the non-productive tweaking I have to do with the crappy tools I have now every time I make a cut.

[louieatienza]

I would check out the Solsylva.com plans. I built mine with a circular saw, and crappy drill press. It's very tolerant of cutting errors, and easily adjustable. Plus you can build one relatively inexpensively. It uses regular framing lumber also, which I feel is a bit surdier than MDF and screws hold to it better. There arealso full scale plans for the critical parts that you can print, cut out, and paste to your part to cut.

[OCNC]

Second, the problem I have with building from scratch is that I don't have precision tools. I have an old craftsman table saw that MIGHT be able to be calibrated square. But even then, I can only cut perfect 90's on short pieces. If anything's very long, the slop in my miter guage slots and such will prevent square cuts. The fence has been pretty beat up. I think I had it square once, and it might have stayed that way for one cut, but that's it. I have a Skill brand drill press. I can line up a hole on it, but it does have slop in the spindle. I don't know if it is enough to matter when building a CNC or not. And I can't currently get the table square. I don't have a machinists square, and the various other squares I have aren't perfect enough for me. I'll get a machinists square for setting up the CNC, but I haven't had to drill anything that takes that precision yet.
Basically, I may be able to upgrade my shop to give me enough precision to build a precision CNC machine. But that in itself will eat up most of my budget, what with better guages, better fences, etc. So if I build from scratch, I either need a set of plans designed in such a way as to have enough adjustments (basically, everywhere) that I can make up for bad cuts, or reassurance from the community that the cuts aren't as hard as I think they are in the first place.

An accurate table saw with a sharp blade is a big plus in building anything made from sheet goods. A skilsaw (with a sharp blade), a straight edge, a couple of small clamps and a new tape measure will also work well when cutting sheets but it will take more time to execute the cuts. Cuts can be further cleaned and squared using a router, a straight bit, a cutting guide and a couple of clamps. A mini drill press can be purchased for under $100 and is all you will need for drilling holes. A 4" by 6" precision square (0.0006" accuracy along the length of the blade) can be purchased from Mcmaster-Carr for $15.00. This can be used for setting up any tools. There are all kinds of small and medium sized squares that carpenters use to make layout marks and that can also serve as skilsaw guides that are accurate enough to cut parts for a cnc machine. These are inexpensive and readily available from Home Depot, Loew's, etc. In spite of not yet having a specific plan to to engage your interest all of the deficiencies you site above shouldn't stop you from moving forward. I would start by building that drawer box that keeps coming out skewed with your choice of the tools I list above so that it's actually square. You need to work out a procedure that gets the correct result in a simple form and then you'll have the confidence to take on the cnc project.

[Cavnar]

I saw that one today and am thinking about ordering the plans. I don't know that I like the idea of home-depot wood being the guiding line for the rails, though. I just bought 24 "prime" 2x4's this week (for raised bed boxes). Some were good, and some were twisted. But even the good ones, I don't know that I would trust them to be "true". I don't have a planer. I do have a 6 inch jointer, but that's another thing that will take me 2 days to get calibrated. It's a craftsman, so the outfeed table doesn't move, and it's currently cutting so that one end of the boards end up thinner. I think that means the knives are too low. But even if it's calibrated, it's not going to help me get the edges parallel. From what I saw of other builds, the y rails are on each edge of a piece of lumber, so I'm paranoid about the lumber not being parallel. But at least it doesn't need a torsion box (at least on the build thread I read). That's one of my biggest fears with something like the JGRO build; I may be able to adjust everything else, but if my joints for all the interlocking pieces of the torsion box are off, the base is going to be pretty funky. That seems like an awful lot of places for cumulative error to build up. I don't know that anything I build above that would ever be true.

Part of my problem is that I'm a perfectionist. Things like, "lay a straightedge across the table and the knives should be 1 paper thickness higher" (or lower, or whatever it is I've read) just don't sit well with me. How do I know that my piece of paper is exactly the same thickness as the one they calibrated with? I have that problem with all of my projects. But with things like my drawers, or the flower boxes, or whatever home projects I do, I just deal with the imperfections. My drawers may be slightly twisted (although somehow I managed to get that out of most of them) but there's enough gap around them that they still fit in the cavity and slide along the rails. But I don't want to build a CNC machine that is going to mean that everything I use it to build is going to be off by that much before the blade even touches the wood. I want to know that if I want to make two parts fit perfectly, like an inlay or a perfect dovetail, that I can.

[shipbldr2000]

I
Part of my problem is that I'm a perfectionist. Things like, "lay a straightedge across the table and the knives should be 1 paper thickness higher" (or lower, or whatever it is I've read) just don't sit well with me.

Bootstrap - Build a crude but working cnc. Use it to cut surprisingly accurate parts for your "real machine"

The "crude" bootstrap machine you build will probably be accurate to 0.01

Lots of guys on buildyourcnc.com have built a "book machine" for $150.

After the bootstrap machine cut all the parts are cut for the "real machine" then transfer over all the hardware from your bootstrap machine into your real machine.

Make the bootstrap from MDF, the real machine from MDO.

Then throw away or give away the bootstrap machine parts to someone else doing the same thing.

Or if you are close enough by... maybe some one on the forum with a router table will just cut the parts you need for you (beer appreciated). If you are anywhere near Columbia, MD stop on over. Have table, will cut. Have the ability to manufacture 20 foot steel rails from cheap 2x2 steel angle iron... in about 2 ice teas flat.

Who else else can give this guy a hand? Anybody know where we can grab some cut files for a 4 x 8 router table? Mine is a 7 x 14 and I think that is a bit of overkill...

[adprinter]

Your mention of a budget of $1900 limits what you can do. Someone inspired me a couple of months ago to research what it would take to duplicate my machine at today's prices. I was shocked to learn, that it would exceed $3500 to build the same machine today!

I am not a professional machinist by trade, however I have tinkered with machining for the past 20 years or so. The tools I used in my shop to build my CNC include: floor model drill press with an X Y mill table, which I installed DRO's on each axis. I was trying to use it as a vertical mill at the time, but learned that the spindle just can't handle the side-load vibrations of using end mills. (Test cuts resulted in the taper falling out of the quill, due to the vibration which end mill cuts exert on the spindle).

However, the DRO's proved indispensible in acurate hole locations for the various parts of the CNC frame. The main thing to remember when drilling holes in metal, is to ALWAYS start the hole using a center drill. Then, change to the correct size drill bit to finish drilling the hole. This will prevent the "wandering" of normal drill bits, since the center drill creates a dimple in the metal which the normal drill bit will follow (at dead center). You can actually get some pretty acurate results, by assembling the various pieces, and use C clamps to clamp everything together when drilling holes.

For holes which are to be tapped (threaded), always clamp the mating pieces together with C clamps, and start the hole with a center drill, then change the bit to the size needed for the particular thread size, and finish drilling through both pieces. Then, unclamp the pieces, and change the bit to match the bolt size needed and drill through the top piece only. Then use a tap to cut the threads into the hole in the bottom piece.

Also, large size holes should be drilled at the slowest speed settings on the drill press- otherwise you will quickly BURN the drill bits. Coolant should also be used. Cool Kutz, or similar spray can products works great for this. A metal cutting band saw is also needed. The Harbor Freight 4x6 Horizontal/Vertical is what I used. It took me about 4 months to build my first machine. I also have a 13x40 metal lathe, which I used for some custom made parts. However, one could get by without one of these. Then main tools you will need, are the drill press and band saw. A machinist's square, carbide scribing tool (for marking hole centers in the metal pieces), a digital caliper, tap and die set (and matching drill bits), a counter sink bit (for cleaning burrs from drilled holes), mill files, sand paper, and a dial indicator are all recommended minimums for "tooling up" in preparation for a DIY build.

While I have not specified the dimensions needed for my build, you might find the information contained in the attached MyCNCv1.PDF helpful. I have since completely redesigned (and rebuilt) the entire machine, since I later added a 4th axis -which required an increased Z axis travel to accommodate the 4th axis. Hope this helps!

[OCNC]

Part of my problem is that I'm a perfectionist. Things like, "lay a straightedge across the table and the knives should be 1 paper thickness higher" (or lower, or whatever it is I've read) just don't sit well with me. How do I know that my piece of paper is exactly the same thickness as the one they calibrated with? I have that problem with all of my projects. But with things like my drawers, or the flower boxes, or whatever home projects I do, I just deal with the imperfections. My drawers may be slightly twisted (although somehow I managed to get that out of most of them) but there's enough gap around them that they still fit in the cavity and slide along the rails. But I don't want to build a CNC machine that is going to mean that everything I use it to build is going to be off by that much before the blade even touches the wood. I want to know that if I want to make two parts fit perfectly, like an inlay or a perfect dovetail, that I can.

There is no perfection. No parts ever fit together 'perfectly'. There is only acceptable tolerance. That piece of paper you're agonizing over is 'any ordinary piece of paper' (a piece of newspaper, a piece of copier paper, a sheet torn out of a magazine). The principle of setting the jointer knives is to make sure that they are not below the level of the outfeed table but also 'only just above'. A piece of paper is a handy reference for 'just above'. A jointer even when properly set up is always cutting a slight taper but it is within an acceptable tolerance for secondary operations (a jointer is a surfacing tool designed to generate 'flatness' only that's why it is typically used in conjunction with a thickness planer to establish 'parallelism' of opposite sides or edges). Even the most expensive machine in the world is built to a tolerance and not to 'perfection'. What you seem to be lacking is a knowledge of the procedures to get to a particular tolerance. If you don't practice though, obtaining a particular tolerance on even the simplest form will remain a mystery to you. So sorry to hear that you suffer from 'perfectionism'. It seems like such a debilitating 'disease'. Maybe the 'cure' is somewhere here in the 'Zone'. You should be able to tune the jointer in less than 15 minutes.

[louieatienza]

I agree with OCNC's statements. There will never be 100% perfet fitment between parts; you always have to work ot a certain acceptable tolerance. Even if you could build a machine to near zero tolerance, your raw materials may not be "perfect." Consider wood, which can expand/contract at a rate far greater than the accuracy of your machine. A round hole one day can be oval the next, or even within a few hours, unless you have climate control. Even the most expensive machines have coolant running through the linear bearing components and ballscrews to help with thermal expansion, and even then these machines need to be "warmed up" before any cuts are done.

Also "perfection" comes at a cost. Not only does the cost of precision parts go up with the level of precision, installation of these parts become more critical.

Tha said, depending on what you want to cut, a JGRO might not be the best choice, because of the unsupported linear rails. A Joes2006 can bedaunting to build by hand; it's intended to be machine cut. The MOMUS design was made by the designer using very basic hand tools.

If you want better dimensioned lumber you can peruse a hardwood lumberyard. They can even dimension lumber to your needs for a small fee. And the hardwoods will generally be more stable and heavier.

Don't get trapped into a Xeno's Paradox type of thing. When done this will be more accurate than any tool you have in your shop right now. Learn how to make precise parts on you tablesaw. It's a pretty basic skill to make a crosscut sled for a tablesaw for example. Perfection is in the eye of the beholder too; I've done work at multi-million dollar estates, by hand, with people watching over my back. You learn to do things the best you can, and hope others see that you did...

[Drools]

I just want to add a comment on the cost.
There is no way I could have spent the total cost of my machine in one shot.
I decided to do a lot of research here and other sites and from there I started to collect the parts that best suited what I wanted to make. Collecting the parts took time, example: purchase a G540 kit now experiment with how it works with the software so when the time comes to actually mount the motors you will have that part done.
In the mean time while you are learning the electronics side you can scour, hunt scrounge the parts for the frame. I spent over 2 years searching eBay for linear rails in the size I wanted. I ended purchasing a good lot from a user on the zone.
I guess what I’m trying to say is this, if you have the time to spread the cost out over a period of time in the end you will end up spending quite a lot but you won’t need to refinance your house to do it.
Now that I’m almost done with my machine(s) I have some “left over” parts that I will sell to help with the purchase of upgrades and tooling.
Just a note on cutting 8020, you can purchase it cut to the exact lengths you need which is what I did.

[harryn]

I think ShopBot would disagree with you. Their machines are directed towards small cabinet shops and sign shops. Plenty of people use these machines to make a living from. That by definition is a commercial machine. ....... ShopBot has sold thousands of machines on the same principles.

RoshMosh - you are right. ShopBot does have some very nice machines in this category. Thanks for pointing this out. In fact, there is a ShopBot "Buddy" that would physically be very interesting for many people to at least consider a close look.

As you note though, $ 2 K is very constraining.

Knowing what I know today, which is still a long way from really experienced, my suggestion is that you try to find someone willing to rent time on a cncrouter. Maybe you can find someone willing to do it for $ 15 - 20/ hr or maybe $ 100 - 200 / month with you supplying any bits and consumables. (do it for perhaps 1 year)

If that sounds expensive, consider that just electricity cost here to run a home size unit will be around $ 2 - 3 / hr, not to mention wear and tear on some expensive system parts. If your business is going to be a go, it should start to become clear within 18 months.

Perhaps post what state you are in.

[mgman71]

Hmm, this is a good thread, i'm kinda new here and am interested in building a 4x8 machine.. I can build the structure no problem but not sure where good places for all the parts are online. I'm from canada.

[brucer]

Hmm, this is a good thread, i'm kinda new here and am interested in building a 4x8 machine.. I can build the structure no problem but not sure where good places for all the parts are online. I'm from canada.

I used, vxb for bearings, moore gear for racks, hubbard cnc for motors and cheap tb6560 just to start with, QTP for timing pulleys, ebay for timing belts.
friend bought an enclosure at a salvage auction that was built out of extrusion for $40 he give me what I needed for my table, 1 1/2x1/4" angle was like $75.

if I had it to do over again I would have built it all from 2x3x3/16 rectangular tubing..


I have right at $1000 in mine, I do all my own machine work, mig welding and tig welding. http://www.cnczone.com/forums/diy-cn...cnc_build.html

I will buy individual drivers and bob pretty quick, just using the tb6560 to get it running.. I probably spent a good 3 months planning and getting parts and materials rounded up.