[jinns]

First let me say, hi everyone! I'm new around here. I stumbled across the site a few days ago and MIND = BLOWN!

Potentially about to embark on my first build after spending a lot of time cutting out some letters for a sign on my scroll-saw (long story). I knew there were better and more accurate ways to do this.

The other key to the design is affordability. All costs must pass the "house budget committee", not that I have to get the whole thing from the junkyard and assemble with duck tape and popsicle sticks, but all expenses should examined. I like to look at the cheapest solution and calculate is the upgrade will pay dividends. i.e. all thread vs. ACME vs. ball screws.

So what where do we start? First, let us look at what it will be used for.

• hobby woodworking
• amatuer signmaking
• building v2.0

I'm guessing I should first look at the size of the desired cutting area. Any suggestions?

So all help is appreciated. Please speak up if (more like when) you notice me do or say something stupid.

I'll try to keep this up to date will all of my decisions and considerations.

[jinns]

I should also mention that I am strongly leaning towards an MDF build. I have experience working with the material and the tools to do it. Aluminum extrusions look nice, but I suspect considerably add to the price of the build.

[jinns]

Tossing ideas around in my head.

Trying to weigh the differences between the angle iron (aluminum) designs and the black pipe ones.

I know that the pipe is stronger and will resist deflection, but the designs I have been seeing with 2 on each side (and torsion box supports) of the x-axis seem like overkill.

Anybody want to steer me one way or the other? I'm actually leaning to the aluminum angle, looks cleaner and appears more simple to build.

I think I have a size figured out (almost). My most recent project needed a shape cut that was 20" x 20" roughly. That would be my minimum working area, though I would prefer a 24" width. I really dont know why. For length the longer, the better, but 4' would be the maximum I would be thinking of stuffing into my already stuffed shop. Reducing that dimension by about 12" for the unusable length for the gantry etc. would result hopefully in a roughly 36" cutting area.

So there we have it... 24" x 36" I hope.

[CarveOne]

I recommend 3' x 4' overall, which gave approximately 28.5" x 35.3" actual working area for my most often used machine. This machine has recently made a sign that was 24" x 36" x 1.5" because it was cut to size (no waste material) and everything that was to be carved did not exceed the 35.5" working length. For a machine this size I would recommend laminating two layers of MDF for some of the structural parts that absorb machining stresses. All-thread rods are too slow, and you will quickly change any notion that you may have that it doesn't matter. The cost difference is sizable though. Use multi-start ACME rods of 1/2-10 2 start or 5 start so that the motors don't have stalling problems from maxing out the rpms trying to get higher feed rates. A good operating range is 100 to 300 inches per minute jog speeds. 1/2-10 1 start is too slow. There are some who have been using 8 start and like them. Mine uses 5 start.

The #1 motor driver is the Gecko G540 and 380 oz-in size motors for this size and larger machines. They are not low cost, but they are well worth the cost in performance and in minimal frustration level. They are available in plug and play kits from a few CNC Zone advertisers. My recommendation is CNC Router Parts or Keling, Inc. The low cost TB6560 drivers have a "somewhat colorful history" around here. Some people get them to work, some people haven't been so successful, and some of the drives have been DOA out of the box. YMMV

[jinns]

Thanks CarveOne.
Just checked for a rough price on the 5 start, looks like $34 for 3' lengths. So for 3 I'm looking at $102. Actually not as bad as I thought it would be, but OUCH at $30+ per nut that's a lot to swallow and would require 6 if I was to try to rig up some sort of anti backlash design.
Total $282 for just the lead-screws from MC.
Looking at the Gecko G540 setup from one of the mentioned suppliers is listed at $519.
Now we are up to $801 (before shipping, taxes etc)

That's already pushing the budget. Anybody have suggestions on where to save money?

Cheaper motors or controllers? I'm not crazy about rolling the dice on eBay, but if there is a supplier on there that people have had success with I would be willing to put it on the list of considerations.

Is the acme 5 start really needed?

Should I be thinking of scaling down the operation? Would that help?

[Trotline]

dumpsterCNC - anti-backlash solutions for home and industrial linear motion for relatively low cost ACME anti-backlash nuts, as well as some nice collars and couplings. The biggest cost-saving measure is to spend a lot of time on these forums, looking at how other people have built their machines. Look for machines with the called "Solsylva" in particular. That's a very low-cost design.

The bottom line though, is that CNC is for the most part a pretty expensive undertaking. The bottom-end, most DiY machines, made from MDF and driven by all-thread do work, but even they will run over $1k, most often. Those machines are generally very slow, which might not sound so bad, but in use, becomes really annoying, and leads to (costly) upgrades and redesigns.

Take some time to dig through these forums, and get a realistic sense of the costs involved. It might mean waiting a few extra months, but spending more money on your initial build can end up saving you more, in time.

Luke

[jinns]

I know buying unknown parts from China is a risk, but the idea of a cheap kit to get started with is tempting. Nema23 Stepper Motor 270oz in 3A Dual Sahft 3 Axis Board CNC Kit Longs Motor | eBay

Also my idea may get scaled down considerably. Although the work area of the shape that put the CNC router in my head was on a surface 20" x 20", the piece itself was only about 18' x 18". Maybe some downsizing is needed. A bigger machine would be v2.0.

If scaled down, would a normal 1/2 10 single start acme be reasonable? What about using a 16tpi all thread?

[CarveOne]

Don't buy everything at one time. Do it in steps as you need it and as income allows. Spread your build out over a few months and the financial pain won't be so bad. Start with buying the materials to build your basic frame parts, but know in advance what the hardware and electronics will be. Get the hardware second. Get the electronics last. You can buy one motor during the assembly build so that it can be used for determining hole patterns or you can get that info by doing a web search for drawings for the NEMA 23 (or NEMA 34) motors that you will use. Having a motor in hand gives you an idea of the size and room that it will need around it. If you are building from plans, having a motor available is less of a need. My first three machines (so far) took around 8 months to 1-1/2 years each. Solsylva 25x37, a self designed 5' x 12', and a self designed 3'x4'. The old Solsylva's mechanical and electrical parts are now available for machine #4, but I also have a new G540 complete kit from CNCRP waiting for some new project when I can get to it. As for software, you can spend nearly nothing or as much as you want to for that stuff. Get that last. Mach3 from Artsoft and anything that Vectric Ltd makes are my recommendations. I started with EMC2 (now called Linux CNC).

[Devastator]

Thanks CarveOne.
That's already pushing the budget. Anybody have suggestions on where to save money?

Cheaper motors or controllers? I'm not crazy about rolling the dice on eBay, but if there is a supplier on there that people have had success with I would be willing to put it on the list of considerations.

Is the acme 5 start really needed?

Should I be thinking of scaling down the operation? Would that help?

Saving money depends on what tools you have, and how much talent and time you have. I built a 24X48 Solsylva for ~$500 but I made everything I could including the delrin AB nuts, and most everything else was dumpster dived. And it took two years.

I did find 291 oz in motors @ Circuit Specialists for $25.50 ea. and if you solder up the hobbycnc ezdriver you can have a pretty affordable setup. I'm using those on mine and with 5 start screws I'm getting up to 400 IPM rapids. I will eventually upgrade to Gecko, but so far it's been rock solid.

Do not skimp on the leadscrews. That is one place not to cut corners because if you do, you'll spend the money twice when you redo it out of frustration because it is too slow. I'm not talking too slow for your patience, but too slow to get proper chip loads and you'll burn up your cutters and material.

Not to scare you off, but the part that kills me is how expensive the cutters are. I probably have about as much in tooling as I spent on the machine already. The cost of learning.............