This is a purely CNC-oriented thread that I started because quite frankly, I am interested in it, and have been most seriously considering building one to help me do things like make PCB's.

I probably even have enough raw material around here to make one. What's stopping me? Lack of focus is the most likely gremlin; a hazard of hanging out at places like this, I suppose. I've been doing fits and starts on this thing for a few years.

So, let's get down to business, and git 'er done.

I have a couple of pairs of bipolar steppers, and a whole boatload of unipolar steppers. The bipolar steppers have higher resolution than the unipolars, and are much more robust.

Time for brainstorming

 

You may bed able to pick up some tips on things to consider from this PART-ONE of a THREE-PART article on "build-you-own" cnc machine.

hgmjr

 

The question is which choice is the more precise, move the "tool" or move the "work".

hgmjr

 

I would vote for moving the work. The bed can run on a flat, solid surface, and the milling head can be very rigidly mounted. Once you have the plane of the bed set, it should be adjustment-free (except for the zero-zero reference point).

 

I would tend to agree Bill. I suspect that moving the "work" would be the better choice since that entails moving the lower mass. Tools are generally heavy and require the management of the power cord and on/off control cable.

This approach would also permit anchoring the tool solidly so that it would form a stable reference during the milling operation.

hgmjr

 

I've also been leaning towards moving the work. I've gone so far as to build a rather solidly-reinforced small plywood platform to that effect.

I just need to get this thing done.

 

Some pictures, Perhap?

hgmjr

 

Let me mention there are several DIY sites dedicated to CNC. This is probably common knowledge to most of us. Here are links to three that I visit occasionally:

http://groups.yahoo.com/group/DIY-CNC/
http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/
http://www.cnczone.com/

IMO,the Yahoo groups tend to get a bit winded and off subject. CNCzone has some nice completed projects.


I am not so sure which has the lower mass, the cutting tool or the work + work table. The work itself is probably pretty light, but it has to be mounted to something. Take a Bridgeport, for example, the tables and base needed to get 3 axises are far more massive than the little head. Moving the work in X and Y seems pretty straight forward. The Z axis seems to offer more complexity and is where a lot of the mass in a Bridgeport is located (I know a Bridgeport has two ways to move the Z-axis. It moves the tool for one of them.)

In contrast, it seems most of the home units are designed around moving the tool with a gantry or like design.

John

 

If you're doing PCBs, then moving the spindle is fine. They actually do it on commercial machines. The main reason to move the piece is for that extra mass of damping material. One word of advice, build it big and heavy! Never underestimate the rigidity required to do quality work.

Wook,

I would shop around for linear guides on Ebay, don't bother with crappy gibs or linear bushings. Make sure the guides are the type where you have many circuits of recirculating balls, rather than two. It helps with stiffness in all possible directions except the intended.

You need a good cnc program like Mach3 Mill to interpret your gcode and run your machine. I use Mastercam to do all of my gcode, since it is extremely easy to use. I used a PIC-SERVO IC and some power IGBT module to do my servo control with a brushless motor along with an optical encoder for each axis.

I bought a Taig Mill and converted it for CNC operation. I am not happy with it at all though and wish I had built one. You kind of need a machine shop to do it in the first place though, which is a problem for many people. One day when I have more money, I was working on a design using granite surface plates and linear guides that I want to try out.

I guess it really depends on what you want to do. My taig can eat through any plastic at 40 IPM, but has trouble with metal. You really need to limit passes to 0.05", which equals a lot of time per part. If you're doing PCBs, it is more than capable of it. I actually did a small PCB on it and was surprised how well it worked.

I think I may be rambling on, not used to working fulltime. I miss my laid back student life

Steve

 

Coding would be an issue though, wouldn't it? On a homebrew rig?

 

perhaps a wrench in the discussion, but most CNC machines i've seen the tool moves. i guess it depends on how many axis you are planning to build.

 

In the world of auto bonders, which don't have much strength but bocu precision, I've seen both schemes used. Currently for speed it is the head that does the moving.

 

From the responses in the thread thus far, I am beginning to sense that the preference for moving the "tool" or the "work" hinges on the nature of the application. At least for movement of the x-axis and the y-axis that would seem to be the case.

If complex curvatures are to be handled then the z-axis may involve moving the tool. Or I suppose one could break the process up into two steps with the work reoriented so that the y-axis or the x-axis could serve as the z-axis for the second operation.

hgmjr

 

What's stopping me? Lack of focus is the most likely gremlin

From a Jar-Head?! Say it ain't so!



I've done precious little machining myself, but most of the books I've read on the subject agree movement of the work provides better accuracy.

Unless the mill will be very light duty (PC boards do count as very light duty), I suggest servo motors instead of stepper motors.

 

I think steppers are better for PCB milling, since it is light work. Steppers will keep their precision under light loading, but not so much under heavy loads. You can, of course, buy really heavy duty steppers that'll be great for almost any milling.

I had a hard time with my servo controller and didn't really like tuning the PID. I wished I just went with simple stepping motors

I definitely want to build another machine one day, but I will have to wait until I get my hands on a decent manual mill first.

Steve

 

i frequent http://www.cnczone.com/ when i think about building my own cnc table to move my plasma torch around for cutting sheet and plate. i'm just too busy these days to tackle such project. i'll probably just buy something...

 

Wookie, I have built a CNC table from scratch using inexpesive parts.
My goal for the CNC was to build propotype circuit boards (mechanical etching), Pen plotter, face plates and engraving all using light materials.
I think the size of the table will determine whether you will want a moving table, mine is a 20"X20" so I have a moving tool.
What I used.
3 Nema 27 50mm bipolar steppers (small but seem to work just fine for what I'm doing)
for the axes I used precision grounded rods and bushings (slides would be the way to go but expensive) http://www.mcmaster.com/
You could use a Dremel tool to cut your work.
I'm using Mac 3 tor run it. Very good program and inexpensive, another program you can look at is DeskCNC it has the Gerber to Gcode converter built in.
If you are going to do mechanical etching there are some things you should look for.
You need the right routing bit. http://www.thinktink.com/stack/volumes/voli/store/mechmill.htm
Control of cutting depth is very inportantant if you are expecting good detail. (Less than 25mil path )
I am now routing 2 sided circuit boards with excellent results
In building the CNC I had access to a machine shop so it did not cost me much. The project can get very expensive if you have to buy slides.
I could post a picture if you are interested
Hope this helps

 

To add to my last post
An inexpensive hobby plotter can be built using good quality drawer guides with ball bearings for Y (moving table) and Z axes, plywood and 1/2 inch aluminium plate for table.
Ready rod can be used for lead screws.
Good site for controller or if you want to build your own http://www.students.tut.fi/~kontkant/a3977.php.html

 

i agree, cnczone web site is the place you want to look. write your explanation, they will help. believe me, you want help, as there are many pitfalls in cnc building. you can spend months (& $) debugging a problem maybe easily avoided at the beginning.