Retroactive build Log: DIY mill from surplus parts

[janbbeck]

At the beginning of the year, I decided that I wanted to build a CNC mill. I took pictures along the way, but not very many. Today, I took it apart and cleaned it to move it to another room, and took new pictures of the components. I now have a set that can be used to describe this build, I think. I am going to mix pictures of the current setup with pictures I took while I built the mill to fill in the gaps.

So I am basically writing a log of my progress over the past year. I did not have much money available, but my work has a machine shop and part of my fringe benefits is access to the shop and stock room. I was able to find quite a few nice things that saved me a pretty penny.

For this first post, I am going to include an image of the current mill as a preview of the log.

I am hoping that the info here helps or encourages someone else to build a cnc mill.

This machine currently works and has been making parts, including some for itself.

[janbbeck]

Ok, so this whole thing started by me finding a small x-y table that had only a motor on one axis.

All I needed to do is:

- scrounge up a second stepper motor and mount it
- find a controller board
- find an old PC with a parallel port to drive the board

A friend gave me several stepper motors as an extended loan as well as a controller board to drive them. There were some partial PCs in the stock room from which I built a 700 Mhz Pentium III with an extra parallel port card - for a total of 2 ports.

[janbbeck]

Here a pic of the controller board. It is from a little while later, when I was cutting some plastic, so there are chips all over it. Of course at this time, there was no way of cutting yet, just the x-y table.

[janbbeck]

Scrounging around the stock room, I found an x-table, that seemed large and rigid enough to be my z axis.

In order to mount it vertically to the x-y table, I used a cast iron tooling plate and right angle cast iron knee, both of which I found in the stock room as well. so far no money has been spent.

[janbbeck]

Next, I mounted the x-table to the knee.

Then, i mounted a Skill wood router I had in my garage to an aluminum plate. This plate was then mounted to the z-axis (previously called x-table).

I also mounted a CNC vice for a Roland CAMM machine. (I bought this one by mistake years ago, and it had been collecting dust in my garage)

I also taped a shopvac hose to a height gauge to have some dust extraction.

At this point I was able to make small parts. No money spent. A couple of weekends worth of work.

[janbbeck]

The controller I had borrowed was quite old and limited to 3 axes. I wanted 4 axes or more, but was a bit hesitant to buy a board I found on ebay that could handle 5 axes at reasonable power. My wonderful wife made up my mind by buying the board for me.

I should not that in the previous post an autotransformer is visible, which I used to vary the speed of the router. Of course it too came from the stock room. From there also came a nice bench power supply to power the new controller board - but I only got that one as a loan :P

[janbbeck]

In order to make my 4th axis, I decided on retrofitting a spin indexer, which I also scrounged up. The bracket to mount the stepper motor to the indexer is the first part the machine made for itself.

I was able to scrounge the drive belt and a pulley for the stepper motor. I then proceeded to machine an aluminum adapter to grab the spin indexer in the machine shop. I also put teeth into it to grab the belt, but somehow I got the spacing between teeth wrong. I then bought a plastic pulley off of amazon and mounted it to the aluminum. It cost me about $5

I also mounted an aluminum plate to the bottom of the indexer, so that I could clamp it into the vice with an offset.

This setup allowed me to use the 5-C collets from the machine shop to clamp to the work pieces.

This ends the part of the machine where I incurred almost no cost (hardware, at least - the mach3 demo only does 500 lines of code, but there is always emc2).

[janbbeck]

at this point the machine could do small parts up to 4 axes. Here some test pieces. the smaller piece is from a broomstick or something that I put into a collet on the 4th axis.

[janbbeck]

In order to be able to machine larger pieces, I bought a 4 jaw chuck on ebay that fits into the spin indexer. Using a 4 jaw instead of a 3 jaw enables me to mount plates, not just cylindrical pieces.

[janbbeck]

I found that the runout on the Skill router was awful. After some research I found that this is due to the collet and that precisebits.com makes precision collets for other routers.

Therefore, I ordered a Bosch router and a high precision set of collets to replace the crummy Skill. That reduced my runout to 0.0002 an inch down from the collet.

I should note that wood routers are universal motors. They are very unstable under changing load conditions. So I had a friend from the motor winding business show me how to rewire this thing as a shunt DC motor.

This was detailed in another thread:
http://www.cnczone.com/forums/genera...nt_torque.html

The whole operation was totally worth it. It does not slow down one bit from the preset RPM even with heavy cuts.

[janbbeck]

At this point, the 5x5 inch travel on the x-y table seemed puny, so I decided to replace it. As luck would have it I knew somebody who had planned on building a cnc mill and had already parts on hand, but decided against it. He sold me some pieces cheap and gave me others for free.


Among these pieces were two linear tables, a long one for my x and a short one for my y. I mounted them with the help of some aluminum plates to the tooling plate. On top of the y axis, I mounted a large aluminum plate to act as a table to clamp the workpiece, vice or 4th axis to.

[janbbeck]

The reach of the arm holding the router on the z axis was now too short and so I built a longer one. Predictably, there was very little stiffness in this arrangements, so I reinforced the z arm to stiffen it up a bit.

I first tried to alumiweld the reinforcements, but my torch did not give nearly enough heat. When I took this to a friend to TIG weld it, he laughed at me saying that his welder did not have nearly enough amps to do this. He also mentioned that if we did use a heavier welder the workpiece would likely distort.

So instead I screwed the braces on.