Final Assy - Part II

5 Bears Home Homebrew CNC bench mill

In Part 1, the mill was assembled from the base up. The X and Y axes were both finalized; aligned and secured. All that really remains is to mount the Z-axis spindle plates, and install a few switches, route cables, and other boring but necessary chores. After these were done, it was time for a simple test!


Front view, spindle backplate
The Z-axis spindle backplate was secured to the NSK bearing blocks with eight 8mm cap screws, with the orientation correct and the master register rim hard against the master block. This is a relatively heavy piece of the spindle assembly - the thickness will both dampen vibrations somewhat and will also suspend the spindle centerline farther out over the table, needed by this particular design.

Once the plate was in place, the Z-axis ballnut carrier was secured to both the ballnut and the backplate with the same method used on the other ballscrew systems, namely a progressive tightening of the various cap screws so as to avoid stressing the ballscrew.

The first spindle to be mounted for tests was the Sherline industrial ER16 system, mounted on its own carrier block. This spindle will be fine for getting the mill going for basic functionality tests, but it will lack the speed and horsepower for what I'd ultimately like to do.

I'd love to get my hands on a CAT30 box spindle of perhaps 1 hp. Automatic tool changes, and correct Z offset!

A view of the business end of the X-axis. I will probably have to create protective swarf plates for the currently open coupling system, or shield it in some fashion, as metallic swarf will jam inside the oldham coupler and cause grief.

Ultimately, I am very pleased with the appearance and fit of the X-axis, and consider it to be the best part of the mill in terms of design and quality of execution. The iron work table, cannibalized from the cheap X-Y mill-drill table, has worked out perfectly so far.

Note the 3 sets of limit switch wires exiting the mill upper right. These are simply 6-wire phone line with mini-DIN couplers.

A close view of the Z-axis limit switch.

I originally had the switch mounted on the Z-axis ballscrew backplate, and it was tripped by the ballnut at the extreme upper limit of its travel. I soon realized that an adjustable Z switch would be a nice thing to have, as my Z column is relatively tall for a mill this size, and 90% of the time I will not need to home all the way to the top of the machine.

At 6:00 in the picture is the right-hand NSK linear bearing block. The tan band is an oil-bearing polymer wipe which is a dynamite protective device for the NSK truck.

Note how the NSK block body trips the switch (black, behind the slotted aluminum strip), and how the aluminum switch carrier itself can be adjusted in Z for perhaps 5". Here, it is as high as it can go.

The spindle backplate is to the left.

A few installments back (I forgot which!), I mentioned that some Z-axis mockups showed me that the weight of the spindle assembly (spindle, spindle block, backplate, and the 2 NSK bearing trucks) was going to be more than I anticipated, and sure enough, with the power off, the spindle freewheeled downward until it contacted the table. I tried to install a braked motor, and while that stopped the freewheeling, the sheer effort of lugging the spindle up and down would wear the system quickly, heat the motor, and generally cause grief. I needed a countering force.

I messed a bit with coil springs, but hated their kludgy appearance. I decided a gas spring would do, but all of the springs I found with a 10" throw were over 100 lbs force! I needed one with ~30 lb force, and a stroke of >10". I finally did find an answer... this is an adjustable gas spring, meaning I can bleed off nitrogen from the body at that brass ring in the picture. The stroke is 11.5", and the system is set so that the Z-axis movement is in the middle of the spring's travel, avoiding the excessively oil-damped ends of the travel. This system works great. As far as the motor can tell, it may as well be horizontal. Equal forces up and down.

Not too expensive at around $30. It looks decent too, partially hidden behind the left side of the spindle backplate. The bottom of the gas spring shaft was rounded with a grinder, and set in a modified cap screw on a bracket like a golf ball on a tee. The lower bracket is attached to a pair of T slots near the bottom of the mill base.

Overall view of the X and Y axis. The aluminum angle on the front of the table will hold a sheet-metal swarf guard.

I should have though harder about swarf protection and the routing of the cables. The X-axis is perfect, totally enclosed, but the Y and especially the Z are both exposed. I also need a proper machine base in a tray form, a flood coolant system, and ultimately some type of cabinet to contain the mess.

And the entire mill as it now exists. I hate this picture because of the background clutter and lack of good detail.

As of this posting (10 Dec 2003) I have verified the performance of the mill with some nice CNC work. The mill is very tight, very fast, accurate. The spindle will probably be the limiting factor. I have some installments coming up in which a portion of the KAVO spindle holder is CNC milled using the Sherline spindle (the mill is helping to build itself! Shades of Isaac Asimov), and they will detail the entire process from a Rhino model, to Visual Mill for the toolpath, all the way through to the cut. Hopefully it'll be pretty cool.

  But before all that happened, I exercised the system with a simple engraving on a scrap of aluminum. Using DeskEngrave (free at DesKam.Com) I created a simple toolpath shown here, and imported it into Flashcut.

A dremel engraving bit was chucked in an ER16 collet, and the spindle turned on full speed. The result is shown here.