Denford Easimill restoration and upgrades

[Jonathan]

A couple of years ago I bought a Denford "Easimill 3" and used it for a little while in student accommodation before putting it into storage. I've now moved it into my own place so it's time to get it working properly. I'm no audiophile so I already flogged the originals controller complete with g-code on magnetic tapes. Apart from needing a good clean up and an unfriendly amount of backlash on Y, there's not much wrong with it.

This is how the machine looked when I bought it:



It's quite a nice size machine, somewhere in between a desktop "hobby" mill and a standard e.g. bridgeport size. Split into pieces it's not terribly difficult to transport, so for now it's a pile of parts and will remain so while I fix the main limitation - Z-stepper motor driving the quill not the knee.

I planned this slightly badly by lending my working CNC mill to a friend last week, so I'm now having to make do with my lathe and CNC router, which are 90 miles from the milling machine...I've there dismantled the X&Y axes, measured up and taken some parts home so I can replace the trapezoidal screw that lifted the knee with a ballscrew.

After some accurate scribbles...



I machines these few parts to mount a Nema 34 size motor inside the knee:



Which fit as shown (using the Y-ballscrew to test):



I've replaced the 6204 bearings in the knee with 7204BEP angular contact bearings. Not as nice as the 60 degree contact angle bearings on X and Y, but a lot more affordable so they'll do for now. I'm sticking with trapezoidal belts like on X and Y, as I feel they'll be best for low backlash. Found a local shop for the belts and bearings who were within 10% of the price online and more importantly open.

Here's a little part I made today for the ballnut, which will affix to the base:



I cleaned up the Y-axis ballscrew and bearings to find that the source of the backlash was endfloat on the screw due to a loose locknut, so that wasn't very interesting. The Y ballnut feels smooth, but doesn't feel like there's any preload. This was confirmed when I reassembled the axis for a quick check and measured 0.01mm of backlash - better than the 0.2mm before but I'm hoping to improve it with an NSK ballscrew I've got on the way from eBay.



Here's some general photos for interest:



Now I'm shopping for ballscrews for the knee. I've sized up based on a 32mm ballscrew for Z, as that seems to give a reasonable compromise for stiffness, at least according to the script I posted here a few years ago:


Denford_motor_sizing.zip

The important values here are the estimated axis stiffnesses, based just on the combined ballscrew nut and bearing stiffnesses - not the machine structure. The X and Y ballscrews are 25mm. This makes Z a similar stiffness to Y, which I think is acceptable since although Z can see very high forces while drilling the accuracy is less important.

Here's a simple drawing of the ballscrew I need:



I've been making inquiries with some of the many different sellers on Aliexpress:



I used a C5 grade TBI one on my X3 conversion, and the preloaded double nut with zero backlash was very pleasing, but seems a bit of an indulgence here since the weight of the table itself will do a reasonable job of preloading the nut. They also seem to have gone up in price significantly. I'm after something better than the lowest quality ones though. The seller in bold seems to be saying there's are better quality - steel instead of plastic inserts. Also seems to be a manufacturer not a distributor. Has anyone tried these suppliers, except BST automation who seems to be a popular choice. He's suspiciously cheap compared with the rest.