SIEG X2 Mini Mill CNC Build - Introduction

Sunday, April 9, 2017
Some of you might remember that several years ago I used to own a Harbor Freight Mini Mill, AKA SIEG X2. At one point I upgraded the X-Y table to the LMS large table assembly and the old small table has been collecting dust under my workbench. I couldn't force myself to throw it away but couldn't find a use for it either. Until a few weeks ago, that is, until I stumbled across a series of three videos by Stefan Gotteswinter in which he rebuilds a Chinese angle plate. In the introduction to the first video he explains that he bought it as a "set of castings" and proceeds to completely re-machine the whole thing. This gave me an idea...
I've been thinking about getting a small mill (Taig or Sherline) and converting it to CNC. While either of those machines would be of much better quality than a standard SIEG X2, the latter has two advantages: it's made out of cast iron (weighing 2x as much as a Taig and almost 4x as much as a Sherline mill) and uses R8 spindle, same as all my mill tooling.
That said, out of the box SIEG Mini Mill has several annoying issues and generally poor finish, but with some patience and know-how it can be made into a decent machine and produce very nice and accurate results.

The Plan

Stefan's video gave me an idea - why not build a CNC machine out SIEG X2 castings. This might sound a bit radical, but having owned a mini mill, I know pretty it's shortcomings. A few problems are caused by questionable design, such as the infamous column droop caused by excessive backlash in the Z axis' fine feed and column rigidity issues caused by the column tilt "feature". Most of the other issues stem from poor fit and finish, which can be relatively easily improved. All of the parts fit the working envelope of my SX4 mill and 1024 lathe; the rest can be addressed by lapping, sanding, scraping and filing.

Quick Disclaimer

This project is not designed to promote or bash SIEG X2 mini mill (or import machinery in general). My main motivation for the project is to have fun, hopefully learn something and share the things I learn. I'm also well aware that this is completely impractical, but neither is building model steam engines, rebuilding an old metal shaper, or a number of other things I've done. Now that I think about it, the set of castings is actually cheaper than Stuart Models Twin Victoria…
With the above in mind, my plan is as follows:
  • Finish the cast iron parts, fix the geometry where needed, scrape the ways, etc. and build up the bare machine frame.
  • Install ball screws on all three axes; make whatever brackets, mounts, etc. as needed.
  • Add stepper motors and electronics, and setup the controller using LinuxCNC.
There are several improvements I want to make to address some design issues and generally improve the mill.

Get Rid of Column Pivot

SIEG X2 comes with the peculiar column tilt feature which I never had a use for. The drawback is that the column is attached to the base by the pivot bolt. Making matters worse, the bolt is tightened over a hollow column, so quite naturally, column rigidity suffers. I will address this problem by skipping the pivot altogether and replacing it with a solid column bracket. The bracket will take care of deflection along the X axis. To add rigidity along the Y axis I will add a brace on the back of the column that will be bolted to the mill's base. If done properly, those two things will minimize or even eliminate column rigidity issues.

Improve the Spindle

There are three issues with the stock spindle configuration on the Mini Mill.
  • First and foremost, SIEG uses low quality radial bearings which aren't appropriate for this application since they can only withstand radial forces.
  • Second, drive gears in the X2 header are plastic and besides being rather noisy, are prone to damage. While the replacement parts are relatively inexpensive, the process is a major pain.
  • Third, top speed for a stock SIEG X2 is 2500 RPM, which is pretty low for milling aluminum with small cutters.
To improve the spindle, the cheap radial bearings will be replaced with high precision angular contact ball bearings or tapered roller bearings. Better bearing will minimize the spindle runout and will have higher RPM rating. To take advantage of the higher RPM rating I will make a belt drive kit and use it with a 6700 RPM treadmill motor.

Add Thrust Bearings to Y Axis

For some reason SIEG decided to use thrust bearings only on the X axis lead screw, making backlash and/or binding all but inevitable on the X axis. Since I will be using anti backlash ball screws to drive the axes, proper axial bearings are a must. This means that I will need to add provisions for thrust bearings to the Y axis stepper motor mount and an adapter to couple the ball nut to the saddle.
No axial bearings on the Y axis is far from ideal

Improve Way Lubrication

The only way to lubricate the ways on the mini mill is to wipe them with oil every so often; there are no oilers anywhere on the machine. This is far from ideal, especially for a CNC machine. At the very least I will drill some oil passages in the saddle and the head mounting for oiling; ideally, though, I would like to add a one-shot lube system to the mill.


This is the first post in what will likely be a long series of posts (and potentially videos) about me building a CNC mill out of a SIEG X2. Rather than threating this as a CNC conversion, I decided to approach the project as if I'm building a new machine out of a set of castings. While it's to my main goal to save money, the fact that I already have some parts will save me a good chunk of cost of new mill. At the end of this endeavour I will hopefully get a decent quality small CNC milling machine capable of making small parts for my other projects.


  1. this will be fun to watch. Are you planing on using the touch dro to know location?

    1. I don't know yet, frankly. I'm leaning towards "C7" screws, and those aren't extremely accurate. I'll need to dig into it a bit, but there might be a way to use TouchDRO with glass scales as a feedback loop (or may be just the glass scales).