My rotary broach project left me wanting a sensitive drill to drill very small holes on stuff in my lathe. So for the past few months, I’ve been working on making one, using the old spindle from my CNC 2.5D router, from before I upgraded the router to a three-phase spindle. Here it is at full 50mm extension:
And retracted, from the back:
(The fan on the back looks ugly from running into some red tape while it was running on the router.)
Prologue
I first tried what looked like an easy solution, 3D printing an adapter to use my dremel in a toolpost holder.
However, it just wasn’t stiff enough. It bent with the slightest pressure. I’d been thinking of making a toolpost grinder all the way back last year when I was Building an Electronic Lead Screw for my G0709 lathe — and even put some of the electronics together at that point. That meant I’d just be completing a project, not starting another one, if I made this project! 
Mechanical build
I iterated on and off for some weeks on a design in FreeCAD. Ultimately I discovered that I didn’t think enough about the assembly process, but assembly is only awkward in places, not impossible.
As I started, I thought it would be helpful to chain-drill near the edges of the pockets for the MGN9 linear rail carriages. I ultimately roughed it out with a larger end mill and the chain drilling was just a distraction. I wouldn’t do it again.
But ultimately I was able to fit the carriages and rails into the frame:
At some point I realized that while actually assembling this, I was going to need some “keepers” for the balls in the linear rails, and this time 3d printing was my friend. In order to assemble it, I actually needed four smaller ones, because I had to remove them one at a time on each rail.
The black screws in the face of this go into the carriages on the other side:
The rails mounted, and here the screws that will hold this to the steel block that goes into the toolpost holder are inserted to show where they eventually go. The slot to the left of the screws delimits the 50mm of travel and there’s a screw that rides in that slot.
Screws stick out through the slot that is keyed to the steel bar that holds this into the QCTP holder. The countersunk holes are for screws that hold the two sliding halves fixed when this is not being used as a sensitive drill; in some of the pictures you can see the threaded holes the screws go into. (When those screws are removed to allow the spindle to slide back and forth, the screws go into “keeper” holes on top of the spindle clamps, where you can see them in the first pictures in this post.) The teal pieces are the 3d printed MGN9 “keepers” holding the recirculating balls in place.
Most of this is aluminum, but the part that goes into the tool holder is mild steel. The slot is an interference fit for the steel bar.
Two holes in this pieces are just for tightening the screws holding the linear rail in place with all the carriages installed on the rail! I actually used these to tighten screws at the front end, tightened the ones at the back with the frame slide forward, then took everything apart, tightened all the rest, and then assembled it again! There is 0.3mm between the two plates. It’s not really visible, won’t let swarf enter and get onto the rail, yet they don’t rub at all.
With the spindle installed, metric scale for banana:
Once I went to install it, I discovered that I hadn’t accounted for the tool holder height adjustment wheel. 
So I bored out a bit of a recess to fit it in:
Electrical
The printer I used for this test print of the panel was a little off in Z, so the surface looked bad, but it was enough to test-fit parts:
(That was before I selected a different connector for the spindle.)
The wiring is all ferrules and crimp connectors of various sorts, and the pictures just look like spaghetti. The basic idea is:
- One high-voltage four-conductor cable to the green momentary remote switch for the contactor; one pair pulls the signal input up to 120V, the other is connected to the 240V power to the spindle power supply and glows if the spindle power supply has power. This connects through a liquid-tight four-conductor latching plug.
- One six-conductor low-voltage cable, of which three wires are connected to the speed-setting potentiometer, and two connect the switch above the green button to the enable port on the spindle power supply. This connects through a normal GX16-8 aviation connector (8-pin because I installed it before I worked out my plan and it was easier to leave it than to change).
- One two-conductor cable for the DC spindle, which goes through the DPDT on-off-on switch at the upper left to change direction. (The spindle needs to turn forwards to drill, but backwards to grind against a turning piece. Both the connection from the console to the power box and the connection from the spindles to the console use a liquid-tight two-conductor latching plug.
None of the three connections uses the same type of plug, so it’s not possible to accidentally hook it up wrong. That was by design.
I finally turned my pencil sketch of a subway map circuit diagram for the accessory electronics box into an inkscape diagram of my electronic lead screw circuit and toolpost spindle circuit combined, printed it, laminated it, and put it into the box.
Lathe fully powered down, with the spindle not connected and the splash cover in place, here’s the new console, sticking up above the backsplash and easy to use.
Spindle plugged in:
Lathe powered on, with the clough42 electronic lead screw console lit up, but the spindle power supply not powered up:
Spindle power supply powered on, the green button is very obviously bright:
I’m still thinking about using white letter on clear or black label tape to put labels on the console, but… it all seems really obvious to me. ON/OFF is already labeled for the switch that tells the spindle power supply to actually turn on the motor. The potentiometer works like a normal potentiometer, counterclockwise to go slower and clockwise to go faster. The switch above the potentiometer points towards the lathe headstock to “drill into the stock” (normal rotation) and towards the tail to run backwards.
In Action
Here’s my first test drill into a piece of scrap ¼" brass:
Epilogue
Of course, in the end I forgot that I started this because I broke a drill bit off inside my 4mm broach, saw that I had started a hole, and decided to finish the job with my new drill, which worked great until I hit the fragment of drill bit and broke yet another drill bit!
Having four carriages on two rails to act like a virtual very large linear rail was an interesting idea. However, it’s sufficiently imperfect that it has a bit more friction than I had hoped while sliding. It’s still good enough, but could be smoother.
This is not a super high quality spindle, and I haven’t measured the runout—at least since it was on the CNC router years ago and I’ve forgotten since then. I haven’t tried using it as an actual toolpost grinder yet.
I wish that the latching contactor would forget its state when it lost power. I understand why this one doesn’t — it’s really meant to remember a brightness setting as a remote three-way light switch — but I want one that always starts off, and a momentary press of the switch toggles state. But it doesn’t matter very much.
While I was working on this, I started some threads on Mastodon:
I may later make another set of spindle clamps and screw them into the side of the head of my cute little CNC-converted HF mini-mill. I think that the power supply on the mill could run this spindle instead of the much slower motor, and I could just have a switch between the built-in motor and yet another 2-connector liquid-tight plug and just plug the spindle in there to have a high-speed spindle as well as a low-speed spindle on the CNC mill.