16C Collet Chuck

So... I have a CNC lathe that takes 16C collets, and it would be great to use the same collets on my DoAll 1340 lathe, but Bison 16C collet chucks are expen$$$ive - around $1400 US which with the current US-to-CAD exchange rate is about my annual salary. So I got it into my head that I would make my own collet chuck instead. Now it's not going to be as fancy as a Bison 16C scroll chuck, but with some collet closer ingenuity it should do nearly as well. I picked up this piece of 6" diam x 5.5" long 4140PH from my crack-deal.. er, metal supplier, and chucked it up in the four-jaw. The DoAll 1340 lathe has a D1-4 camlock spindle nose, so the smaller piece of 3/4 x 10 is for the camlock studs.

After finding some D1-4 spindle-nose specifications here, I modeled everything in my Alibre CAD program. From the model, I produced a couple of drawings, namely a cross-sectional view with dimensions, and a bottom-up view of the drill-hole positions for the studs and lock-screws. Everything except the 16C collet model is mine. Whoever drafted the free 16C collet model made it .002 larger at the taper than it should be, proving that indeed you get what you pay for.

The D1-4 spindle-nose has a 7.125 degree taper from centerline, and clearly there's no way I'm going to hit this simply by looking at the divisions on my compound... but I'll tell you, when I eyeballed the angle at about an eighth over the mark and traversed the compound with an indicator, I got 0.0623 per 1/2" of travel on the very first try. A bit more indicating while banging the compound around brought it to exactly 0.0625 per 1/2", or 7.125 degrees.

I locked down the carriage and used the auto cross-feed to put a nice finish on the face. The point of the parallel is to shine a flashlight behind it and look for any light coming through at the interface. Any light and it's concave or convex. Not even a glimmer of blue light (0.0001 gap) means it's bang-on.

So on to drilling, and drilling, and drilling. Drilling is really the best way to remove metal. Since I bought my lathe at the beginning of 2016, I have accumulated a lot of large MT3 and MT4 drills - they're incredibly efficient at removing material, and I also started collecting extra-long drills too. Unlike jobbers, extra long drills save me from having to flip the part and drill from the other side, resulting in the inevitable two wobbly offset holes meeting in the center.

Next step is to bore out the back end in front of the spindle where the collet-closer drawbar threads will pick up the end of the collet. The 16C collet handles up to 1-5/8, but my spindle bore at 1.66 minus the wall thickness for a drawtube, I end up with a max 1.4 ID. It's not ideal, but this project cost me all of $70 in material, so I'll take it.

With the compound at 7.125, I bored out the taper and snuck up on 2.5000 (+.0005 / -0) at the sharp edge, then broke the edge.


Now to flip the material around and square up the other end before drilling and tapping the camlock stud holes.

The finish is hit and miss... 4140 gets nice and shiny with deep cuts with these i-dunno-what-brand inserts I'm using, but sucks on the light passes. Doesn't matter anyway. All that material will be turned away to form the collet chuck nose.

Putting it on the mill, I'll find center and using the hole drawing from earlier, I'll drill, counterbore, and tap six holes.

This is what I'm using for reference (I later found an actual drawing but had to wait for my membership on a forum to be approved so I could read it.) Doesn't matter though. None of the dimensions are critical. These are M12-1.0 threads, 3/4" long, the body is ~1.4ish long, and the big semi-circular groove is 3/4 diam placed 0.44 off the centerline. Same with the 3/8" milled slot that will fit a socket-head capscrew that stops the stud from rotating.

I wanted to use M12-1.0, but didn't have a 1.0 pitch matching tap, so I went with 7/16-20 instead.
Turning to .624"

Tapping threads. I need a better die holder solution to keep junk out of the tailstock bore.

Side by side comparison - all the length dimensions were eyeballed by holding the original part beside the chucked material and bringing the cutter up against the surfaces... like a copy lathe, sort of.

So far, so good. Next step is to mill the 3/4 radius. I probably should have milled the 3/4 semi-circular cut before putting on threads because holding these little buggers proved troublesome. I put one in a 5C collet block and started milling, but the cutter pulled the part out of the bore and broke my end mill. I'll fix this a little later, but for now, let's get back to drilling and tapping the chuck body.

Presto. Good fit, but the studs could have been longer to put more thread in those holes.

Zeroing off the end. The center of this curved slot is .625 from the end of the stud.

Making sure it's 3/4" before I start cutting... This random endmill seems to fit nicely :-)

Success. Ok, so this is the idea: When I previously milled the stud while only being held by the collet, the endmill yanked the part out and broke the endmill in the process. So, that little piece of A36 that has been turned and tapped will screw onto the back of the stud after the stud is mounted in the collet holder. Once the collet is tightened, this extra bit will screw up to the rear of the inside of the collet and prevent the stud from pulling out. I also milled using the full side of the cutter rather than plunging down. It squeaked and squealed in the final depths of the cut, but it worked.

Oh man! Looking for a 3/8 ball end mill, and though I have a million end mills, there's not a single ball end in the lot. That shallow 3/8 radiused groove for the screw stops will have to wait.

I tried mounting it on the lathe, but I was missing an important feature. Those 3/4 slots aren't what makes the camlock cinch up on the stud. Toward the top of that curve, there is a little 45-degree cut that makes that curved cutout look more like a J - it is the 45 degree cut (the straight part of the J) that the camlock pins turn up against. I'll drill and tap for the lock screws later.

*Sniff* ... not there yet.

So this is my home-brew collet chuck for mounting 5C stuff in my Nikken tilting/rotary table. Bottom right: The collet holder with a .625 5C collet inside, then counterclockwise: the "drawbar" that grabs the threads, the A36 tool I spoke of earlier now turned to fit up through the bore of this new setup, then a washer and collar with 1/4" holes for tightening. The carbide bit's shank is used to tighten this fixture until I make a tool.

Sorry about the pictures. I didn't really take any shots until afterward. This shot is showing the stud held in a 5C collet which is installed in the table with the B axis tilted to 45 degrees. The cutter milled across the top of the curve to give it that 45 bevel needed for the spindle nose to pull against the stud.

The three studs now have the 45's on them. I smoothed the transitions with a file then re-installed them in the chuck body.

This time, they worked perfectly. The camlock sits exactly midway between the two V's on the spindle nose and cinches up very nicely.

Next steps will be to bore out to 1.89" for the 16C collet body, carve out a 20 degree taper, turn down the body, then drill/tap at least one hole for an alignment guide pin. I am also thinking about leaving some meat around the nose for a thread, or possibly some step-chuck closer rings of some kind... maybe.

More to come...