A nice new MT3 blank in the spindle.

When I removed the existing chuck, one of the studs unscrewed from the chuck not the nut. Irritating. The nut proved so tight, I had to make a split nut to clamp the free piece of stud in vice grips to get it undone.

Right, now I was ready to attack the Taig chuck. A little patience soon had the chuck mounted backwards onto the arbor and running true to within a 0.0005 TIR.
 

This was definitely the scary part. After all, this was the first dimensionally critical piece of turning I had ever done and I was a little unsure. The back of the chuck was fairly quickly recessed to 4mm deep and 50mm diameter. Whatever it is made of, it cuts nicely.

With most of the metal removed, I went as carefully as I could with much pausing and figuring of handwheel movements until the template just fitted.

In spite of my best endeavours, the recess was too big by about 0.001. Not bad I guess but not what I had hoped for. There was definite movement in the template even though it would not drop out of its own accord.

Scary part done, I moved on to what should have been the easy part - fitting the studs. I really ought to have known better...

Digging around in the scrap box revealed a suitable bit of steel rod. This was soon converted into four pins sized to fit into the spindle holes with a small conical pip at one end. With the pins in the plate, the chuck was fitted over the spindle A bit of scrap behind the pins and a sharp tap with a brass hammer (actually a lump of bar stock) soon had locating spots on the chuck.

There was no need to harden the pins. They are one-shot disposable items.

While cutting the pins to length, I wanted a quick way to reliably move the saddle back and forth by known amounts.

I discovered that my six-inch rule just sits on the bed behind the V-way without fouling the saddle. It is no trouble to read off movement quite accurately and, of course, you have an easily resettable zero.

The simplest things are so pleasing

Now things started to go a bit pear-shaped. I had a drill stand for my hand drill and used that to make the first stud hole. I was careful, mindful of the lack of rigidity, centre drilled and went up in small steps. However, I was too stupid to check that the damned thing was drilling perpendicular to the table. My beatifully inserted stud was clearly at an angle other than 90 degrees and I could not see how to get the rest of the holes in.
 

After a bit of a break, I decided to mount the new chuck on a bit of 5/8the steel bar clamped in the toolpost. I was not certain that it would be sturdy enough but it worked like a dream. My drilling was all done from the chuck with bits just held in the 3-jaw.

Here you see the centre drill hunting for its next victim.

A bit of brass scrap with a convenient hole made a fine support for the back end of the tap to hold it square and I soon had a suitable hole.

In the UK, the studs are M6. Everything except the leadscrew and power supply is metric on mine. No problem here and, although it is not convenient in many (largely historical) ways, there is a certain simplicity to the metric system.

Once the stud were in, that was it - job done. The wobbly stud was a bit of an embarassment. In the end I had to use three of the holes from the 90 degree set and one from the 120 degree set. Bit of a bodge, but balance is not much of an issue since the 4-jaw is used specifically to hold stuff eccentrically.

Possibly as a result of the screw up, when the new chuck is mounted, I can detect no movement against the backplate by hand. Once tightened down, TIR is 0.0015. I guess I will live happily with that since it is hardly critical for a 4-jaw anyway so long as it cannot move against the spindle in use. I am fairly sure it can't - a tap with the hammer will soon find out.

Overall, I am pretty pleased with the job. The way I have done it may not be the best but it is pretty simple and taught me a lot on the way. What more could you ask?