Bridgeport Manual to CNC conversion...

[Muzzer]

IIRC, there were oil pipes for each of the X and Y acme nuts and I was able to reinsert them into the ballnuts. The X can be fitted when the table is still off but I fitted the Y with the table and saddle in place (forget why) and I recall a little bit of messing about inserting the pipe into its ballnut, via the hole in the front of the knee.

Nice solution, getting a spare head casting. I wasn't impressed by the crude approach used by Elrod, South West Industries etc that relied on a Z axis yoke sticking way out from the quill to clear the features on the standard head. When combined their massive castings (you could have kept a rabbit inside the housing Elrod "designed" and supplied for his conversions), the end result is enormous and expensive - but not even very robust around the ballnut and quill.

I designed my own Z axis drive using a 16mm ballscrew that replaces the threaded feed adjustment rod. This allows a ballnut yoke with a minimal overhang and the installation is compact and neat.

Unless you plan to do a lot of heavy drilling, you don't need a big motor for the Z. As a starting point, if you ignore friction, the torque and force are related like this:
Torque x angle = force x distance where angle is number of radians in one turn (2*pi), distance is the ballscrew pitch (metres) and the torque (Nm) / force (N) are your input and output.

So for my 2Nm Leadshine closed loop stepper and 4mm ballscrew, the stall force would be ~2 x 2 x 3.14 / 0.004 = ~3000N, which is equivalent to the weight of about 300kg (or three fatties). That's a lot of force and would be a test for most ballnut installations. Any pulley reduction ratio would increase that further - in fact mine is 1.8:1, resulting in about 1/2 tonne of force.

In practice, you lose a fair bit of your motor torque along the way, especially if you don't align your ballscrew carefully, but even so, you can see that a lot of these conversions you read about are stupidly overspecified in terms of stall torque, reduction ratio etc. There's one good reason for fitting limit switches...

Keep us updated on your progress!

EDIT - missed out a factor of 2 on the left side of the fag packet sums. Answer was right, though.