Then the motor itself had to come out because it causes self-induction when spun by my stepper motor, and this makes the rotation less smooth.

And finally the bearing had to go. This involved an arbor press and a ball bearing to focus the force (and a hemostat to hold the ball bearing, which didn't actually work the first time — it's out there somewhere).

Now it was time to make mount pads for the stepper motor and servo assembly. I felt like a tornado cutting through a trailer park as these little chips and things got in the path of this huge end mill.

The original hole needed to match the mount ring on the stepper motor.

And the original screw holes didn't align, of course.

At this point, it looks like this, with the stepper motor firmly attached.

The arm of the read/write head needed some rocket surgery. Only the top head is being used because it's the only one visible, so the others were eliminated.

Now it's time to make standoff posts for the little servo that drives the arm. It's not designed to be mounted upside down, so I had to get creative. Here's trimming down 1/4" fiberglass rod to about 3/8" long. It's a carbide bit because fiberglass is rough on high-speed steel. This is the "vise in a vise" trick, where the little guy is sideways so the process is repeatable across all four standoffs.

And there needs to be a hole for the mounting screw.

All four ended up like this.

And they attach this way.

The servo can now be mounted quite nicely.

The head assembly needed a small hole for the drive shaft. I broke three bits doing this. Not only is that a waste, but they each got stuck in the assembly, which meant I had to drill the next one slightly off target.