RE pulleys, many different suppliers have them. I've just looked at McMaster Carr for design purposes to see their dimensions. Before beginning the project I would need to find who has the best deal. As I've mentioned, for me this project is behind a few other things in the queue, so I won't expect to get started on it for quite a few months.
RE tensioning, I intend to build a new housing for this spindle, rather than trying to reuse the old one. I'll be milling some slots for the motor so it can slide to provide that tension. I could also mount the motor on a plate in a slideway with a threaded bolt to provide the tension. We will see.
One design issue I'm still pondering is speed ranges. I would like to get 6K to 8K rpm for aluminum, but steel wants to run a lot slower. If I cover that whole range with just the VFD, there is not going to be much torque available for steel. It seems inevitable that the system wants a Hi/Lo range for best performance over a broad range of spindle speeds. Still contemplating the best way to go about providing that.
I may investigate a tumbler back gear arrangement at the motor end that can halve the motor speed for cutting tougher materials. I hate the thought of changing pulleys, but that is another simpler possibility.
If any of us actually get the silly thing to work well for $300 in bearings and a $65 spindle, it will be a lifetime achievement award and very cheap for a decent spindle. Try not to think too hard about what it'll be like if you get the preload wrong and burnup a few sets of those bearing--Doh!
Spindles have to be one of the more finicky and difficult projects one could tackle.
Best,
BW