I have a 4th axis of sorts for my mill via a rotary table which I had CNC converted (IIRC, I got the cheapest rotary table in the store; at the same time also getting a drill-press vise to use for holding stuff, and made some aluminum smooth jaws for it).

For this, had used a 69 oz-in NEMA17 connected to the handle assembly via a belt drive.
Seems fast enough, and seems to still have a fair bit of torque, and have it set to speed of around 720 degrees/minute, with an approx 90:1 reduction (approx 4 degrees per revolution). AFAIK, internally I think it is a worm gear or similar. Main failure modes are generally belt related (like the belt coming off the pulleys or slipping under sufficient load). Had machined a pulley both to go on the handle assembly and also a similar pulley for the motor.

It is actually capable of going a bit faster than 720/min, but this tends to make it not hold position as well, and makes it more frequent for the drive belt to jump the pulleys. There is an adjustment screw that reduces these issues though (by keeping the motor aligned with the handle pulley). Technically this is using a repurposed vacuum cleaner drive-belt with some strategic use of knurling (a proper timing belt and pulleys would probably have been better).


I was basically unable to mount hardware directly to the rotary table, as the HSS tools I have were pretty much entirely ineffective against whatever type of cast iron it was made out of. Best I could ultimately do was sand the paint off in that area and then JB-weld on another piece of steel which I had modified into a mounting brace for a NEMA17 motor bracket. Could also have possibly JB-welded on a piece of steel plate and then machine it on the table (to face and drill/tap some mounting holes).

But, yeah, I suspect a NEMA23 is probably overkill if doing something similar to this (more so, this isn't even the largest size of NEMA17's which exist).

Overall, it seems to basically work (excluding not having good mounting hardware for mounting stuff to it).