Evening all, Im rebuilding a spindle for a precision drill / sort of jig bore. Its 'home made' from a SIP Inspection Microscope. Pic here:
current spindle has a drill chuck on the end, which I intend to replace with ER16 collet direct in the spindle. The current spindle has a ball thrust, then a single row bearing at the nose, and another one at the tail. The bearings are just normal run of the mill ones.
The spindle cartridge is 30mm OD, and the ER spindle needs a shaft dia of about 15mm. This seems (from a quick poke around at the AET website) to rule out the use of angular contact/taper rollers.
Given I wont be milling on this, only precision hole locating is the existing bearing arrangement a good idea, or could it be improved upon easily?
You could consider getting a normal ball bearing with a C5 (larger than normal) clearance and using it as if it were an angular contact bearing. For a 15mm shaft the clearance is quoted as 25um vs the standard of 5um. It's not quite as good as an angular contact and the geometry is not perfect, but it is a servicable route to a poor-man's angular contact solution and is noticably smaller OD.
The only likely looking bearing I can see listed is a 61802 which is
15x24x5. A more common 6002 is 32OD so you would run out of meat.
Other alternatives could be to use a standard ball bearing with reduced internal clarance, but you would need to ensure the adjustment of the thrust bearing was finely adjustable or ensure the ball bearing can float so that there would be negligible end thrust on the zero clearance ball. A small end thrust on a zero clearance ball bearing gives large radial loads which might be excessive. Also you might use the existing thrust bearing and a large-ish needle roller for the radial location - very stiff, nil clearance, high load capacity. An NKJ15 has a 15mm bore with an inner ring and an OD of 27. If you could run the needles direct on the shaft you could reduce the OD to 23 with an NJ15, but you would need to harden and grind the spindle.
Depends on how much you like tinkering, I'd probaby give it a go as is and consider the above as a fall back if it proves a problem.