Surface grinder spindle bearing question

I am working on my Boyar Schultz surface grinder. The spindle bearings are a normal ball bearing at the nose, and a smaller angular contact at the rear with a sprung spacer setting the prelude on the angular contact. Both bearings are Barden.

Is there a technical reason for using the radial bearing at the nose vs two angular contact bearings? Do angular contacts only function properly if they are the same size? Was it a cost saving measure?

I'd say you have the typical basic spindle setup of an angular contact pair with a floating radial to allow for spindle growth. The angular contacts are geared more toward thrust loads and overall stiffness, whereas the radial will aid in the handling of the radial loads, which on a grinder are plenty.

I have an old Boyer Schultz myself. 12 X 18 hydraulic. Been ages since I've looked in the books about it.

EDIT:

Just looked in my books. My spindle is angular contact DB pair at the front and radial at the rear. Not seeing a spring anywhere. The radial is a bit smaller then the angular pair. Not uncommon I would say. The radial is fully floating axially.

This seems backwards.
Are you sure some other owner did not do things?
Front should be a preloaded pair mounted take your pick. The back end should be standard an allowed to float as the shaft expands and contracts during use.

It does seem a bit odd, but that is the design. It is a BS 612, their small grinder, that is why I asked about cost as reason. Here is the drawing showing the radial bearing at the front. This is from the later model drawing, it mentions the dot on the race, which is normal for high precision angular, but I have never seen a radial bearing with the high point dot. My bearings are Barden 205SS (SS=stainless steel shields) front and 204H (H=regular 15 degree angular) rear, neither of which has a dot.



You can see the spring arrangement pressing the outer race of the angular at the rear. As the spindle expands the springs hold constant preload on the outer race of the angular. (if we assume that over the infinitesimal length of expansion the spring force remains constant)

Wow... see something new every day. For a spindle with only two bearings, there sure are a lot of parts to that assembly.