After reassembling the taper and using the machine later, I discovered there was a bit of spring back in the cross-slide. I isolated this to two things. One was the little taper pin which anchors the slide. It had a bit of rocking motion. I was able to re-lap this pin and replace the lock square bolt with an allen bolt. This held the assembly rigidly but did not completely solve the issue.
Further investigation turned up the two bearings that secure the cross-feed screw (which you can see in the picture). These had a little bit of spring to them. I believe I read somewhere in one of the brochures or manuals that these two bearings were duplexed. I replaced these bearings with a couple of radial bearings and made shims for the outer race to induce some pre-load and increase the stiffness. This worked to some degree, but it still had .002 of movement in the cross-feed screw. If I increased the size of the shims, it would increase the stiffness but increase drag. I wasn’t happy with this.
So what I did was to use a double row angular contact bearing. This bearing has no movement radial or axial. I did have to make a bushing and some collars to adapt to the screw. This bearing is held in place in the slide by a screw-in collar which holds it securely. This was finally successful, and movement was reduced to a few .0001’s with the greatest force I could pull on it.
This got rid of any spring-back when threading or making heavy cuts. It made the lathe more predictable. What I’m unsure of is if the bearings that was in it were original or if someone had replaced them along the way. Without a parts manual it was hard to know what Pratt and Whitney had used here. If you have this type of taper, this is a modification I would recommend.
The last picture is of the angular contact bearing I used.
I’m not sure why Pratt and Whitney chose this particular design. It requires a little more work to set up than the taper attachments on some of my other machines, but it is very rigid and very smooth. Everything is either clamped or moving in ball bearings. There are no sliding parts. I also wonder if part of this design wasn’t to make it better able to accommodate the relieving attachment.
I know this post is from back in the thread, but....
I have a square head 12C and I have also tried to correct the exact same "spongy"
cross slide condition. My shot pin seems in OK shape. Was also thinking about
replacing the square head screw with an Allen screw. I like the lapping the pin idea.
Might have to re-visit that.
Anyhow, I also took apart my ball bearing taper attachment.
It was missing most of the balls. Ha ! I think the reason they made it ball slide
design has something to do with the releaving attachment option.
Anyhow, mine lathe also has the 2 bearings. I speculated there was looseness
there. What I tried was to add a shim washer (I think between the inner races)
so that when the pin spanner bearing nut is tightened, it actually preloads the
races. Now you have to be careful not to over tighten the retaining nut, as now
it effects preload, not merely retains them like it used to.
This worked somewhat. I would tug (by hand) on the toolpost and see some
dial indicator needle movement, but it was much less. Before I added the washer
between the bearings, I had like .005" of deflection. Now I had .001" with about the
same tug force. It is still not rigid as I would like it to be.
I like your idea if a double row angular contact bearing. I will have to size on up
and try it out. I am glad you mentioned this bearing, because I was going to re-
visit the shot pin and maybe make something more rigid (and less convenient).
But now I will check out the bearings again. I would rather not have to re-invent
the wheel when working on machines (but I do, it seems on every machine I touch).
----Doozer