Jason wrote:
I picked up a 1951 Singer Featherweight sewing machine recently.
It's
in great shape, except that even after applying lots of oil, the
shaft
which rotates the bobbin assembly resists rotating through about half
its travel. Presumably this means the shaft is very slightly bent,
although I can't detect any distortion visibly. So, while I can
still
turn it by hand, the motor isn't quite up to the task. Here's a
photo;
the problem is the light shaft running down the middle:

http://www.lunkwill.org/tmp/singer.jpg

The friendly university machinist helped me force out one of the dowel
pins, which are indeed tapered. I put one end of the shaft in a drill
press, and using a clamped pencil as a dial indicator noted perhaps a
millimeter of wobble at the bottom, then ran out of time for today.

The bobbin-side end cap was ground in places on the face next to the
bearing, so that seems to account for most of the resistance, although
it still doesn't turn as smoothly as I'd like even without the end cap
forced against the bearing.

The machinist pointed out the tapered cuts down into each end face of
the shaft, and it seemed like he was saying they're useful for checking
straightness and whatnot. But I can't figure how to use them for that
-- seems to me like I'd just clamp one end in the lathe and use a dial
indicator at the other end. (As you can see, I've reached the extent
of my true-Machinist skills)

Anyway, it seems like the options now are some combination of turning
down the shaft, grinding down a bit of the end cap so it doesn't bind,
and just manufacturing a new shaft. There are some scratches in the
journaled shaft surfaces; the machinist kept pointing out the journaled
surface -- is there some trick to producing such a surface on a lathe?

Also, I suspect the end caps are cast iron. All I can tell from Google
is that it's possible to grind cast iron, but it's different from
dealing with normal steel.

As to machining a new shaft, the biggest trick seems like it would be
measuring the angles for the dowel pins, since they're offset by a few
dozen degrees. First I'd have to get the other pin out, which doesn't
seem possible by pounding (even with heating); he said that'd entail
end-milling out the pounded up pin end before I could drill it -- that
just means milling it flat, right? Once the pin is out it seems like I
could line up one end so the drill sinks through the hole, then note
the table rotation required to sink through the other hole in order to
get the angle. But I'm hoping it won't come to that -- I think with
just a tiny amount of reduction, the shaft will go from
finger-turnable-with-a-little-effort to free-spinning, and hopefully
that won't cause noticeable errors in the sewing stitch.

Thanks for your comments! I think that unless fine sandpaper and maybe
the suggested valve grinding compound fix up the shaft, I'll have to
talk one of my engineering buddies into an exchange of food for
machine-shop tutoring.

-J