Re: [CAD_CAM_EDM_DRO] Ball Screw Options

natchamp_87 wrote:

>Well, I've taken the plunge and purchased my mill for the CNC
>retrofit. I bought an older Bridgeport with an M head. It has a
>9x32 work table and its perfect for me because I get the full 12" Y
>travel but the overall machine will fit in my garage much better than
>a full 9x42.
>
>

Is this a round-ram machine? I have a 1938 round-ram that originally
came with an M
head. It only had 9" or so of Y travel. My CNC retrofit involved
making a new yoke for
the ballnuts, and I made some errors in it, and now have only 7" of Y
travel. Sometime I
need to pull it apart and remove a little metal on my yoke. But, I'll
still be stuck with about
9" of Y. Maybe with a little more carving I can get all the way up to
10", but that will have
the saddle coming at least 1" past the front of the knee. I have a 9 x
31" table, approximately.
My machine is serial # 1388, certainly one of the oldest Bridgeports
still running!

>Amoung other things, I'm trying to resolve is what I'm going to do
>for the ballscrews. I would really like input on what my options
>are. I've looked into the archives and am starting to get the angle
>of the dangle.
>
>1) Is one option to install anti-backlash nuts on the existing stock
>screws? I've read about making/installing anti-backlass nuts but
>can't figure out if that applys only to ballscrews.
>
>

This generally won't work, as the screws on all old machines are
differentially worn. Remove the
4 socket head bolts on the Y handle and unscrew the Y screw from the
yoke. This is good to
check for inadequate lube, and see how many tons of chips are inside the
knee! I filled a
trash can with what was inside mine! Looking at the lands of the Acme
threads, you will
almost certainly see they are thinner in the middle of the screw. Any
attempt to take up
all the slack at the middle of the screw will cause binding at the
less-worn ends.

>2) If I were to install new ballscrews and nuts what specs should I
>look for? Is 5 tpi the standard? How about the diameter?
>
>

5 TPI is very common, but 4 is also seen sometimes. Very fine-thread
ballscrews need very
small balls, making the balls wear out faster. Very coarse threads
allow rapid motion, but
need more torque to deliver linear force. So, 4-5 TPI would be the best
choice in X and Y.
I have a 10 TPI ballscrew on the Z axis that works fine.

>3) If I do #2 above am I right in that I have 2 options, either buy a
>bolt in kit or buy the ballscrews and fabricate how the nuts attach
>to the table?
>
>

Yes. Measuring all the heights to make everything fit right is
daunting, and I worked VERY hard
to get it right, and I still ended up putting as much as .060" of shims
here and there!

>4) What do I have to do/consider for attaching the ends to my
>servo/stepper motors? Do I have to do any machining to them for this?
>
>

You need to use pairs of angular contact bearings at one end of the
screw to firmly locate it
against axial thrust. You can use pairs of some standard radial
bearings for this, but without
a lot of testing, angular contact guarantees they will restrain the
screw. Most likely, unless
you have the screws machined for you, or you get used screws, you'll
have to turn steps, etc.
on the ends.

If you use servo motors that don't have integral encoders, a good setup
is to mount the encoders
coaxial to the leadscrews, and couple them with helical-slit couplers.
Then, you can use
toothed belts to give a speed reduction between the motors and screws.

Jon