Re: [CAD_CAM_EDM_DRO]Backlash & Wobbly Spindle on my Bridgeport

Carlos Guillermo wrote:

> I did try to measure the backlash as you described I zeroed out
> the table in
> "X" and then I moved it an inch with the indicator on one side I
> measured
> the the difference in the indicator to be .0025" Then I zeroed the
> table
> again (brought back to the zero position) moved the indicator to
> the other
> side and displaced the table an inch in the other direction and
> measured the
> difference to be .0030". Does this mean that the backlash is
> .0055"? If it
> is WOW!!!!

Hard to tell. If you have some sort of object with flat faces, you can
lift it out when doing the moves, and then seat it back after the machine
stops. This allows you to move either way without having to move the
indicator. If you are using a 1" travel indicator, then you could use the
.500" position as the zero reference, and move .25 or .5" either way
without bumping the indicator out of place. Move left and back to ref
pos, then right and back to ref pos, and compare readings. the difference
is the backlash.

.005" backlash is not surprising if the screws are worn. the good news
is that worn ballscrews can be rebuilt for much less than replacing with
new screws. For slight wear, or when the wear is uniform over the length
of the screw, the balls can be replaced with larger ones. If the wear
is localized, then the screw needs to be ground.

A second cause of backlash and other anomalies is wear in the dovetail
ways. The center of the bottom of the table wears, causing the table
to describe a slight arc as it moves. Also, when in the center, the underside
of the table is concave, and the top of the saddle is convex, allowing the
table to rock on the saddle. Since the X axis thrust is below the table,
the table will rock first, then start to move. Making measurements at
different heights above the table, or placing a master precision level
on the table will show this effect.

> With respect to the gauge blocks......well I think I'm gonna have
> to ask
> that one again 'cause that was way over my head.

The idea is you can make a series of measurements at regular intervals,
like in 1.000" steps, with a set of gause blocks. This can indicate the
cumulative accuracy of the screw and other parts.

Jon