Re: [CAD_CAM_EDM_DRO] Ball-screws

James Owens wrote:

> Hi Jon,
>
> I should have made this clearer, I wasn't referring to backlash when
> writing
> about error. The error I was referring to is the 0.004"/foot error in
> the
> thread. This is the value I thought could be programmed out if it was
> known
> and in which direction. The screws claim a repeatability of 50 micron,
>
> correct me if I am wrong but 0.000005" is quite small.

Sure. If the threads are uniform, you just enter the appropriate
constant to
relate the number of encoder counts to a user unit. I'm told by people
who
do thread rolling that once the machine is set up, it will generate
threads that
are very uniform, even if not terribly accurate. So, that should work
out
fine for you. I checked whether I could improve accuracy of my machine,

but I have "High Precision" ground screws that are so accurate that they
are
within one encoder count per foot, so there is no sense in tinkering
with
the calibration constants.

> I was under the impression that it is repeatability is the thing to go
> for
> in CNC? The other question I have, is the difference in price between
> a
> Bridgeport kit say $750.00 and the low quote I got of $166.00 plus two
> extra
> nuts worth me paying. I cannot return it if it is wrong and the kit
> doesn't
> fit. I can however make a 1" ball-screw fit my machine provided I do
> the
> machining of the ends?

No, I wouldn't get the Bridgeport fit uness I was sure the machine was a
VERY
close replica of the Bridgeport (which some machines, in fact, are).
I'm using a .875" diameter screw, which is definitely considered too
small
for the Bridgeport, but it works quite well.

> In short the question is will this work, does anyone have experience
> of the
> McMaster-Carr ball-screws.

I'm sure it will work, but you are going to have some serious machining
to do.
Do you have a toolpost grinder? I found it to be very useful when
preparing the
ends of the leadscrews. You need to prepare one end of each screw
for a pair of radial contact bearings to constrain the leadscrew from
axial
motion. So, you need a bearing 'nest', a bore from each side with a
step in
the middle for the outer races to press against. The leadscrew has a
step
that presses against one inner race, and a collar screws onto the end of
the
leadscrew to press against the other inner race. This works like the
tapered
roller bearings in a lathe spindle, and the collar provides the
preload. This is
how I built up the Z axis mechanism on my mill from scratch.

Jon