Re: How bout this idea??

--- In CAD_CAM_EDM_DRO@y..., Doug Fortune <pentam@c...> wrote:

> sspeed33317 wrote:
>
> > I came up with this idea , using regular DC motors , not servomotors
> > mounting linear encoders right to all the axes
> >
> > would eliminate the need for Servo motors , would eliminate
> > the need for Zero lash ball screws ,would eliminate the need
> > for expensive servo motors , wouldn't have to compensate for any
> > backlash because it would be reading actual table movement .
> >
> > And the benefit of the ACME screws friction to help hold
> > the table in position .
> >
> > And you could tie in for DRO use ,or tie into an existing DRO
> > for CNC use if you already have one.
> >
> > Good idea??? Any body have input??? .
> > Already in service??
>
> Yup, great idea ... been using it for a few months now
> (see my webpage where I sell this type of acme leadscrew
> with an encoder nib on the end for mounting the encoder).
>
> The only minor catch is that you still have to make sure
> your motor-coupler-screw (or motor-gearing-screw) interface
> has very little backlash (compared to the resolution of the
> encoder mounted on the screw), otherwise a nasty oscillation
> can be set up (I'm guessing that a few encoder counts of
> backlash is OK).
>
> Now your table will still appear to have backlash if your
> table nut has backlash (this is a candidate for compensation
> in software). Luckily I have found using long length nuts
> (greater than 3:1 length:dia ratio) have under 1 thou backlash
> (when new & unworn). I use a 6:1 ratio and there is no measurable
> backlash.
>
> Note that (expensive but worth it in some applications) ballscrews
> have other desirable features beyond just the low/no backlash
> ballnuts, so just because you can eliminate backlash doesn't mean
> that ballscrews aren't the perfect answer (for some applications).
>
>
> Doug Fortune
> http://www.cncKITS.com

Hi Tim and Doug,

Doug, know from our experiments how and encoder on a lead screw can
cause trouble. There has to be next to no backlash in the system. So
again, your long acme nuts or spring loaded backlash nuts would have
to be used.
I also have an idea that may be crazy, but here it is anyway. There
have recently been a couple of threads discussing the virtues of
ballscrews vs acme (and all its grades) vs regular threaded rod. All
of these methods are bad for reasons of cost or poor accuracy. How
about if you removed the encoder from the screw altogether. I saw a
photo somewhere recently of a benchtop mill that had a dro where the
encoders were driven by a cable. The fellow had even used a pulley
system to increase the resolution. So what would happen if you had
inexpensive screws and anti backlash nuts of some kind, and cable
driven encoders. Seems to me that your accuracy would not be affected
by the screw at all. Just a thought.

I also like the idea of using just any dc motor and I have tried
several with good results. They are a dime a dozen, and are good
enough for hobby use. I just look for motors that have a relitivly low
rotor inertia. I can usually tell how a motor will act by spinning it
by hand. A good one is usually long and skinny, and wont spin by hand
very easily. Hook it up to 1 or 2 volts and see if it cogs. If it
doesnt, its a keeper.
Dean