Re: how important is backlash in a simple home setup ??

BOTTOM LINE:BACKLASH IS BAD! YOU NEED TO DO EVERYTHING POSSIBLE TO
ELIMINATE AS MUCH AS YOU CAN> Ideally your system will have no
measurable backlash. Practically, what you are going to need to do
is find the source of it and eliminate it. Continue doing this by
tightening everything up, adding a second preload nut, adding
bearings to your screw mounts, adding a thrust bearing, saving up
for a couple of decent lead screws. On and on and on. Everything
you do will increase your quality and satisfaction. There is
nothing worse than spending a bunch of time and money on something
that looks awfull. Good luck.

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Keith Clark" <mail@t...>
wrote:

>
> Johan,
>
> I am not sure if it will work well, but I plan to do the same

thing

> you have done only I plan to use a second simple nut to set up a
> system similar to that sold to eliminate backlash. The idea is to
> turn one nut against the left side of the threads and fix it in

place

> then and the second nut that can be adjusted against the right

side

> of the threads and fixed. That way you can remove the backlash.

If

> you look at the website of companies that supply them you can see

how

> their systems work.
>
> Unfortunately I can't find the site this minute but could search

for

> it later. One could fabricate such a system with minimum of

expense

> provided the equipment is available for the fabrication, most
> importantly a mill.
>
> Keith
>
>
> --- In CAD_CAM_EDM_DRO@yahoogroups.com, "Johan" <wykj@m...> wrote:
> >
> >
> > Hello again everyone, I'd like to know what sort of backlash can

be

> > tolerated without serious mishaps ?? I'm finalising the drivers

for

> a
> > cnc router I built, now before I spend too much cash I'd like to
> get
> > an honest opinion on how much backlash can I tolerate if fir
> instance
> > I'd like to do small plaque engraving, PCB drilling etc.
> >
> > I didn't have access to any expensive anti-backlash nuts etc.

(main

> > objective is to do it as cheaply as possible, if I like the

hobby

> I'd
> > spend more bucks)......... so I just used a single nut on some
> > threaded rod, plus the homemade couplings between the steppers

and

> the
> > threaded-rod I used also has quite a lot of play in them, so as

I

> > measured I'd have about 0.8 mm error, I suppose that's a lot,

but

> as I
> > measured it's only out by that much whenever I change direction

on

> a
> > axis, once it's moving in a single direction the error becomes

much

> > smaller, now if I use software backlash compensation does that

add

> 0.8
> > mm for every step or only when I change direction ??