Re: Max stepper motor speed - rpm

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "Scott Riddle" <sariddle@l...>
wrote:

> Great info. I have (2) 7/8 dia ball screws and my (4) 36" long THK
> rails are limited to about28" of travel when I take out the length and
> spacing of the 2 bearing blocks. So I am looking at about 28x28" for X
> & Y and I am still up in the air as to Z stroke. I was thinkg
> somewhere around 6-10". I want to be able to do flat work and also put
> a 4th rotary axis on the table to mill round pieces. Kind of a
> universaly capable machine. I have a great scrap pile that I have been
> able to salvage some pretty sweet motion items from. My table is a
> 3x3" X .25" wall tube frame. It was made for something else but I am
> able to work with it. I weighs in around 300+ lbs. It was to much for
> me to lift up onto the trailer so I just rolled it on.
>
> I have a question about the ball screw. Do you, or have you seen one,
> that preloads the screw itself in slight tension to get better frame
> stiffness? There are many things to consider like thermal growth and
> shrink but I have a very easy way(self adjusting) to do this and was
> thinking to stiffen up the frame I would do this. It does put a small
> load on the screw bearings and more drag for the motor to overcome but
> I think it would give much tighter machine dynamics. I have been
> looking around the web at many different designs for about the last 8
> years and I definately know what to stay away from for the accuracy
> and strength of machine I am looking to build. It wont be a plywood
> and round rod guide machine. ;-)
>
> For fun I have built a egg plotter that uses 2 steppers kind of
> arranged like a lathe. One stepper with a rubber cup on it holds a egg
> with a spring loaded rubber cup on the other end. The second motor
> sits at 90° to the first and has a arm to hold a pen or pencil. It was
> a fun little project and produced some cool Easter eggs. The only
> problem I had was the IMS drivers I had were full/half step only so my
> resolution with the pencil looked a little rough. I am hoping to use
> the 10 micro step Geckos to see how much better it can look.
>
> Thanks a bunch! Great info.
>
> Scott
>

Couple of issues here. The general rule for microstepping is that you
gain smoothness but not accuracy. Resolution and accuracy are not the
same. Resolution is the distance your machine moves with one discreet
pulse from the signal source. Obviously if you move 1/10 of a full
step then the resolution is 10X better.....well sorta. Because
fractional steps are done by current modulating multiple windings and
setting up opposing magnetic fields to move the rotor through the
areas between poles (vector currents), it works better as a dynamic
operation that trying to stop (hover) at a position along the path.
There are too many forces and variables to say with certainty the
position will be accurate between poles. The only true accuracy is
when the rotor is aligned with a pole. Most steppers have 200
positions. You can say in a loose way that resolution is finer with
microstepping and there HAS to be some gain in accuracy, but it is not
10X. The primary benefits of microstepping are the smoothness of
motion and the reduction of resonance points that can cause lost steps.

The problem with putting a ballscrew under tension is it needs to be
really accurate mountings. The charts on leadscrew flex are usually
presented with one end captured (double thrust bearings at the driven
end , no bearing at the other), One end captured with the other in a
bearing (no tension - shaft has some play in the length) and finally
both ends under tension. Of the setups, the one I have seen used the
most on commercial machines is #2 where the driven end has opposing
double thrust bearings and the far end is supported with a bearing but
there is some play lenghtwise to allow for things like thermal expansion.

The best way to get zero backlash is to use zero backlash ballscrews.
They typically consist of two ballnuts with a tension device between
them to take out all back lash. There are single nut anit-backlast
nuts that use varying sized balls in the nut and have to be assembled
precisely at the factory.

There are lower cost spring loaded acme nuts and screws that work
pretty well for router applications. Your accuracy on a router is not
as critical as a milling machine where you are expected to be able to
make parts that will hold .0005 to .001 tolerance on a daily basis.
Wood is more forgiving and if you can hold .005 overall you have a
good machine. Wood will shrink and grow that much with just changes
in humidity. While there is a level of self satisfaction in building
a highly accurate machine, accuracy you don't need is just wasted time
and money.

As you have probably seen, accuracy of your machine is a function of
all the errors summed so you have to work at each stage to get the
smallest amount of slop.

There are hundreds (maybe thousands) of us that have built routers,
engravers, CNC mills, CNC lathes and plasma cutters. Each time we
build one we immediately start mentally designing the next one! There
is always something you would have done differently.