Re: [CAD_CAM_EDM_DRO] Linear vs. Rotary Enc.

John D. Bussema wrote:

>Howdy. I have a few questions about encoders. . . the main one being: Is it possible to use linear and rotary encoders interchangeably? If both give the same quadrature feedback. . . is there even a difference besides the obvious mechanical differences (and price)? For that matter, I guess I could ask the same of laser linear encoders. Has anyone tried to use linears on a small machine tool? I would sure like to see some pictures, because it seems like mounting them might be a little tricky.
>
>

The biggest difference is that very high resolution is easily achieved with
inexpensive shaft encoders. For instance I have 1000 cycle/rev encoders
on all 3 axes of my mill, direct coupled to the leadscrews. This gives 4000
counts/rev in quadrature. So, with the 5 TPI screws on X and Y, I get
20,000 counts/inch, or .00005"/count. On the Z axis, with a 10 TPI
ballscrew,
I get 40,000 counts/Inch, or .000025"/count.

These are Sumtak encoders, and I paid $75 each for them some years ago,
but similar units are often available on eBay for $30 - 50. Try to get
these
resolutions with a linear encoder, and you are almost certainly talking
thousands
of $ for an analog scale and an interpolator box. It is generally
considered
necessary to have an encoder resolution 10 x better than the "system
resolution"
which is 0.0001" on my EMC system. or the resolution you are programming to
(would be more like 0.001" for me). I'm somewhere in the middle of those.

The typical linear DRO scales are good to .0002", generally (really 50 uM).

The other big gotcha is backlash between the motor and the encoder. With a
shaft encoder, this is no problem. With an Acme screw and a linear encoder,
you will have a dithering disaster, as the motor pounds from one end to
the other
of the backlash, trying to control the encoder position.

Jon