Re: ATM Direct drive?

[Donald Clement <clement@inetworld.net>]

Andy,

The 0.0001" is certainly doable. Electroglas wafer probers have this
kind of accuracy. I feel that a considerable effort goes into making a
precision gear reduction system. For a direct drive stepper, a curved
array of permanent magnets and would have to be built. But this would
not necessarily be expensive. Requires learning a new technology though.
Perhaps some on this list have some experience with this technology.
Direct drive would not have the problems of backlash and wear. It would
not require lubrication. A larger assembly of many coils in the
stationary part could average the error of fabrication to increase
accuracy. Electroglas probers calibrate the nonlinearity of the X-Y
position and store this in a lookup table. Same sort of thing could be
done with a direct telescope drive. In an Alt-Az type of drive there is
already a computer. So nonlinearity could be easily stored by software.
Perhaps it could even learn while the system is being used. 

Don Clement
San Diego/Running Springs

iss@pvtnetworks.net wrote:
> 
> Don C writes,
> 
> >Perhaps a similar linear stepper could be incorporated in a large curved
> >sector. The same drive circuitry that works with standard steppers,
> >including microstepping, would apply to the direct drive.
> 
> I suspect this technology would not be cheap to drive in < .5 arc sec
> increments. At 20.6" radius 1 arc sec = .0001"....
>