Appeal No. 95-2413
Application 07/936,865
should not exist in an actual magnet, such as in figure 6.
Moreover, we believe that appellants misinterpret the shaded
regions in figure 9 to be the magnetic path instead of the
areas swept out on the target by the arcs of the curve 67 in
figure 8.
Appellants argue (Br8):
On the other hand, the "modified" design of FIG. 6 fails
to meet the total arc length limitation of Claim 36. For
example, the patent admits that the actual design is such
that, "Each area swept by the arcs a-b and b-c is almost
proportional to radius Or-a and Or-b respectively."
(Column 5, lines 64-66, emphasis added.) Thus, it is
admitted that the actual magnet design only approximates
the total arc length limitation and no teaching is
presented which suggests how to actually achieve this
limitation with a closed-loop magnet.
Suzuki expressly teaches that the total arc length (i.e.,
'L(r)) at a distance r from the axis of rotation divided by
the distance r should be constant, i.e., 'L(r)/r = constant.
Suzuki teaches a plasma region having the disclosed spiral
shape and a very narrow width will produce an erosion rate
that is "almost the same" (col. 6, lines 3-4). While Suzuki
admits that 'L(r)/r is not exactly constant, the disclosure of
"almost the same" is considered to teach "substantially
proportional" as recited in claim 36, because this term does
not require exactness. Given that real magnets having finite
- 18 -
Page: Previous 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Next
Last modified: November 3, 2007