Opinion
Patent Appeal No. 8955.
July 12, 1973.
Kenneth R. Schaefer, New York City, Roland A. Linger, Arlington, Va., attorneys of record, for appellant.
Ernest F. Marmorek, Marmorek Bierman, New York City, attorneys of record, for appellees.
Appeal from the Board of Patent Interferences.
Before MARKEY, Chief Judge, RICH, BALDWIN and LANE, Judges, and ALMOND, Senior Judge.
This appeal is from the decision of the Patent Office Board of Interferences, adhered to on reconsideration, awarding the junior party, Nagata and Ono (hereinafter "Nagata"), priority of invention as to count 4 relating to an electronic amplifier circuit employing an insulated gate field-effect transistor, on the ground that the application of the senior party Carlson, lacks support for the count. We reverse.
Minoru Nagata and Minoru Ono, involved on their patent No. 3,311,756, issued Mar. 28, 1967, on application serial No. 376,322, filed in the United States June 19, 1964, claiming priority of a Japanese application filed June 24, 1963, and assigned to Kabushiki Kaisha Hitachi Seisakusho.
David J. Carlson, involved on his application serial No. 245,063, filed Dec. 17, 1962, assigned to RCA Corporation.
The Contested Subject Matter
The invention relates to an electronic amplifier circuit employing an insulated gate field-effect transistor (IGFET). Priority as to independent count 1 and counts 2-3 dependent thereon was awarded to Carlson, while priority as to count 4, which is also dependent on count 1, was awarded to Nagata since the board found that the Carlson application does not support count 4. Count 1 specifies (our emphasis):
1. A transistor amplifier comprising a field effect transistor provided with a semiconductor substrate; a channel layer formed on said semiconductor substrate; source and drain electrodes fitted on both ends of said channel layer; an insulating layer formed on said channel layer; a first gate electrode fixed on said insulating layer and a second gate electrode fixed on said semiconductor substrate; a bias voltage source; an operating power source; a single [signal ?] source; a load; means to connect said load and operating power source in series with said source and drain electrodes; means to connect said bias voltage source to one of said first and second gate electrodes; and means to connect said signal source to the other gate electrode.
Count 4 specifies (our emphasis):
4. The transistor amplifier as defined in count 1, wherein said signal source is connected between said first gate electrode and the operating power source, and said bias power source is connected between said first and second gate electrodes.
The critical limitation of count 4 which the board found unsupported by Carlson is the requirement that "said bias power source is connected between said first and second gate electrodes." The issue on appeal is therefore whether the Carlson application supports this limitation of count 4. If it does, then priority on the count must be awarded to Carlson.
Appellant tells us that this should more properly be treated as reciting a bias voltage source. The board's treatment of this limitation as requiring a "source of bias voltage" is consistent with this more accurate characterization, and count 1, upon which count 4 depends for antecedent basis, recites a bias voltage source.
The Carlson Application and the Board's Decision
Without describing the details of the insulated gate field-effect transistor amplifier circuit which is the invention here, we will deal directly with those portions of the Carlson application which we find to describe the bias voltage source. We reproduce Carlson's Figs. 6 and 8:
The insulated gate field-effect transistor 70 (70' in Fig. 8) is seen at the center; it is a four terminal, or contact, device. Reference number 72 (72') indicates the "source," 74 (74') the "drain." The "first (insulated) gate" is 76 (76'), and the substrate, or "second gate electrode," is 78 (78'). The issue, therefore, is whether there is a bias voltage source connected between the first gate electrode 76 (76') and the second gate electrode 78 (78').
To place things in perspective, we note that Fig. 6 clearly shows no bias voltage source between 76 and 78, because there is a short-circuiting connection 88 therebetween. The Carlson specification states of Fig. 6:
Signals to be amplified from a source 84 are coupled through a signal coupling capacitor 86 to the [first] gate electrode 76 and to the substrate [second gate] electrode 78 through a direct current conductive connection 88. The gate and substrate electrodes 76 and 78 are biased to the desired operating potential by an adjustable source of operating potential 90 which is connected in series with a resistor 92 between the gate electrode 76 and ground. As is shown the gate and substrate electrodes are biased negatively with respect to ground.
It is Fig. 8 and the accompanying description upon which appellant primarily relies for a description of the connection of a voltage source between the first and second gate electrodes. Appellant's specification states, in pertinent part:
If desired, the cut off characteristic of an amplifier may be tailored to a predetermined characteristic by applying different biasing voltages to the [first] gate and substrate [second gate] electrodes as is shown in FIGURE 8. The circuit of FIGURE 8 is similar to that of FIGURE 6, and like components are given like reference numerals. The main difference between the circuits of FIGURE 6 and FIGURE 8 is that the gate electrode 76' and the substrate electrode 78' are isolated for direct currents by a d-c blocking capacitor 98. In addition, different d-c bias voltages are applied to the gate electrode 76' and substrate electrode 78' through resistors 100 and 102 respectively. The initial biasing voltages on the gate and substrate electrodes may be set to provide an initial maximum transconductance characteristic. The control bias applied to one of the electrodes 76' and 78' may be made more negative while the bias to the other of these electrodes is maintained at a fixed potential, or after a predetermined delay made more positive.
The board, considering appellant's argument that the pertinent limitation of count 4 is supported by Fig. 8 and the above language in the Carlson specification, said:
We fail to find in this any suggestion that the respective terminals of a source of bias voltage be connected to the two gate electrodes which we consider to be required by count 4. Rather, the entire thought appears to be independent selection or variation of the biasing potentials applied to the gate and substrate electrodes, apparently with respect to ground. Accordingly, the Carlson application lacks support for count 4 and Nagata is entitled to an award of priority as to this count.
The board's apparent conclusion that count 4 required "that the respective terminals of a source of bias voltage be connected to the two gate electrodes," was apparently based only upon consideration of the language of the counts because the board noted early in its opinion that "since Nagata [appellee] has not contended that the counts are ambiguous, reference to the Nagata specification is therefore not justified," citing Weiss v. Roschke, 425 F.2d 772, 57 CCPA 1264 (1970).
Appellant, in a petition for reconsideration before the board, stated that the impact of the board's decision "was to the effect that the count requires that the bias source be limited to two terminals, one connected to each of the gate electrodes," whereas all that is necessary to meet the count is that there be a difference in voltage between the two gate electrodes. Appellant maintains that his specification, which teaches the independent variation of the value of voltages at the first and second gate electrodes with respect to a reference voltage (the voltage at the source electrode), inherently teaches that a voltage source equal to the difference between the voltages at the two gate electrodes is applied between the electrodes.
The board stated, in a decision on the petition for reconsideration, that:
It is clear that in Figure 8 of Nagata et al [where a precharged capacitor is connected between the two gate electrodes] it is the intention to relate the bias on one gate electrode to that of the other whereas the Carlson specification quoted in our original decision [quoted supra with reference to Fig. 5] clearly indicates that the biasing of the two gate electrodes is intended to be mutually independent. In view of the latter statement we are not convinced by Carlson's arguments that two bias sources fixed with respect to a reference potential such as ground can properly be regarded as a single "bias power source."
The board's reference to appellees' specification was initially made because appellant had argued that Nagata's elements, in particular the capacitor, were the same as appellees', but it seems that the board improperly extended its consideration of the specification to interpret the language of the count which it had said it would not do.
OPINION
We agree with the board that the basic consideration is what the language of the counts defines as the invention and that only if the count is ambiguous should the specification of appellees' patent be consulted to determine its meaning. See Hemstreet v. Rohland, 433 F.2d 1403, 58 CCPA 743 (1970). The board found that the counts were not contended to be ambiguous by appellees and determined, therefore, that reference to the Nagata Specification was not justified. Appellant has not challenged this determination, nor will we. We accept the ground rule of the board that no reference to the Nagata specification is justified and proceed to consider whether the language of count 4, that the "bias power [voltage] source is connected between said first and second gate electrodes," finds support in the Carlson specification. We think it does.
The board was not convinced that the teachings of the Carlson specification that two separate mutually independent bias sources, fixed with respect to a reference potential, supplied to the two separate gate electrodes could be regarded as a single "bias power source." However, we think it inherent in the use of two different independently variable bias sources in the circuit of Fig. 8 of the Carlson specification, as compared with the circuit of Fig. 6, that there exists a potential difference, i. e., a voltage, between the two electrodes to which the separate bias sources are applied. That is all that we find to be required by the language of the count. A source of voltage between two points exists not only where a single two-terminal source, such as a battery, is connected therebetween but also where a different voltage is supplied at each of the points. To the extent that the board may have limited the count to a two-terminal power source device being connected between the gate electrodes, we do not find that one of ordinary skill in elementary electronics would so regard the words of the count.
While the two independently variable d-c bias voltages of Fig. 8 of Carlson could conceivably be at the same voltage level (as is always the case with the two gate electrode terminals in the circuit of Fig. 6), we believe that inherent in the concept of independent variation and in the language of the Carlson specification which states that the "control bias applied to one of the electrodes 76' and 78' may be more negative while the bias to the other * * * is maintained at a fixed potential," (our emphasis) is a disclosure that there be a difference in voltage, or a voltage drop, therebetween.
In Woofter v. Carlson, 54 CCPA 917, 367 F.2d 436 (1966), this court considered a similar contention that the application of Woofter did not contain a written description of the invention giving Woofter the right to make the counts. We there said, 367 F.2d at 441, 54 CCPA at 922:
The test of the right to make counts is to be found in the total disclosure. Thus, the omission of a written description in the specification of an application will not necessarily prevent the reading of a count on a structure shown in the drawing and described in general terms in the specification. Here, the function of the disclosed structure is inherent in the structure shown in the drawings forming a part of the application. * * *
In Padgett v. Warner, 104 F.2d 957, 26 CCPA 1403, this court observed:
As appellee's rollers are so arranged that they are free to move longitudinally in the cups, as clearly appears from the drawings in appellee's application, it is immaterial, so far as the issue here is concerned, that no reference was made in appellee's application to such longitudinal movement.
It seems clear, therefore, that the Woofter application discloses the structure recited in the counts in issue. Also, the functional statement at the end of count 4 seems to be inherently satisfied by the structure shown in Woofter Fig. 5.
We believe that here, as in Woofter and in Padgett v. Warner, the language of the count is satisfied by the disclosure of Fig. 8 and the accompanying language of the Carlson specification which teaches that the voltage levels at the first and second gate electrodes may be different. One of ordinary skill would find inherent in the teachings of the Carlson application a "bias power [voltage], source connected between said first and second gate electrodes." Compare Brand v. Thomas, 96 F.2d 301, 25 CCPA 1053 (1938). Accordingly, the decision of the board awarding priority of invention as to count 4 is reversed.
Even considering the Nagata specification, as the board did (see n. 4 and accompanying text), to show an "intention to relate the bias on one gate electrode to that of the other" we find appellant's teachings that the "control bias applied to one of the electrodes may be made more negative while the bias to the other is maintained at a fixed potential" to likewise "relate the bias on one gate to that of the other."
Reversed.