This is an appeal from the decision of the Board of Appeals rejecting all claims, 1-5, of appellant's application serial No. 782,708, filed December 24, 1958, entitled "Non-Linear Inductance."

The invention is in the field of electronics and, notwithstanding the title of the application, is claimed as "A charging circuit for a pulse-forming network." The "inductance" referred to in the title is part of that charging circuit. The specification does not disclose the use of the pulses produced by the network which is charged by the charging circuit nor does appellant's brief enlighten us in that regard, but one use would appear to be in radar.

The pulse-forming and charging circuit is schematically represented in the specification in highly simplified form in Fig. 1, thus:

NOTE: OPINION CONTAINING TABLE OR OTHER DATA THAT IS NOT VIEWABLE

This shows a pulse-forming network 1 comprising a series of capacitors 2 arranged in parallel between load 3 and various points on an inductance 4. A thyratron, which is a gas-filled three-element electron tube, is schematically represented as a switch 6 connected between the high-voltage end of network 1 and ground so that when the thyratron 6 is triggered it grounds the network. whereupon the capacitors discharge through the inductance 4, providing a current pulse through the load 3 which may be a square wave current pulse.

The charging circuit for the pulse-forming network comprises a d. c. voltage source 5 connected to the network 4 through an inductance 7 and a diode or rectifier 8. It is said to operate according to the principles of resonance charging, which appears to be a known phenomenon which is thus described in appellant's brief:

"* * * The unidirectional property of the diode [8] and the characteristics of the reactor [7] which oppose any change in current flow combine to effect charging of the capacitors to a voltage equal to twice the voltage of the direct current source. * * *"

All of the foregoing appears to have been known in the art. However, a problem existed which is described in appellant's brief as follows:

"* * * Reliable switching in such a system becomes a problem, however. In order for a thyraton switch to change from a conducting to a non-conducting state, its plate current must fall below a given threshold value; a condition which does not necessarily exist after each switching operation of the prior art pulse forming network. * * *"

The reason for the lack of reliability is described in the specification as follows:

"* * * It will be apparent that as soon as the pulse-forming network is fully discharged the valve tube [8] will act to initiate the charging from the d. c. voltage source, and frequently the result is that the pulse-forming network will start to be charged before the thyratron [6] switch mechanism has returned to the necessary original state. * *"

Appellant's invention is a change in the characteristics of inductance 7, which is also known as the "charging reactor," whereby the charging of the pulse-forming network 1 is delayed sufficiently to assure that the thyratron 6 has returned to its stable or open-circuit state before charging starts, presumably so that the charging circuit will not be connected to ground through the thyratron in a conducting state.

One way of doing this is described in the specification and drawing thus:

"In accordance with the invention a non-linearity is introduced into the inductance 7 as shown in Fig. 3."

NOTE: OPINION CONTAINING TABLE OR OTHER DATA THAT IS NOT VIEWABLE

"For example, said inductance 7 may comprise an iron core 9 including as a part of its magnetic circuit a saturable material 10, such as ferrite or some similar product. As a result, at low currents the iron core 9 has a high inductance, but as higher currents are reached, so that the ferrite or other material saturates, the iron core acts as though it had an air gap in it, since the ferrite is no longer acting as part of the low-reluctance magnetic circuit. The inductance at these higher-current values would then be inversely proportional to the length of the high-reluctance `gap'. The effect of providing such a saturable material in the iron core 9 of the inductance 7 is shown by the broken line in the graph of Fig. 2 [which shows a time delay in the rise of charging voltage]. At first the inductance in the iron core 9 is relatively high, thereby opposing the flow of current so that the current remains low. However, when the saturable material 10 becomes saturated, the inductance is decreased by an amount which is proportional to the thickness of the saturable material 10. As a result of this reduction in inductance, the current flow is permitted to increase in the normal manner. The overall effect of the addition of the nonlinear inductance is to delay the charging of the pulse-forming network 1, thereby permitting the thyratron switch 6 to return to its stable state."

Appellant discloses a second embodiment of his invention wherein he uses, instead of the Fig. 3 core with ferrite insert, an inductance with two coils in series each having its own core, one core reaching saturation at an early point in the cycle while the other does not, thus producing the same effect with different means.

The examiner rejected the claims for reasons unknown to us, citing as references three prior United States patents which are not in the record before us and which we need not consider because the board disagreed with the examiner and refused to sustain his rejection. For this reason the record contains no examiner's actions and no examiner's answer; the appellant prevailed before the board in having the examiner reversed.

In his brief before the board, appellant called attention to "page 12 of Vol. 5 of the Radiation Laboratory Series concerning resonance charging of pulse forming networks," as stated in the board's opinion. The board cited a new reference as follows:

Pulse Generators, Vol. 5, Radiation Laboratory Series, page 341 and pages 364, 365 and 366

This is the only prior art of record before us — four pages of what appears to be a book of several hundred pages. That book, as such, is not of record nor have we been furnished a copy.

After citing and discussing the four pages above mentioned, the board made the following rejection:

"* * * Therefore, under the provisions of Rule 196(b) we hold that claims 1 to 5, inclusive are unpatentable over the disclosure in Vol. 5 of the Radiation Laboratory Series.

"The decision of the Examiner is reversed.

"A new rejection of claims 1 to 5, inclusive, under the provisions of Rule 196(b) is made."

Appellant filed a request for reconsideration and argument but the board fully adhered to its position.

"1. A charging circuit for a pulse-forming network, comprising, in combination with a pulse-forming network: a source of d. c. voltage and an inductor linking magnetic material comprising two components which saturate at mutually different applied magnetomotive forces [x] connected in series between said d. c. voltage and said pulse-forming network [y], the inductance of said inductor having a non-linearity such that said inductance is less at currents approximating the maximum current delivered to the pulse-forming network during the charging operation than at lower currents delivered to the pulse-forming network during the charging operation." [Emphasis ours.]

Claims 2, 3 and 4 all depend from claim 1 and add limitations thereto. Claim 5 is another independent claim which is the same as claim 1 except for the insertion therein at [x] of "and a diode" and at [y] of "in such a way as to provide resonance charging of said pulse-forming network." While these two additions make claim 5 more specific, it appears to us that neither addition has anything to do with what appellant invented, both being features of the acknowledged prior art. The principal feature of the claims on which reliance is placed for patentability is the clause we have italicized which gives rise to the "non-linearity" characteristic described in the concluding clause of claim 1. These claim limitations are common to all claims.

In the first board opinion, wherein the only rejection before us was made, the board acknowledged that the claims distinguish from the sole reference on which it relied. The board said:

"Claim 5, the most specific claim, is readable on the above disclosure of Vol. 5 except for the statement of `two components which saturate at mutually different applied magneto-motive forces.' * * *" [Emphasis ours.]

However, the board was of the view that appellant was urging patentability on the basis of using "a charging reactor 7 which saturates at some point during the charging cycle," further concluding, in view of appellant's specification disclosure of two different core arrangements for accomplishing his purpose, that "There then appears to be no significance in the type of saturable reactor used in the charging circuit." [Our emphasis.] Because the board found in the reference pages relied on reference to "approximately linear inductors" and a statement that "Large variations in the charging inductance produced by saturation of the iron core are usually not permissible * * *" (board's emphasis), it concluded that the reference would suggest the use of "non-linear reactors" (i. e. inductances) and the use of them at core saturation, whereby large variations in charging inductance would be permissible. The gist of its rejection appears to be in its concluding statement that the use of two core components which saturate at mutually different applied magnetomotive forces "appears to have no significance different from a [single] core that saturates."

Appellant strenuously argued in its petition for reconsideration that the board's view was erroneous and does likewise in this court.

For the several reasons discussed below, we cannot agree with the opinion of the board.

First, we think the board oversimplified the claimed invention in treating it as the use of an inductance with a core "which saturates at some point during the charging cycle" and in assuming that the type of saturable reactor used is without significance. All inductor cores can be saturated and when they are the value of the inductance changes radically. If appellant's invention means anything, it means that he was not satisfied to use the prior art single core inductors which were capable of saturation. His answer to the board in his petition was as follows (our emphasis):

"Compare now the nature of the single saturation point non-linear inductor of the prior art as described above with appellant's device * *. The composite structure of appellant's inductor successively introduces two different values of inductance into the charging circuit. The first value of inductance, being large, opposes the flow of current and thereby delays any appreciable charging to the pulse forming circuit. * * * When one portion of the composite inductor saturates, the second value of inductance is introduced into the circuit. This second value of inductance, being appreciably less than said first value of inductance, permits the current flow `to increase in the normal manner' * * *. By `increase in the normal manner' is meant in the substantially linear fashion recommended in the cited reference * * *. Thus, the charging current in appellant's device operates substantially as the charging currents in the preferred prior art device with the exception that, by means of the novel inductor element, a time delay is introduced that makes possible the various advantages enumerated in the specification.

* * * * * *

"In summary then, appellant wishes to stress the following points:

"(a) The type of saturable reactor used in the present invention is significant * * *." [Original emphasis.]

Second, the reference, which admittedly does not show appellant's composite core inductor, and particularly in the passages relied on by the board, clearly teaches that the single core inductors under discussion are not to be so used in pulse-formers that they operate at or near saturation, or near the "knee" of the saturation curve, as the solicitor suggested in his argument in this court. In fact, after a rather detailed discussion of the problem, the reference makes the flat statement, "This occurrence indicates a high degree of saturation and, of course, is not permissible in practice." At another point, after extensive discussion of the importance of linearity and some necessary compromises with it to save weight and size, within rather narrow limits, we find the following statement:

"* * * When weight is not an important factor, the majority of charging reactors for pulsers designed at the Radiation Laboratory satisfy the following conditions for linearity * * *.

"1. Linearity. The inductance shall not change by more than 5 per cent measured at rated full-load and at half the rated full-load direct current."

We agree with appellant "That the teaching of the reference is in direct contradiction to the teaching of the Appellant's patent application * * *."

Third, upon consideration of all that the reference pages disclose, we are unable to find therein the slightest suggestion of the claimed invention. All we find is description of the acknowledged prior art pulser on which appellant's invention is an improvement.

While the board has not so characterized it, its rejection is necessarily one for obviousness under 35 U.S.C. § 103 since it admitted the claims are not met by the reference so as to be rejectable for want of novelty under section 102. We can see no reasonable basis in the reference for holding the invention to be obvious. In fact, we do not even find an indication of the existence of the thyratron switching problem which the invention allegedly solves, or any other reason for so constructing the charging inductor as to delay the initiation of charging current.

The solicitor has laid emphasis on the breadth of the claims but there is no objection to them on that ground, and he has noted that the thyratron is not "claimed" by which we suppose he means not included or mentioned in the claims, but we see no reason why it should be. The novelty of the circuit resides exclusively in the changed form of the inductor. He has also speculated on the possibility of the prior art using in pulsers other forms of inductor cores such as cores of certain alloys and cores with segments of less cross-sectional areas and has included in his brief several more "references" or extracts from the literature, which were not mentioned by the board, in support of these arguments. These, however, are new arguments never used in making the rejection and if they are to be relied on we think they should be made in the Patent Office where appellant has the opportunity to answer them as contemplated by the statute, 35 U.S.C. § 132, and the rules, Rule 104(b), and not presented to him for the first time in this court. This court is a court of review and reasons for rejection not made in the Patent Office are not properly before us.

The decision of the board is reversed.

Reversed.