Opinion
No. 67 C 991.
September 29, 1970.
Hanse H. Hamilton, New York City, for plaintiff.
Robert R. Keegan and Michael J. Sweedler, New York City (Darby Darby, New York City, of counsel), for defendant.
Plaintiff has sued for a declaratory judgment that its light-dimming devices do not infringe defendant's Slater patent Reissue 26,119 and that the patent is invalid. Defendant counterclaims for infringement.
It is concluded that the patent is invalid so far as concerns claims 1, 2, 8 and 9, and if the said claims were valid, there would be infringement only by plaintiff's Model No. 671.
The ordinary domestic wall switch used to turn lights on and off is conventionally housed behind a switch plate in a metallic box of standard dimensions, and, electrically, it is located in one of the two wires that conduct the electric power to the lamps. The lights are turned on and off by pressing a button, or rotating a short protruding lever, that actuates a switch which opens or closes the electrical circuit mechanically by removing and restoring a segment of the conductor.
Continuously variable dimmers have long been in use, conspicuously, in the theatre. However, until the period of the patent in suit there were no continuously variable dimmers that could be safely and readily installed in the conventional wall switch box in direct substitution for the familiar two-position wall switch. The central question is whether Slater Re. 26,119, which discloses and claims such a device, evinces patentable novelty, and the related questions are whether certain Leviton dimmers infringe one or more claims of Re. 26,119. It appears to be conceded, and the evidence is clear that such dimmers are useful and are now in wide use in wall-switch installations. And the evidence indicates that Slater's switch was the first in the field.
Slater's dimmer awaited the availability of the semi-conductor devices; it functions through electronic circuitry, the use of a diode and a silicon controlled rectifier, and adroit deployment of mechanical switching.
The device is designed for use on alternating current electrical circuits. Alternating current reverses polarity a stated number of times a second; for example, domestic house electricity in this area alternates 60 times a second, that is, it rises from zero voltage to a peak of 110 volts of one polarity, falls to zero voltage, then increases to a 110 volt peak of opposite polarity, and returns to zero voltage, repeating the cycle 60 times a second. The cycle is graphically depicted as a sine wave (see, e.g., Ex. AC, p. 42, Fig. 93(A), or Exs. K, L, M). Such an electrical supply can be effectually halved by blocking the passage of voltage of one polarity so that a pulsating direct current, that is, a fluctuating current of only one polarity, is conducted, and no current at all is conducted during the time when the current of opposite polarity would have been conducted but for the blockage. One device, long familiar, used to produce such a "unidirectional" current has been the kind of "electronic" vacuum tube or valve known as a "diode." Since the discovery of the semiconductors, compact semi-conductor diodes have been available.
In addition to thus making alternating current unidirectional, methods have been evolved to control the point within each half-cycle at which the unidirectional current will start to flow. One method is through the use of a triode in which a third "gating electrode" is used to introduce into the electronic path between the other two electrodes a current used like a trigger to control the time within each half-cycle at which the flow of current will start — the "firing time." The familiar tube used in earlier electronics in such circuits was the thyratron, a gas-filled tube. Since the introduction of the semi-conductor devices the silicon controlled rectifier ("SCR") has been developed to perform essentially the office of the thyratron.
Slater's Re. 26,119 presents a light dimmer that, through exploitating ( inter alia) the miniaturization potentials of the semi-conductor diode and SCR can be installed in a wall switch-box as a direct replacement for the familiar wall switch. The Slater dimmer uses an SCR connected in one side of the power line in series with the load (as is the conventional wall switch) across which a time-delay, phase shifting, circuit (energized from the voltage across the SCR) is connected to provide a continuously controllable gating signal to the gate electrode of the SCR, thus controlling the power supplied to the load by controlling the portion of each alternating current half-cycle during which the SCR conducts. The current through the SCR is a unidirectional one that cannot exceed fifty percent of the bidirectional alternating current; hence the Slater device incorporates a mechanical switch arranged to bypass the SCR and connect the load directly into the power line at full power after the SCR has reached maximum conduction. The time delay, or phase shifting, circuit is not restricted to any specific circuitry, but it is strictly required that it be energized from the one side of the power line in which the load and SCR are connected in series — a requirement dictated by the goal of making the device directly substitutable for a conventional wall switch. The Slater device by its nature requires a manually operable control for the time-delay circuit, a means of bringing the bypass electrical path to the load into operation, and a means of cutting off altogether the flow of current to the load. In the embodiment pictured and described in the patent, the illustrated time delay control means is a variable resistance the wiper arm of which is operated by a rotatable shaft that extends through the switch plate on which the dimmer is mounted, and the shaft is fitted externally with a control knob; disk cams mounted on the rotatable shaft are of a configuration such that at the "off" or counterclockwise extreme of rotation the time delay circuit is switched off and de-energized; at the clockwise or "bright" extreme of rotation, the variable resistance is shorted out and the bypass switch is closed so that full power is applied to the load. The Slater device is housed in an insulating housing fitted with lugs for attaching it to a switch plate, and binding posts for connecting the device in one side of an alternating current power line; provision is made for "heat sink" elements within the housing to dissipate the heat generated by the device.
Claim 10 does require that the circuit contain a capacitor and "means" (unspecified) for varying its charging rate. Claim 5, cancelled upon reissue, was more specific.
Slater Re. 26,119 provides also for (and separately claims) a dimmer as described with the addition of a diode, connected in parallel with the SCR and of opposite polarization; that diode does not function until switched into operation when the SCR is conducting at full power (50%); the diode would then pass 50% of the power (though of opposite polarity to the current through the SCR) relieving the SCR of all work; the SCR circuit would then be available to supply continuously variable power from 50% to 100%. The Re-issue patent also outlines and claims, in effect, a twinning of the basic SCR and time-delay circuits to furnish a dimmer operable from substantially zero to 100% by simultaneously varying both polarities of the alternating current.
Claims 1, 2 and 9, used to evaluate the issues of validity and infringement read as follows —
1. A power control and switch device providing substantially continuous control of power supplied to a load over a substantial range of power values comprising
(A) two terminals for connecting said device in series in one side of the two sides of the power line from an alternating current power source supplying current to the load,
(B) a rectifier element having a cathode, anode, and gating electrode,
(C) a switch,
(D) means connecting said rectifier element between said terminals to cause said rectifier element to be connected by means of its cathode and anode terminals in series between said two terminals,
(E) means connecting said switch between said two terminals to selectively provide a separate electrical path between said two terminals, and
(F) a variable time delay means energized from said two terminals for providing a gating signal to the gating electrode of said rectifier element,
(G) said gating signal being controllable to initiate conduction by said rectifier element at instants of time bearing a controllable relation with respect to the beginning of alternating current cycles of power available at said two terminals,
whereby a portion of each alternating current cycle, which portion is continuously controllable in duration over a predetermined range, may be supplied to the load connected in circuit with said device.
2. Apparatus as claimed in claim 1 wherein said rectifier element is a controlled rectifier and further including.
(H) a further rectifier element and
(I) means for connecting the last said rectifier element in shunt between said two terminals with its polarization opposite to that of said controlled rectifier.
9. A power control and switch device for use as a writing device to be installed in a building providing substantially continuous control of power supplied to a load over a substantial range of power values,
(J) said device comprising an insulating housing adapted to fit in an outlet box,
(K) two terminals for connecting said device in series in one side of the two sides of the power line from an alternating current power source supplying current to the load,
(L) said power line comprising conductors of a building wiring circuit,
(M) means for respectively connecting said terminals to conductors comprising said one side of said power line,
(N) a semiconductor control element within said housing having first and second electrodes,
(O) said control element normally being substantially non-conductive between said first and second electrodes and being adapted to be rendered conductive in at least one direction by a control signal,
(P) means within said housing connecting said control element in series between said two terminals by means of its first and second electrodes,
(Q) variable time delay means within said housing energized from said two terminals for providing a control signal to said control element,
(R) mechanical switch within said housing arranged to selectively open the energization circuit for said time delay means and
(S) a manually operable control member for said time delay means extending outside said housing,
(T) said switch being mechanically connected to said control member for actuation thereby, and
(U) said time delay means being controllable by operation of said control member to initiate conduction by said control element at instants of time bearing a controllable relation with respect to alternating current cycles of power available at said two terminals,
whereby portions of alternating current cycles, which portions are continuously controllable in duration over a predetermined range, may be supplied to the load connected in circuit with said device.
The Patent Office history of the patent is somewhat complex. Part of the allowed claims ( i.e., claims 1, 2 and 3) derive essentially from Slater's application S.N. 3144 filed January 18, 1960, and abandoned in connection with the allowance of such claims on Slater's application S.N. 35,298 filed June 10, 1960, as a continuation-in-part of S.N. 3144. Claim 4, dependent on claim 1, was, as original claim 11, dependent on a formally different claim, original claim 8. Patent No. 3,103,618 issued on application S.N. 35,298 on September 10, 1963, with seven claims. An application, S.N. 516,821, was filed November 16, 1965, for reissue of Patent No. 3,103,618 on the ground restudy of the patent with a view to its enforcement indicated that the applicant had claimed, particularly in claims 5 and 6, more than he had a right to claim in the patent, and that the claims omitted statements and limitations essential or helpful in pointing out the invention — "which is directed to wiring devices including an electrical circuit for power control rather than an electrical circuit per se." The result of the application was Reissue 26,119 of June 10, 1960, in which claims 5, 6 and 7 were cancelled and new claims 8, 9 and 10 were allowed. Meanwhile, and both before Patent No. 3,103,618 issued and before the abandonment of S.N. 3144 on August 5, 1963, Slater filed on January 4, 1963, an application, S.N. 250,458, as a division of S.N. 3144, and on a further division of S.N. 250,458, filed as S.N. 562,259, Patent No. 3,328,676 issued on June 27, 1967.
All claims of application S.N. 3144, claims of which were in substance granted in the patent in suit, were initially rejected on the ground that the art showed a mercury vapor tube dimmer control with the tube in series with the load and the alternating current source, and with a control circuit comprising a capacitor and variable resistor to control the gating signal initiating conduction (Craig No. 2,001,837), and that to substitute an SCR for the mercury tube was obvious in the light of the earlier use of gated silicon rectifiers in a dimmer control for a theatre lighting system. (Macklem No. 2,920,240), and the known technical characteristics of SCRs (Ross No. 2,877,359); it was said that Morton, No. 2,896,125 showed the use of a (diode semiconductor) rectifier and switch in shunt so as to give either half power through the diode or full power through the switch, and to transfer that shunt switch idea to short out Craig's mercury tube to supply the full alternating current to the load was not patentable innovation.
All claims of S.N. 35,298 (on which the suit patent issued) were next rejected, ten of the twelve claims being rejected as unpatentable over certain of the claims in S.N. 3144, again citing Craig, Macklem and Ross. Morton was not mentioned.
Next the specification of S.N. 3144 was elaborated to embrace (1) a description of circuitry in which the full-power by-pass elements are eliminated and a diode circuit in parallel with but of polarity opposite to the SCR is provided and arranged to shunt half-power around the SCR when the variable control reaches the SCR's maximum so that the SCR can then operate to vary the supply to the load from 50% up to 100%; and (2) the addition of a twin SCR circuit of opposite polarity to the first SCR to supply zero to 100% variability by control of both halves of the waveform. Three claims were added.
Slater argued that his claimed invention was "a control circuit" with advantages of simplicity not earlier obtainable, e.g., (a) that the device can be inserted in one side of the power line (no access to the other side being required) so that it can directly replace the ordinary snapaction switch; (b) that the device wastes little power (in contrast to a variable resistance device); (c) the percentage of power supplied to the load at any setting is substantially independent of local impedance. A part of the advantages was attributed to the energizing of the time delay circuit from the voltage across the SCR — whereas Macklem powered the phase-shift circuit by the voltage across the power supply; that required wiring to both sides of the power line and inferentially ruled out facile installation in a wall switch box; in addition, Slater argued that his device presented a simpler control structure than Macklem. Macklem and Craig were distinguished as showing their time delay circuits as powered from both sides of the supply line; Ross was characterized as merely illustrating SCR characteristics and their use in an unrelated context and function. Morton was distinguished as not showing any means for continuous variation of power. The new claims (beside covering the new circuitry etc.) were said to embody statements clearly distinguishing from the prior art.
Later, comparable material was added to S.N. 35,298. Slater argued that the device of the application differed from that of S.N. 3144 in circuitry, and particularly in that S.N. 35,298 showed a diode in the input of the time delay circuit to supply fluctuating direct current to the capacitor, whereas the only diode used in the basic circuit of S.N. 3144 was in series with the gate electrode, i.e., in the output of the time delay circuit; improved operation of the circuit and switch and reduction in power consumption and rating were said to result. Responding to the rejection based on Craig, Ross and Macklem, it was said:
It should be pointed out that the present application (as well as applicant's prior co-pending application Serial No. 3,144) provides a device with a novel advantage of great practical importance. Applicant's device may be directly substituted for a conventional snap switch because its installation requires access to only one of the two sides of the power line to the lamp (or other electrical device). All of the references relied upon require access to both sides of the power line to obtain power for the dimmer control apparatus.
Even ignoring the absence of this important feature in the prior art, it will be noted that none of the references relied upon shows a dimmer control with a diode rectifier in the input to a time delay circuit, as claimed in this application. In fact, only the patent to Craig has a time delay circuit and it shows no diode or equivalent unidirectional current conducting means.
The role of the diode (specification Col. 5, ll. 13 et seq.) is (i) to supply a pulsating direct current with a "ripple" component of the same frequency as the line voltage, (ii) to make it possible to use a physically very much smaller variable resistance in the time-delay circuit by greatly reducing the power in that circuit, and (iii) to dispense with the diode in the lead to the gate electrode shown in S.N. 3144 by functioning in its stead to furnish current of the correct polarity to the SCR gate electrode.
Commenting on Macklem, Slater said
The patent to Macklem has several diodes as components of a magnetic amplifier, but this teaching cannot be related to the Craig disclosure because the Macklem circuit utilizes a substantially constant variable bias, rather than a variable time delay for gating purposes.
In S.N. 3144 the Examiner next rejected as new matter the material relating to the second diode in parallel and the twinned SCR circuit, rejected four claims as not patentable over a published circuit (Ex. 17, item 9, Fig. 22) disclosing "a controlled rectifier CS with a RC [resistance-capacitor] control circuit therefor between two terminals in one side of the line. To make RC adjustable or house the system in a control can would not involve invention." The circuit referred to was described in the cited reference as a time delay circuit, showed load and three-electrode semi-conductor in series in one side of the line, and a resistance — capacitor time delay circuit in which the junction between capacitor and resistor is connected (although not through a limiting resistor) through a diode to the gate electrode of the rectifier (see Ex. A B, p. 6, ll. 17-19). The Examiner stated that four claims appeared allowable; all four were among those claims said to require rejection of certain of the claims in S.N. 35,298.
Before the Patent Office responded in S.N. 35,298 to Slater's amendment and argument for allowance of the claims in that application, Slater supplemented them by submitting amendments to the specification and three of the allowed claims from S.N. 3144 and other claims. It was stated that upon allowance of S.N. 35,298 Slater would abandon S.N. 3144. Commenting on the Examiner's reference in S.N. 3144 to the time-delay circuit Slater observed that the circuit was a direct current time delay circuit and otherwise substantially different from his invention.
Slater then indicated in S.N. 3144 that S.N. 35,298 be acted on first, and, while he continued to support S.N. 3144, stated that if S.N. 35,298 were allowed S.N. 3144 would be abandoned. The Examiner then stated in S.N. 35,298 that if S.N. 3144 became abandoned, the claims then in S.N. 35,298 would be allowed. S.N. 3144 was thereupon abandoned (after the filing of the division application S.N. 250,458). Further amendments and the claims (19 and 20) later allowed as 6 and 7 were then submitted in S.N. 35,298. The applicant argued:
At the interview applicant explained in considerable detail those devices which had previously been utilized for dimming lamps in installations in homes mentioning amongst other devices variable resistors, variable reactors and pointing out that such devices required that both sides of the alternating current line be available at the switch location.
Not only did these old devices require that both sides of the line be available at the switch box, but also these devices were in general of such physical size that they could not be located in a single gem box but rather required a double or triple box and of course the necessity for removing a portion of the plaster or other wall to accommodate the oversized box.
* * * * * *
However, the claims in the application are limited to a silicon controlled rectifier or, in some instances, to a rectifier having three electrodes. As was pointed out to the Examiner this limitation is unnecessary since a semi-conductor element having only two electrodes can be utilized. Thus applicant's claims do not wholly protect his invention which, as stated above, is in essence the use of a semiconductor element in one leg only of the a.c. supply line and with no control circuit which must be connected to both sides of the line and which control element serves to adjust the brilliancy of the lamp over a range of adjustment.
Claims 19 and 20, added by this amendment, are directed to the essence of the invention as set forth hereinabove and as set forth at some length at the interview.
It is believed that these two claims, which differ from each other only in that claim 20 includes a switch whereas claim 19 does not, are allowable and that no additional search is required.
Allowance of the seven claims followed and patent No. 3,103,618 issued.
Slater knew of Morton's patent No. 2,896,125 in 1959 and, indeed, finding a sample of the patent assignee's device too big to fit into a wall switch-box, devised a high-low wall switch after Morton that did function, and Slater thereafter started manufacturing the switch under license from the assignee of Morton's patent. Slater had also, and sent to patent counsel along with an early sketch of the circuitry of the dimmer of S.N. 3144, a 76 page General Electric publication entitled "Notes on the Application of the Silicon Controlled Rectifier"; Slater commented in the letter to counsel that "nowhere do they show a controlling circuit for the Silicon Controlled Rectifier using a half wave circuit wherein you do phase shifting and thereby control the output from that of the total half wave to practically zero using only the voltage across the rectifier itself such as I have designed in the circuit drawn on the bottom of this letter. Therefore, I think you should apply for a patent on it." (Italics in original.) At page 28 of the GE notes, a firing method for an SCR (using a fixed resistance and diode in series) is shown apparently using only the voltage across the load; at the lower left hand corner of the page Slater drew a sketch that to him, testifying in 1970, "looks like my starting to figure out making it variable, making a controlling circuit out of it instead of just plain rectifying" by putting a potentiometer (variable resistance) across the SCR so that the gate voltage would be adjustable and the rectifier so made controllable. At pages 41-43 of the GE Notes there was a "phase controlled AC circuit" which, it was suggested, "can be used for such purposes as lamp dimming and temperature control"; the circuit, it was said, "draws on the voltage developed across the SCR's during blockage for the unijunction transistor interbase supply and synchronization." The dimmer circuit shown is elaborated, but in fact it is in one side of the line in series with the load, and it is powered by the voltage across the SCRs. The gating impulses to the SCRs, although controlled through a variable-resistor (potentiometer) and capacitor, are communicated to the gate electrodes by transformer rather than by wired connection. A zener diode is connected in the power input of the time-delay circuit.
Slater's testimony is that he did not utilize the teaching of pages 41-43 of the GE notes: his objective was such a radical miniaturization as would enable him to fit the dimmer switch that he was then trying to develop into the conventional wall switch-box (as he had with Morton), and the transformer in the GE Notes circuit implied size to him; testifying in 1970, he judged that, put off by the transformer, he would not have studied the circuit to glean its full teaching. Nor, evidently, did counsel, for the reference was neither located by the Patent Office nor drawn to its attention by the applicant. (Note, however, that Slater's language to his counsel in the letter of October 23, 1959, emphasizes the absence of the specified half wave circuitry from the GE Notes; pp. 41-43 deal with a full wave dimmer; of course, it operates symmetrically on each constituent half wave.)
The critical pages of the GE Notes sent in Slater's October 1959 letter came to the attention of Slater's patent counsel as "a bombshell" four months after the patent was issued, and a month after GE had commenced an interference with Slater in which GE claimed the right to Claim 1 of the Slater suit patent; the "bombshell," counsel said (as a matter of first impression) "seems to us to be substantially identical to the circuit disclosed in your letter and which moreover was apparently published more than one year prior to the filing of your earliest application." (The latter point was apparently not quite correct in fact; the GE Notes were dated December 1968 but, it may be, they were published some time after that date although before February 6, 1969; the "one year" is irrelevant if GE's circuit was genuinely an anticipation, 35 U.S.C. § 102(a)).
The GE interference terminated on November 20, 1964, with the final granting of Slater's motion to dissolve it on the ground that the device of the GE disclosure did not meet the requirement that the device be one "comprising two terminals for connecting said device in series in one side of the two sides of the power line" and "a variable time delay means energized from said two terminals." On November 1, 1965, patent litigation between Slater and GE was settled, and soon thereafter the application for reissue of the patent was filed.
The application for reissue of patent No. 3,103,618, was filed on November 16, 1965, with amendments for the language of claims 5 and 6 and additional claims 8, 9, 10, and 11. There was only one substantial action by the Examiner and response by Slater, and the Reissue was then granted. The Examiner did not comment on the patentability of claims 1 through 4 (see Rule 176). However, he cited Morton, Tsien, No. 2,981,866, Spira (1) No. 3,032,688, Spira (2) No. 3,061,744 and Davis, No. 3,128,440, and he also cited the GE Notes pages 41-43 and a different page and figure (Ex. 1-A, last item, Fig. 41) from the same publication earlier cited in S.N. 3144 (Ex. 17, Item 9, Fig. 22), and rejected Claims 5 and 6 as unpatentable over (1) the showing in the circuit of the GE Notes, pp. 41-43, of an SCR in series with the load and alternating current source, the time delay circuit connected across the SCR and powered by the voltage across it, and the timing of the gating impulses by means of a variable resistor; and (2) Tsien's "cool dimmer" embedded in an insulating housing adapted to fit into a wall-switch or outlet box and equipped with a manually operable control member extending outside the housing and used to vary the current flowing from source to load ( Tsien provided low power through a capacitor at one rotary switch tap, half power at the next (clockwise) switch tap, full power at the next tap, and "off" at the fourth tap.) Claim 7 was rejected on the circuit shown in the GE Notes and the mechanical switch shown in the time-delay circuit of the static switch circuit of Fig. 41 of Ex. 1-A, last item. Claims 8-11 were rejected on the three references — the circuit of the GE Notes, Tsien and Fig. 41. Morton and Spira (1) were referred to as showing dimming devices adapted to fit into wall outlet boxes and to illustrate external control members. (The Spira (2) and Davis applications post-dated S.N. 3144; in an interference involving S.N. 250,458 and Spira (2), Slater was awarded priority as to the subject matter of claims 1 through 6 of Spira (2).)
Responding to the Examiner, Slater treated only claims 1 to 4 and new claims 9-11 as remaining in the case (issued claims 5-8 having been cancelled), and observed that, since claims 1 to 4 had not been rejected on the prior art, they did not need extended discussion. Claims "9-11" (issued with clarifying amendments, as 8-10) were supported with the opening reminder that in the interview
applicant's invention sought to be covered by these claims was explained to be characterized by (1) a mechanical switch within the housing, which physically opens a circuit to inhibit flow of current to the load, (2) a manually operable variable power control member extending outside said housing, and (3) a mechanical connection between the control element and switch so that the switch is controllable by operation of the control member.
The GE Notes circuit was distinguished as not suggesting a "power control and switch device"; arguing that claims 9-11 "pertain to a wiring device * * * which may be substituted for existing on-off switches or the like in a building wiring circuit" the switches in which "frequently only provide access to one side of the power line," Slater asserted that cheap and convenient continuously variable control of power supply presented an insurmountable problem until his invention. Saying that, "Claims 9-11 do not pertain to a circuit * * * but * * * to a new form of wiring device" unsuggested by the prior art, Slater argued further that his new claims uniquely recite "the switch and a mechanical connection between such switch and a manually operable control member extending from the housing of the wiring device." Tsien's four-position rotary protruding shaft and control knob, Slater argued, could not be combined with the circuit of the GE Notes because that "combination of references is in no sense suggested by the prior art," and Tsien's sort of control member could not be used "to control the time-delay means of the GE publication circuit and also open and close a mechanical switch which is not even disclosed in the GE circuit." The idea of adding the Fig. 41 reference (showing a mechanical switch in the control circuit of a static switch) to satisfy the last point was met by saying that Fig. 41 merely showed that it was conventional to have an off-on switch, but did not suggest coupling it to a manually controllable protruding shaft and knob which also adjusts the degree of (variable) time-delay.
Without further substantive exchange between Slater and the Patent Office, the Reissue was granted under date of December 6, 1966.
The trial evidence on the issue of validity did not go significantly beyond the material that was — ultimately — before the patent office. Leviton did, in addition, draw attention particularly to Freeman, No. 1,604,165 of 1928; the other prior art on which it, strictly, relied was Morton No. 2,896,125, pages 41-43 of the GE Notes and a series of related GE (or Gutzwiller) publications, and Figure 41 from another part of the same publication containing the Ex. 17, Item 9, reference (referred by the Patent Office during the S.N. 3144 prosecution).
Freeman discloses a continuously variable wall-switch light dimmer in which a manually operable control member extends outside a housing containing a variable resistance coil and a contact arm actuated by the manually operable control member; rotating the control member to the counterclockwise extreme breaks contact between the contact arm and the resistance coil and turns the current wholly off; further counterclockwise rotation of the control member turns the current full on without gradual brightening over the whole range of the resistance coil. As Freeman presents a simple variable resistance, it evidently would be in series with the load in one side of the line.
Freeman, however, is a simple rheostat; it would as a practical matter produce so much heat in dissipating power to dim the light, if used for ordinary household dimming, that it would not be tolerable in a wall embedded switch. The importance of Freeman is simply that, like Morton and Tsien (both of which are semi-conductor wall-switch dimmers with manually operable control members extending outside a housing), it presents as long familiar in early 1960 the idea of a dimmer that will be directly substitutable for the common wall switch, and will, of course, have built into the manually operable control-member assembly means for turning the current full on and wholly off. Only Freeman — if it could be used at all under building code and fire-safety restrictions — is continuously variable.
The GE Notes concededly provide a dimmer circuit on which the purely circuit elements of Slater's claims can be read. Interrogatory 8 and the answer to it make that explicit. The circuit of the GE Notes may not have provided either Slater's illustrated circuitry or a suitable set of components and of choices of arrangement for Slater's specific task, although the ingenuity of Leviton's personnel managed to produce and demonstrate a working model of the circuit that substantially accomplished the task of squeezing into the required space allowance; but the circuit component of Slater's claims is phrased with great breadth and largely in terms of "means." The "rectifier" or "semi-conductor control element" need not be an SCR, and while in claims 1 and 2 it must have an anode, cathode and gating electrode, in claims 8 and 9 it need only have two electrodes and be somehow adapted to become conductive on signal in at least one direction. Any kind of variable time-delay means will serve provided only it is energized or powered by the voltage across the rectifier element (or semi-conductor control element). The claimed patentable novelty, that is, does not lie in devising circuitry, but in selecting to the combinations of the claims that one of circuits assumed to be available which will function in one side of the line and yield continuous variability of power supply over a desired range. The other elements of the combinations are essentially mechanical and they come down to means for uniting in one accessible control "member" a means for turning the device off, working it through its continuous-dimming range, and turning it full on.
The difficulty with finding patentable novelty is insuperable. Slater emphasized throughout the prosecution that it was important that his device was suited to be connected in one side of the power line, and that neither the time delay means nor any other of the circuitry of the device was connected across the power source. But that was neither novel, in light of Freeman, Morton, Tsien and the circuit of the GE Notes, nor is there any basis for supposing that it was not an intrinsic connotation of the very idea of a domestic wall-switch dimmer. It is not logically feasible to say that, specifically, Morton did not suggest it, for if one contemplated development of a workable continuously variable dimmer that would be substitutable for an off-on wall-switch, that very concept directed thought to the selection of a controllable semi-conductor circuit that would operate in one side of the line, and such circuitry was available in the public domain. There is no real indication in the record that any difficulty attended the progress of idea from Morton's half-wave rectifier for half-dimming to Slater's variable half-wave rectifier for variable dimming down the scale, or attended the finding of such a circuit, and the fact that only the circuit of the GE Notes portrayed such a circuit must be supposed to be due to the relative infrequency of the need for such circuitry. The readiness with which Slater set about adapting the half-wave static switch of p. 28 of the GE Notes to convert it into a variable time delay circuit (all in series with the load) is indicative of his understanding of the availability of means of doing it. And Figure 148 in the GE Transistor Manual (Ex. 17, item 6) and the accompanying text again illustrate the assumption of the Slater patent application and claims — that phase-shifting (time-delay) circuitry was on hand and compatible with operation of the whole device in one side of the line.
The elements of Claim 1, then, specify the circuitry requirements of the device (without specifying novel circuitry) and the combination of such circuitry with switching that will selectively provide a separate electrical path between the terminals bounding the device. That, in fact, is the combination for which patentable novelty is claimed. The combination does not transcend its constituent elements; the elements enact their familiar roles; together, they usefully link without enhancing or altering their inherent functions or providing a new function. The combination, including the very broadly claimed switching, is suggested by the prior art — if that be a valid test and that must be doubtful — for Freeman, Morton and Tsien adumbrate it, and the implementation of the combinatorial idea of the claim is obvious to one skilled in the art.
It must then be concluded that claims 1 and 2 do not present patentable novelty. It is not, apparently, argued that the addition of the second semi-conductor, a diode of opposite polarity, would confer patentable novelty on the expanded combination, and it would, indeed, appear to be simply additive and not a distinct discovery.
Claim 9 is not different in patentability. Again the circuit elements are generalized, but the choice from presumably available circuits is restricted to semi-conductor circuits that will operate wholly in one side of the power line. The combination embraces in addition the insulating housing of the device and a mechanical switch (arranged to open, selectively, the energization circuit for the time-delay means) mechanically connected with and actuated by a manually operable control member extending outside the housing the operation of which also controls, through the time-delay means, the time when the semi-conductor commences to conduct in each cycle. It will be noted that, broadly, this is the content of elements (R), (S) and (T) of claim 9 and so much of element (U) as refines the definition of element (S) to avoid confusion with element (R); it will be noted, further, that the switch, element (C) of claim 1, and its function of selectively opening a separate electrical path between the terminals of the device, element (E) of claim 1, are dropped in claim 9. Claim 9 is explicit that the device be "for use as a wiring device to be installed in a building" and be in "an insulated housing adapted to fit in an outlet box."
The defect of the claim is the same as the defect in claim 1. The new language and elements do not introduce novelty; they do not point out any specific structure that confers novelty on the combination, or that functions in the whole combination in an unexpected or unobvious way. The new language, rather, introduces "limitations" that do not impart a novelty that would be absent without them, but rather measure the retreat from broader claims in which, if in any, the patentable novelty resides. If the device in the — generally — broader aspect presented in claim 1 has not patentable novelty, claim 9 does not achieve it by using language that restricts the embrace of the claim to its most obvious field of utility, for the new language does no more than that; unobviousness is surely not won for the claim by specifying that the "control member" actuate both the time-delay circuit switch and the timer-element in the time-delay circuit: such specification of the almost inevitable (an "off" position in a graduated control is all but implicit in it) presents an unreal sub-combination and adds nothing of pith to the claim.
It follows that claim 9 is invalid. It does not appear to be argued that claim 8 is different in respect of patentability, and it follows that claim 8 is likewise invalid.
Slater's excellently presented arguments for validity have not been overlooked. The argument that the Section 103 standard must not be treated as embracing advanced electronic learning is beside the point. The patent is not on ingenious electronic circuitry but on a device invoking semi-conductor components in familiar roles for combinatorial use in a wall-switch device. The failure of others, well-equipped to do so, to make the invention despite work on the same problem, or, at least, in the same area, is a powerful argument for unobviousness and forcibly presented, but it must yield to the analysis of the invention against the prior art and that comparison fails to identify the required gap of unobviousness that Slater bridged. The explanation may lie in the hurry of priorities of work in the new semi-conductor field, and the high cost of SCRs in the initial stages. The success of the device, in Slater's hands and in that of the alleged infringers, is a tribute to utility, but against the historical background (including the economics) it does not in this case contribute significantly to the inference of patentable novelty.
The deference due to the presumption of validity and to the absence of telling prior art overlooked in the patent office, and the importance of avoiding the pitfall of finding obviousness only because genuinely inventive insight has discovered the solution in a simplicity that is the essence of the inventive faculty and not a mark of a priori obviousness, these considerations are here very much blunted. The patent office history, detailed above, undermines the presumptions here. As always, the procedures are formally impeccable, but the absence of the GE Notes as a reference in the first stage of the prosecution of S.N. 3144 and S.N. 35,298 gave a direction and momentum to the prosecution that the later disclosure of the GE Notes did not cure; the initial condition of the applications was not recreated, and the easy by-passing of any active reconsideration of the allowance of claims 1 to 4 is an index of the extent of the difference in the treatment of the reference that the belatedness of its disclosure produced.
If the claims had been found valid, the issues of infringement would be resolved in part in Slater's favor. Following the order of Slater's brief, the conclusions would then have been the following.
Leviton 6600, treated as representing the whole 6600 group, is claimed not to infringe because its switch (of element R) is so disposed that it shuts off and connects up the entire device; it is, therefore, arranged so that it can only inclusively and not "selectively open the energization circuit for the said time delay means." It must be said that the argument for treating the Leviton switch element as sufficient to answer the requirement of "selectivity" is conducted to exhaustion and with all the vehemence and skill that can be trained on it, but the language and intention are too plain to support the conclusion. "Selectively" means designed to enable the user to single out the identified circuit for opening. The switch exists in an embodiment that also includes the by-pass switch of Claim 1, though that switch is not an element of claim 9. It may be that here the claim is the blind slave of Figure 1 of the patent and the first presentation of the switch sequence in S.N. 35,298 (Col. 5, ll. 32 et seq.), and that the idea of placing it ahead of the junction between the SCR and the time-delay conductors (as in the parent application S.N. 3144) failed to win through to the claim language, although it was brought into the specification (Col. 5, ll. 51-54). The result, whatever its origin, is plain.
The point seems small for the location seems to contribute nothing to the combination and it has little to do with the substance of claim 9 and the subject matter of the disclosure particularly as amended. But the change in the location and function of the switch is not unreal. The position of the switch in S.N. 3144 signified that in "off" position there was no leak across the SCR such as, theoretically, existed under claim 9 (and Figure 1) when the switch was turned to "off" and the SCR, although in non-conducting status, was still connected between the load and power line.
Claim 9 (typical of claims 8, 9 and 10 so far as the 6600 group is concerned) is not infringed.
Claims 1 and 2 are not urged against the 6600 group.
Leviton 671 is claimed to infringe claim 1 and that is not denied. Infringement of claims 8 and 9 is not insisted upon in light of the concession as to claim 1.
Leviton 671 is claimed to infringe claim 1 and that is not denied. Infingement of claims 8 and 9 is not insisted upon in light of the concession as to claim 1.
The Leviton contention is that the 670 does not reflect element (E) of claim 1 and it argues that element (E) in claim 1 and in claim 2 is satisfied only by the switch member 29 that opens and closes a direct electrical path to the load that by-passes the SCR and time-delay paths. Slater argues that the phrase is not limited to the switch member 29 but extents to the switch that opens and closes, in Figure 8, the circuit through the second reversed-polarity diode added as an element of the combination of claim 2; hence, Slater argues, in the 670, which, like Figure 8 of the patent, contains the reversed-polarity diode circuit and a switch analogous to the switch 101 in Figure 8, the requirement of element (E) of claims 1 and 2 is satisfied; the argument in part is sought to be supported by noting that S.N. 3144 had used the term "a direct electrical path" (and had no second diode circuit) while S.N. 35,298 uses the term "separate electric path" (and always contemplated a second diode circuit); it is argued that "direct" may be aptly descriptive of the path controlled by the switch element 29 but that "separate," when used in implicit contrast to "direct," accommodates a path that is either direct or through a diode.
Slater's argument must be rejected on the language and intention of elements (C) and (E) and the description of Column 5, lines 32 et seq. Leviton's interpretation avoids using the 102 switch element to satisfy both elements (E) and (I), and is faithful to the simple meaning of elements (C) and (E).
Since the 670 has no direct by-pass circuit and switch to apply full power to the load, and can apply full power only through its rectifiers, it does not infringe claims 1 and 2.
Leviton argues that enforcement of the reissue patent should be refused because of the nondisclosure of the GE Notes in the original applications and the belatedness of the disclosure to the patent office after the reference was adverted to by Slater's patent counsel. Certainly the sequence of event is unfortunate, but the evidence does not impugn the good faith of Slater, and Saul Slater's evidence on the precise point was impressive in its candor and explicitness.
The evidence and argument have not dealt with claims 3, 4 and 10, although claim 10 has been referred to. The validity of claim 3 would seem to depend on the validity of claim 1, but the additional elements of claim 4 have not been argued or embraced directly in the evidence, and it does not seem well to consider the thrust of claim 10 without record support. In these circumstances the determination of invalidity is limited to claims 1, 2, 8 and 9. As to infringement, it is determined that if claim 1 were valid Leviton's 671 device would infringe, but that the Leviton 670 and the group of Leviton 6600 devices would be non-infringing.
As in every invalidity case, the puzzling search for a touchstone technique to determine unobviousness as a fact, Graham v. John Deere Co., 1966, 383 U.S. 1, 17-19, 86 S.Ct. 684, 15 L.Ed.2d 545; Anderson's-Black Rock, Inc. v. Pavement Salvage Co., 1969, 396 U.S. 57, 60, 62, 90 S.Ct. 305, 24 L.Ed.2d 258, leaves a residue of dissatisfaction that the process of decision does not emanate in a self-evidently objective determination measured out upon a verifiable scale of obviousness particular to a precisely delimitable "art." But John Deere warned that the comparable frames of reference are those of negligence and scienter, and, in final analysis, that remains the difficulty, that the standard must seem Protean because it must retain adaptability for use in a limitless range of "arts."
An award of counsel fees in this case does not appear to be warranted.
The stipulated facts are found as set forth in the Pre-trial Order of March 3, 1970.
It is
Ordered that on the findings herein the Clerk enter judgment in favor of plaintiff and against defendant that
1. Claims 1, 2, 8 and 9 of United States Reissue Patent No. 26,119 are invalid;
2. Claim 1 of said Reissue 26,119 would be infringed by Leviton Full Range Dimmer Switch Model No. 671 if Claim 1 of such Reissue 26,119 were valid;
3. The Leviton Dimmers Models Nos. 670, 6600, 6670, 6673, 6681 and 6683 would not infringe claims 8 and 9 of said Reissue 26,119 if claims 8 and 9 were valid;
4. Defendant Slater Electric, Inc. take nothing upon its counterclaim for infringement of United States Reissue Patent No. 26,119 and that such counter-claim is dismissed on the merits.