OPINION
CONNER, District Judge;
This is an action for alleged infringement of three United States Patents owned by plaintiff and relating to power supplies for high-intensity lighting systems used in graphic arts photography.
This opinion incorporates the Court’s findings of fact and conclusions of law pursuant to Rule 52(a), F.R.Civ.P.
FINDINGS OF FACT
The parties and issues
1. Plaintiff, Berkey Photo, Inc. (“Berkey”), is a Delaware corporation having its principal place of business at Berkey Avenue, Woodside, New York.
2. Defendant Theimer International Corporation (“Theimer”) is a New York corporation which at the time of the institution of 70 Civ. 3216 had its headquarters at Mamaroneck, New York.
3. Defendant Klimsch-Repro, Inc. (“Klimsch”) is a New York corporation that does business in the State of New York and elsewhere, with headquarters at 100 Avenue of the Americas, New York City, New York.
4. The action against Theimer, 70 Civ. 3216, was commenced July 28, 1970. The action against Klimsch, 72 Civ. 3216, was filed August 2, 1972. The two actions were consolidated for trial. Shortly before trial, defendants moved for a summary judgment dismissing the complaints in both actions as to two of the patents in suit on the ground of non-
infringement. The Court denied the motion because of the existence of “genuine issues as to material facts” but, “because the affidavits filed in support of the motion create substantial doubt” as to infringement, directed a separate trial of that issue. The five-day non-jury trial of that issue took place commencing October 29, 1974.
5. Berkey is engaged,
inter alia,
in the business of manufacturing and selling electrical lighting equipment for graphic arts purposes. It is the owner of all right, title and interest in U.S. patents No. 3,174,076 issued in the name of Maksymilian Miehalski; No. 3,309,566, also issued in the name of Miehalski; and No. 3,309,567 issued in the joint names of Robert A. Flieder and Miehalski. It is the successor in interest of American Speedlight Corporation, the original assignee of these' patents.
6. The 1970 action against Theimer alleges infringement of Miehalski patents Nos. 3,174,076 and 3,309,566 because of the sale by Theimer of equipment sold under the trademarks “Xenolux” and “Violux.” Theimer ceased business operations in 1972. The 1972 action against Klimsch charges infringement of all three of the patents in suit; it was instituted when Berkey learned that Klimsch had taken over the sale of the “Xenolux” and “Violux” equipment previously sold by Theimer.
7. During the trial, defendants conceded their infringement of the 3,309,566 patent. Thus, only the infringement of the other two patents in suit need be considered here.
The Miehalski Patent No. 3,174,076
8. The ’076 patent was issued March 16, 1965 on an application filed July 27, 1962. ’ It is directed to a power supply circuit for an electric discharge lighting device, such as a xenon tube, the stated objective being to provide an output voltage which is substantially constant despite fluctuations in the line voltage input, thereby insuring a substantially constant light output, as is desirable in graphic arts photography for uniform results.
The patent discloses a power supply circuit including an autotransformer having a primary winding connected across the power line and a secondary winding connected in series . with the primary and wound on the same steel core; a capacitor connected across the series-connected transformer windings; and gas-discharge lamp and a saturable reactor connected in series across the capacitor. The saturable reactor serves as a switching and pulse shaping device to control the intermittent discharge of the capacitor through the lamp.
9. A power supply circuit having the same topology, that is, containing similar components connected in the same way, was disclosed in Figure 4 of the prior Wiley U.S. patent No. 2,938,149, issued May 24, 1960 on an application filed May 2, 1957. Wiley was referred to in the specification of the ’076 patent and cited by the Patent Office Examiner in initially rejecting Michalski’s claims. Michalski distinguished his claimed invention from the disclosure of Wiley on the basis that in his circuit the capacitor and the transformer formed an LC Circuit which was resonant at the frequency of the input voltage (60 cycles), which contributed to the uniformity of output voltage despite fluctuations in the amplitude of the input voltage.
10. This feature of resonance at the operating frequency is thus an essential characterizing feature of the invention disclosed and claimed by the ’076 patent. The patent repeatedly emphasizes this feature. Even the title of the patent is “Electric System for Discharge Device Utilizing Resonant Circuit to Provide Constant Current Output.” And the specification of the patent states that the patent’s objective of “[c] onstant light output from the discharge lamp even under considerable variations in the voltage of its input supply * '* * is accomplished by providing an electric system for the discharge lamp including
a capacitor in a resonant circuit with an inductance * *
11. Of the 14 claims of the patent, all but Claim 2 are in suit. All of the claims in suit are limited, by one form of expression or another, to a circuit which is resonant at the input frequency, thereby to provide substantially constant output. Claim 4, for example, requires:
“a capacitor, an inductance connected in a resonant circuit with the capacitor for resonance at a predetermined frequency, the voltage of which may be fluctuating, means for impressing an alternating current of the predetermined frequency on the resonant circuit, * * * whereby the output of the discharge device is maintained substantially constant irrespective of the variations in the voltage of the alternating current.”
Claims 5, 13 and 14 contain similar language.
Claims 1, 3 and 6 through 12 also require a “resonant circuit,” and further specifically recite, in substance, the standard definition of resonance — that is, that the capacitive and inductive reactances are substantially' equal; thus, being of opposite algebraic sign, they mutually cancel so that the total circuit reactance is zero and the current is at a maximum (limited only by the applied voltage and the circuit resistance). Claim 1, for example, recites:
“the reactance of the series connected windings at least approximating the reactance of the capacitor so that a resonant circuit is formed therewith, the resonant circuit operating at a frequency equal to the frequency of the voltage impressed on the first winding so that the potential across the capacitor is maintained substantially constant irrespective of fluctuations in the input voltage * * * whereby the output of the electric discharge device is substantially constant irrespective of fluctuations in the input voltage.”
Claims 3 and 6 through 12 contain language which is either identical or of similar import.
12. Plaintiff argues, in effect, that since any circuit which contains both inductance and capacitance is “resonant” at
some
frequency, and since every circuit may be said to be “operating at a frequency equal to the voltage impressed” as recited in Claims 1, 3 and 6-12, these claims are infringed whether or not the operating frequency and the resonant frequency are substantially equal.
Ingenious as that argument is, it is inconsistent with any reasonable interpretation of the ’076 patent. If Claims 1, 3 and 6-12 do not require that the resonant frequency and the operating frequency be substantially equal, all of the portions of these claims referring to the substantial equality of the inductive and capacitive reactances and to the operating frequency become meaningless surplusage. Moreover, as discussed hereinafter, according to the patent’s own teaching, the claimed result of substantially constant output could not otherwise be achieved.
13. Finally, as previously explained, resonance at substantially the operating frequency is essential to patentability. If the claims are interpreted to eliminate this requirement, there is nothing left in the claims to distinguish them patent-ably from Wiley. Wiley’s transformer winding and capacitor are resonant at “some” frequency, and his circuit, like all others, is “operating at a frequency equal to the frequency of the voltage impressed.”
14. In its Reply Brief, plaintiff argues that Wiley does not disclose a “saturable transformer.” However, none of the patent claims recites a “saturable transformer” in those terms. Instead, certain of the claims describe the means by which the transformer is made saturable. For example, Claims 5, 6, 10, 12 and 14 recite a “magnetic shunt” between the transformer windings, and Claims 7, 8 and 9 recite a “high leakage reactance path” in the transformer core, which is a less specific way of describing the same thing.
Wiley specifically describes the auto-transformer 21 in his Figure 4 as “loose
ly coupled” and states that the resulting “leakage reactance” serves the function of a “regulating reactor” as previously described. This is old art. The prior U.S. patent No. 2,143,745 issued to J. G. Sola on January 10, 1939 discloses a “constant potential transformer” which includes a “high leakage reactance path” formed by a magnetic shunt 12b similar to the shunts 10 disclosed in the ’076 patent.
Therefore, all of the claims must be construed in the manner clearly intended by the patentee and represented to the Patent Office Examiner in order to secure the allowance of the claims: i. e., as requiring resonance at substantially the operating frequency.
15. The ’076 patent is charged to be infringed by defendants’ Xenolux system. The accused system includes a power supply which incorporates all of the circuit elements recited in the claims of the ’076 patent, in suit, connected in the manner recited. The only questions as to the infringement of these claims are therefore (1) whether the Xenolux circuit is resonant at the operating frequency and (2) whether the voltage across the main storage capacitor and the light output are substantially constant despite fluctuations in the input voltage.
16. With respect to the first question, plaintiff contends that, even if the claims of the ’076 patent require that the circuit be resonant at the operating frequency, defendant’s Xenolux system satisfies this requirement. This contention is three-fold:
A. Plaintiff argues that although “The terms ‘resonant frequency’ and ‘resonant condition’ are commonly used to identify a single frequency, *
*
* the term ‘resonance’ is used to indicate a band of frequencies.” Even assuming the validity of that highly dubious contention, it is of no avail to plaintiff since all of the claims of the ’076 patent use the admittedly narrower term “resonant” and not the allegedly broader term “resonance.”
B. Plaintiff further contends that, because the transformer of the Xenolux unit is saturable, so that its inductance varies in inverse relation to the current, and because the input alternating current varies in sinusoidal manner from zero to a maximum and back to zero again every half cycle, the inductive reactance of the transformer and the capacitive reactance are “precisely equal at the predetermined frequency at least once during each half cycle.” Plaintiff argues that this instantaneous equality is sufficient since the claims do not require “ ‘resonance’ continuouslyto the contrary, the specification of the patent teaches, at Column 4, lines 49-56, that the resonant frequency changes when the inductance of the saturable transformer decreases due to an increase in the input voltage. However, the patent makes clear at Column 4, lines 32-40 at the “predetermined flux density” (corresponding to the selected nominal input RMS voltage)” the reactive inductance [obviously meaning the inductive reactance] will equal the capacitive reactance of the circuit. Therefore, the circuit will be at resonance * * The specification further states at Column 4, lines 49-56 that when the input RMS voltage increases, “the circuit has passed resonance and the Q value is lower,” with the result that “further input voltage increase does not result in proportionate voltage increase across capacitor 20.”
Although the specification continues by asserting that, because of the resultant lowered Q, “resonance is maintained,” it is obvious that the desired regulation of the output voltage is achieved in substantial part by virtue of the fact that the circuit, which is resonant at the nominal input voltage, becomes non-resonant when the voltage increases, and this detuning of the circuit reduces its transfer efficiency sufficiently to nullify the increase in input voltage and keep the output voltage substantially constant. Equally obviously, if the circuit were designed so that its resonant frequency is
lower
than the 60-
cycle input frequency at the nominal input voltage (e. g., 220 volts), if the input voltage increased above the nominal value, the resulting reduction in the inductive reactance would raise the resonant frequency, bringing it closer to the operating frequency, enhancing the transfer efficiency of the circuit and
aggravating
the effect of the input voltage increase, notwithstanding the fact that twice during each half cycle the instantaneous value of the constantly varying inductive reactance may equal the capacitive reactance. It therefore appears that what is significant in the regulatory function is not the rapid swing in the inductive reactance which takes place during each quarter cycle, or 240 times per second, as plaintiff claims, but the much slower changes in the
average
inductive reactance which result from fluctuations in the input
RMS
(root-mean-square or effective average) voltage. This is the teaching of the ’076 patent.
Thus, the “resonant” condition reffered to in the claims is the substantial equality of the effective values of the inductive and capacitive reactances over a multi-cyclical period, and this requirement is not satisfied by instantaneous equality within each quarter cycle.
C. Finally, plaintiff argues that defendants’ Xenolux unit must have a resonant circuit because it achieves voltage regulation and voltage multiplication.
That argument is in part factually wrong and in remaining part a
non sequitur.
As discussed more fully hereinafter, the Xenolux circuit does not achieve nearly the degree of regulation of light output claimed for the patented resonant circuit and, indeed, is no better in this respect than the acknowledged prior art. The degree of voltage regulation it does achieve is attributable largely if not entirely to the saturability of the autotransformer. Finally, the voltage multiplication is attributable principally to the step-up ratio of the autotransformer. The fact that some further voltage enhancement may be achieved because the LC circuit is operating on the skirt of the resonance curve does not mean that the circuit is “resonant” in the sense of employing the principle taught by the patent.
17. The circuit of defendants’ Xenolux device is not “resonant” within the meaning of the claims. I have not been persuaded by defendants’ argument that in the operation of their Xenolux unit, resonance is conventionally applied to unpacitor is fully discharged through the xenon lamps each half cycle, so that there is no energy left to be returned to the inductance. It is true that the term resonance is conventially applied to unloaded or moderately loaded LC tank circuits in which the electrical energy is transferred back and forth between the capacitor and the inductor each half cycle at the frequency at which the capacitive and inductive reactances are equal. However, interpreting the term in the light of the disclosure of the ’076 patent, it is clear that it must apply to a circuit in which the capacitor is substantially discharged through the lamp every half cycle, because this occurs in the circuit disclosed in the ’076 patent, as clearly shown in Figure 6, and as described at Column 5, lines 55 et seq.
18. With respect to the remaining question whether, in defendants’ Xenolux unit, the light output is maintained substantially constant despite fluctuations in the input voltage, tests of the unit indicate that for each one-volt (or roughly 0.5%) change in input voltage, there is approximately a 1.5% change in light output. This is a far greater variation in light output than the 0.5% change in light output for each 1 volt change in input voltage claimed in the patent at Column 5, lines 38-48, and indeed, is the same as the 1.5% change in light output achieved by the prior art devices, such as that of the Wiley patent, which the ’076 patent criticized and sought to improve.
19. Plaintiff argues that defendants fail to achieve the maximum regulation of which their circuit is capable only be
cause the circuit is overloaded so that the optimum benefits of resonance are not obtained. Plaintiff’s expert testified that the Xenolux unit, which is used with a 6-kilowatt load (four 1500-watt lamps), could much more closely approach the degree of regulation claimed in the patent if it were used with a 4-kilowatt load. Plaintiff argues that defendants have “placed a six-pound load in a four-pound bag” and that one cannot avoid infringement merely because he has thus “impaired the function of the device.”
However, the evidence established that the Xenolux device could not be operated at a lower power level merely by reducing the number or the power rating of the lamps in the load because, with any lamps now commercially available, this would merely burn out the lamps. Thus, used in any manner in which the unit was intended to be or can be used, it does not satisfy the requirement of the claims for substantially constant light output.
20. Plaintiff argues that defendants are estopped to contend that their Xenolux unit does not have substantially constant light output because, in their sales literature, defendants assert that their Xenolux unit provides “relatively constant light output with normal line voltage fluctuations.”
This contention is without merit. Infringement cannot be established by “puffing” statements in a defendant’s sales literature, particularly where the statements are couched in vague, comparative terms which do not duplicate the wording of the claims.
21. Defendants’ Xenolux unit does not incorporate the invention described in any of the claims of the patent because the circuit is not resonant at the operating frequency and because it does not provide substantially constant light output within the meaning of the claims.
The Flieder et al. patent No. 3,309,567
22. The Flieder et al. patent was issued March 14, 1967 on an application filed October 22, 1965 as a continuation-in-part of two earlier applications filed respectively on July 25, 1960 and July 27, 1962. It discloses and claims a starter circuit for an electric discharge device of the pulsed type, such as a xenon lamp. The starter circuit is used in combination with a power supply, for example, of the type covered by the Michalski ’076 patent, to supply high-power pulses during the first few seconds of operation to cause the lamp to fire; once the lamp has ionized, a lower voltage is required to keep it conducting, and the starter circuit is turned off by an automatic tuner switch, leaving only the power supply to furnish the lower energy pulses required for the continuing operation of the lamp.
The starter circuit consists of a high-frequency relaxation oscillator coupled to the power supply circuit through a saturable step-up transformer. The oscillator circuit includes a cold cathode thyratron tube whose control grid is connected to the midpoint of an RC circuit (a resistor and capacitor in series) bridged across the plate and cathode of the thyratron and across a DC power supply. Also connected in series from the plate to the cathode are a triggering capacitor and the primary winding of the saturable coupling transformer. The triggering capacitor is charged through a current limiting impedance or choke in order to limit the current input to the starter circuit.
As is well understood in the art, this circuit will oscillate at a frequency dependent upon the time constant of the RC circuit. The values of the resistor and capacitor therein are chosen so that the thyratron tube discharges the triggering capacitor at a rate of 120 to 600 pulses per second (or 2 to 10 pulses for each cycle of the line voltage). These pulses are stepped up in voltage by the coupling transformer and applied across the tube to cause it to fire.
23. There is nothing novel about the oscillator circuit
per se.
What is asserted to be patentable is the use of such a circuit as a starter for a pulse dis
charge lamp, wherein the oscillator pulses “between two and ten times during at least each alternate half cycle of the low frequency source [the line voltage].” (The discrepancy between the disclosure in the specification of 2 to 10 pulses per cycle, and the recital in the claims of 2 to 10 pulses per
half cycle
is not explained, but is of no consequence to the issue of infringement).
24. Plaintiff charges that the ’567 patent is infringed by defendants’ Violux lighting system. The Violux unit likewise incorporates a high-frequency pulsing starter circuit which is coupled to the power supply circuit through a step-up transformer; moreover, the starter circuit oscillates at a frequency of 3 pulses for each positive half cycle of the power source. However, the power supply circuit itself is entirely different, since it is designed to operate a high-pressure metal halide (mercury vapor) tube which, once started, requires a continuous, non-pulsed current, preferably of substantially sinusoidal waveform.
25. However, Claim 1, the only independent claim, of the ’567 patent recites a number of other elements or features which the defendant contends are lacking in its Violux device:
(a) “a main capacitor connected in series with the operating circuit current limiting impedance across said low frequency [power] source,”
(b) “an operating circuit [power supply] current limiting impedance,”
(c) “a saturable reactor including a main [secondary] winding connected in series with said lamp across said [main] capacitor,”
(d) “the triggering capacitor and the auxiliary [primary] winding [of the saturable reactor] connected in series with the starting circuit current limiting impedance across the source of potential for the charging of the triggering capacitor,” and
(e) “a resistance — capacitance circuit means * * * to effect closing of the electronic switch [conduction of the thyratron] between two and ten times during at least each alternate half cycle of the low frequency [power] source.”
Defendants are correct insofar as concerns all but item (e).
26. Insofar as concerns item (a), the Violux circuit has no main capacitor; its metal halide lamp is energized directly from the secondary of the autotransformer by continuous alternating current ; thus, no main capacitor is required to accumulate a charge for pulse discharge through the tube.
The capacitor “Cl” in the Violux circuit is connected across the primary of the autotransformer (not the secondary as shown in the schematic diagram); it serves merely to correct the power factor. Although the capacitor “C2” is connected across the secondary winding of the autotransformer, it is a small (0.5 microfarad) capacitor which serves merely as a return path for the starter pulses (bypassing the secondary).
27. With respect to item (b), since the Violux circuit has no main capacitor, it neither requires nor has a current limiting impedance to limit the input current to the capacitor.
28. As for item (e), since the Violux lamp is not pulsed, it neither has nor requires a saturable reactor for switching and pulse shaping. Its coupling transformer “T2” is non-saturable and provides substantially linear response.
29. As for item (d), the triggering capacitor and the primary winding of the coupling transformer are connected in
parallel,
rather than in series, across the source of charging current. However, this is matter of no consequence since they are connected in
series
with each other through the thyristor switching device when it conducts, thereby discharging the capacitor through the primary to create a starting pulse. Thus the Violux circuit incorporates the full equivalent of the connection recited in item (d).
30. As for item (e), the Violux starting circuit generates approximately three pulses for each
positive
half cycle,
and none during the negative half cycles. Defendant argues that Claim 1 of the '567 patent requires that from 2 to 10 pulses be generated
every
half cycle. While it is true that Figures 3 and 4 of the ’567 patent show that the disclosed circuit generates pulses on both positive and negative half cycles, Claim 1 is clearly not limited in this respect, but requires only the generation of from 2 to 10 pulses “during at least each
alternate
half cycle.” The positive half cycles are obviously “alternate” half cycles, and the Violux starter circuit accordingly provides, during the interval of its operation, between 2 and 10 pulses each alternate half cycle.
31. Therefore Claim 1, and all the other claims of the '567 patent (which are dependent upon Claim 1 and thus incorporate by reference all of its limitations), are not readable on the Violux device.
32. Plaintiff argues that the '567 patent is nevertheless infringed under the doctrine of equivalents. However, the Violux circuit includes neither the main capacitor nor the current limiting impedance called for in the patent claims, nor any other components which perform similar functions; indeed, these functions are wholly irrelevant to the type of lamp employed in the Violux system.
DISCUSSION AND CONCLUSIONS OF LAW
1. This Court has jurisdiction of the subject matter and of the parties.
2. The '076 patent is not infringed by defendants’ Xenolux system.
Plaintiff’s argument that the “substantially constant light output” called for in the patent claims could be achieved by the Xenolux circuit if it were operated with a four-kilowatt rather than a six-:kilowatt load is not persuasive. Infringement cannot be established by proof that the accused device
might
be used in the manner described in the claims if such use has not actually occurred, or at least has been suggested or anticipated by the defendant. Hap Corp. v. Heyman Mfg. Co., 311 F.2d 839, 843 (1st Cir. 1962); Marvin Glass & Associates v. DeLuxe Topper Corp., 284 F.Supp. 558, 562 (S.D.N.Y.1967).
While statements made in defendants’ advertising are evidence which a Court should consider in determining whether defendants’ accused device operates in the manner described in the patent claims, Johnson & Johnson v. Carolina Lee Knitting Co., 258 F.2d 593, 598-599 (4th Cir. 1958); Gibbs v. Triumph Trap Co., 26 F.2d 312, 314 (2d Cir. 1928), such statements do not es-top the defendants or bind the Court, particularly where, as here, they are merely “puffing” claims of desirable attributes, couched in comparative or qualitative terms.
3. The '567 patent is not infringed by defendants’ Violux system.
Plaintiff’s argument that the patent is infringed under the doctrine of equivalents is misplaced. Each of the patent claims is directed to a combination of elements. Every element recited in a combination claim is assumed to be material and the claim is not infringed unless the accused device includes each of the elements, or its functional equivalent. Reiner v. I. Leon Co., 324 F.2d 648, 649 (2d Cir. 1963); Kennatrack Corp. v. Stanley Works, 314 F.2d 164, 167 (7th Cir. 1963); Alex Lee Wallau, Inc. v. J. W. Landenberger & Co., Inc., 121 F.Supp. 555, 557 (S.D.N.Y.1954).
In drafting combination claims, the patent applicant includes elements at his peril. A court is powerless to relieve him of the mistake of reciting elements which are superfluous to patent-ability; his only remedy is by reissue, and even this must be done within two years after the original grant. 35 U.S.C. § 251; I Deller, Patent Claims, § 42, (2d ed. 1971), and cases cited therein.
Moreover, no single element or group of elements can be deemed to represent the “heart” or “gist” of the invention, so that infringement may be found despite the omission of other, insignificant elements from the accused
device. Aro Mfg. Co. v. Convertible Top Co., 365 U.S. 336, 344-346, 81 S.Ct. 599, 5 L.Ed.2d 592 (1961).
Thus, in considering the application of the doctrine of equivalents to combination claims, it does not suffice to argue that the overall combination of elements present in the accused device is the functional equivalent of the overall combination recited in the claims; instead, the determination of equivalency must be on an element-by-element basis —the Court must find each recited element, or its equivalent, in the accused device. As stated in Sears, Roebuck & Co. v. Minnesota Mining & Mfg. Co., 249 F.2d 66, 68-69 (4th Cir. 1957):
“the doctrine of equivalents has no application where one of the elements called for is entirely omitted.”
The Complaint must be dismissed as to the Michalski patent No. 3,174,076 and the Flieder et al. patent No. 3,309,567.