Opinion
Nos. 4622, 4709.
May 16, 1932. Rehearing Denied July 12, 1932.
Appeal from the District Court of the United States for the Eastern District of Illinois; Walter C. Lindley, Judge.
Suit by Ephraim Banning against the Hartman Furniture Carpet Company for infringement of patent. Decree was entered granting plaintiff partial relief [ 49 F.2d 331], and defendant appeals.
Reversed, with directions.
Clifton V. Edwards, of New York City, and George E. Mueller, of Chicago, Ill., for appellant.
F. Allan Minne, Wm. Nevarre Cromwell, and Franklin M. Warden, all of Chicago, Ill., for appellee.
Before EVANS and SPARKS, Circuit Judges, and WOODWARD, District Judge.
Appellee brought suit to restrain the infringement of four patents and to recover damages for their past infringements. The patents sued upon were No. 1,680,370 covering a radio supply unit, No. 1,682,492 covering a radio battery eliminator, No. 1,682,778, and No. 1,699,523. All patents were issued to P.E. Edelman and were by him assigned to appellee before suit was begun. Appellee did not press his suit on patent No. 1,682,778, and the court held the claims of patent No. 1,682,778 relied upon to be invalid.
The court upheld claims 1, 2, and 3 of No. 1,680,370 and claims 3 and 4 of No. 1,682,492, sustained appellee's motion to dismiss his complaint without prejudice as to patent No. 1,699,523, and dismissed the suit as to patent No. 1,682,778.
Appellant denies validity and infringement of the claims of both of the patents sustained.
The date of Edelman's discovery of his earlier patent is sharply controverted, and its exact fixation is of vital importance in view of the issuance of other patents prior to his filing date. The court found that "Edelman first conceived his invention defined in claims 1, 2, and 3 of patent No. 1,680,370 on April 10, 1922 * * * and immediately thereafter set out to reduce the same to practice, the first working device having been built by the 20th day of April, 1922. * * *"
After the appeal was perfected, appellant sought leave from this court to apply to the District Court to reopen the hearing and to submit further evidence upon this issue of the date of Edelman's discovery. Such leave was granted, but upon condition that the District Court prescribe the terms and conditions upon which the rehearing might be allowed. The District Court thereupon granted appellant leave to reopen the hearing and permit the introduction of further evidence, but upon condition that the witnesses be produced and examined in open court. A hearing was thereupon had, at the conclusion of which the court filed a memorandum, the material portions of which are as follows:
"Without discussing the details of such evidence, I am satisfied that nothing therein included is sufficient to warrant alteration or modification of the findings and conclusions heretofore adopted. Though there appear in the record some sharp conflicts so far as parole testimony is concerned, I am of the opinion that such controversies are explainable upon the ground that human recollection as to events ten years old, at the best, is faulty and subject to error. The documents submitted by plaintiff directly bearing upon the issue of the date of invention * * * impel the court to the findings and conclusions heretofore adopted. There will be no modification or alteration thereof."
The decree was re-entered, and a second appeal from such interlocutory decree was taken. The two appeals were presented and heard at the same time.
The first question which necessitates review is the above finding that the date of Edelman's discovery, disclosed in patent No. 1,680,370, was April 10, 1922. This finding must be accepted. There is evidence to support it. The trial judge, upon the second hearing, had the benefit of seeing and hearing the witnesses testify in open court. As there was but a single issue to be tried, the parties were fully prepared. Discrepancies in the testimony of certain witnesses on this and the previous hearing appeared. They were, however, as the trial judge observed, or at least might have been, traceable to the uncertainties of human memories. Certainly, the District Judge gave both parties full opportunity to be heard, and he was in far better position than we to weigh the oral testimony. While the burden was upon appellee to carry back his effective date to April 10, 1922, we are unable to say that the judge failed to place upon him the proper burden, or to say that his evidence did not overcome this burden. It is true appellant has made a rather persuasive attack upon Edelman's story as to the date of his invention. The discrepancy in the testimony of plaintiff's witnesses and the conflict between Edelman's recollection of dates and that of his former employees have been duly considered. However, in view of Edelman's positive statement, the finding becomes one that is extremely hard to upset on appeal. Uihlein v. General Electric Co. (C.C.A.) 47 F.2d 997, 1002.
Accepting then, as we do, April 10, 1922, as the effective date of appellee's discovery, much of the prior art which appellant cited is eliminated.
With this fact issue settled, we approach the study of the claims and the specifications with only such art as existed April 10, 1922, for their background. The opinion of the District Court, 49 F.2d 331, describes appellee's power pack, its operation, and its purposes. It greatly shortens this opinion, for repetition of the very satisfactory statement there appearing would be inexcusable. In view of that opinion our consideration of the arguments presented on this appeal and applicable to this patent may well be limited to the single contention that the combination of elements described in each of the three claims under consideration spelled aggregation instead of a patentable combination. This urge of appellant was not passed on by the District Court in its opinion.
Edelman generally described the invention in this language:
"This invention comprises certain new and useful arrangement of co-acting electrical elements whose generic object is to provide a small compact electrical convertor unit having a plurality of output circuits suited for furnishing power supply to the cathode heating and the anode to cathode circuits of radio vacuum tubes, provided with means for eliminating objectionable hum in said output circuits, provided further with a grounded enclosure for said electrical elements whereby induction therefrom with respect to the ground is prevented from reaching radio apparatus to which said unit is to be connected. More specific objects of this invention provide a common transformer as primary source for all output circuits whereby the variations in current draw from each of said circuits tend to cooperate in the regulation of the combined output of said transformer. Also choke coils are distributed on both sides of the output pilot conductors whereby remaining fluctuations of current in the output due to capacity leakage with respect to the ground potential are eliminated. Further the output potential intended for plate supply to vacuum tubes of a radio set is adjustable via series resistances which do not comprise a shunt load limiting the output when a rectifier of small milliamperage output is used in the power unit, and do not cause the loss and voltage drop that a shunt potentiometer type of load regulator would. A further specific improvement comprised in this invention is that the storage elements and the choke coils can now be made much smaller in size than formerly required due to the operation of the filter systems starting at source end with much higher potential than at output end, each successive stage operating at a lower potential than the preceding stage, and in the case of the filament supply output circuit, the first storage cells having a higher output terminal voltage than the next cells of the series, thereby overcoming the resistance effect of smaller size of choke coils employed, so that a compact unit of small dimensions is obtained. The grounded enclosure for the elements also assists in dissipation of heat arising during operation. Further, to avoid radio frequency currents travelling back thru the pilot conductors of output circuits to cause undesired regenerative feedback effects in a radio set to which this unit supplies power, the end terminals of said power unit are shunted with by-pass condensers, and the terminals are provided for universal use whether the connected radio set requires B minus to A plus or B minus to A minus, binding post connection."
Three claims are involved. They read:
"1. A radio power supply unit comprising a transformer, means to connect said transformer to a source of alternating current, a plate current supply fed by said transformer, a series cascaded filter connected to said plate current supply and including adjustable series resistances to adjust the output potential thereof and to prevent the sudden discharge of current therefrom, a filament current supply fed by said source of alternating current, and a metallic container for enclosing said plate and filament current supplies and grounded to one of said supplies.
"2. In a radio power supply unit, the combination of a transformer, filament and plate current supplies energized from said transformer, and means for suppressing magnetic lines of force extending from said transformer and associated apparatus, comprising a metallic container for enclosing said transformer and associated apparatus, said container being connected to one of the terminals of the filament current supply.
"3. In a radio power supply unit, the combination of a transformer, a rectifier connected thereto, a filter circuit operatively connected to said rectifier, said filter circuit having a plurality of output terminals, and means for reducing the potential at the terminals of said filter circuit and limiting the rate of discharge of current therefrom, comprising resistances connected between the filter circuit and one or more of the output terminals."
It is apparent from a reading of the claims that each combination describes an apparatus to harness, control, regulate, change the gait of, slow down, or step up that miracle of modern times — the wonderworker electron — in its ceaseless and irresistible activity. True, the regulation and control of a current of electricity are in this case for the purpose of using it in a radio set; that is, in a device adapted to receive undulations similar in form to sound waves and to translate them into sound waves as perfectly as possible. Nevertheless, the elements of the combination describe apparatus through which runs a current of electricity. It is, roughly, the same current which entered the first element that comes out of the last element. It has been at one place "stepped up" inductively and at another its voltage has been varied by resistances. It has been changed from an alternating to a direct current. Its undulations have been smoothed out, and the even flow of the incoming voice current undulations, is the boast of its producer.
But how has this been accomplished? By a combination of elements coacting, one on the other, or by successive steps through old means, each producing its result independently of the other? We think the answers to these questions must be unfavorable to the inventor.
The differences between aggregation and patentable combination have been defined times without number. Siekert Baum Stationery Co. v. Stationers Loose Leaf Co. (C.C.A.) 51 F.2d 326; Palmer v. Corning, 156 U.S. 344, 15 S. Ct. 381, 39 L. Ed. 445; Grinnell Washing Machine Company v. E.E. Johnson Company, 247 U.S. 426, 38 S. Ct. 547, 62 L. Ed. 1196; Powers-Kennedy Contracting Corp. v. Concrete, etc., Co., 282 U.S. 175, 51 S. Ct. 95, 75 L. Ed. 278. Particularly interesting is the opinion in Sachs et al. v. Hartford Electric Supply Co., 47 F.2d 743 (C.C.A. 2).
A few tests are decisive.
A new combination of elements, if it produces new and useful results, may be patentable, although all of the elements of the combination were old and in common use before the new combination was conceived. In other words, a new combination may consist of an association of elements one or more of which are new, or it may consist of an association of elements all of which are old but arranged in a new or novel manner. The discovery may be (not always) patentable. The result, however, to a certain extent supplies the test by which the said new combination is to be measured. To be patentable, the new combination must produce a result which is the production of the combination and not a mere aggregation of several results, each the complete product of one of the combined elements. The combined results of various elements of a combination are not necessarily novel nor old results obtained in a new manner. Such combined results may be the mere aggregate of several results each of which is independent of and in no way affects the other.
Bringing old devices into juxtaposition and allowing each to work out its own effect without the production of something novel is not invention. Bringing together several old devices without producing a new and useful result, which is the joint product of the elements of the combination and something more than the aggregation of the old results, is not invention.
While inventions do not cover results, the latter must be studied to ascertain whether they are the product of a patentable combination or of a mere aggregation of separate, disassociated elements. A combination to be patentable must produce a different force or effect — a new or a better result — and it, the new result, must be the product of the coacting influences of the various elements. There must be a new or improved result produced by their union.
Analyzing the claims we find they contain the following elements in a radio power supply unit:
Claim 1: (1) A transformer; (2) means to connect said transformer to a source of alternating current; (3) a plate current supply fed by said transformer; (4) a series cascaded filter connected to said plate current supply and including adjustable series resistances to adjust the output potential thereof and to prevent the sudden discharge of current therefrom; (5) a filament current supply fed by said source of alternating current; and (6) a metallic container for enclosing said plate and filament current supplies and grounded to one of said supplies.
The elements of claim 2 are: (1) A transformer; (2) filament and plate current supplies energized from the transformer; and (3) metallic container for enclosing said transformer and associated apparatus, connected to one of the terminals of the filament current supply.
Claim 3 covers: (1) A transformer; (2) a rectifier; (3) a filter circuit operatively connected to said rectifier having a plurality of output terminals and means for reducing the potential at the terminals; and (4) resistances in the output terminals.
Applying the tests heretofore laid down to the claims in question, we find the first element in each claim to be a transformer. This element was ancient long before Edelman was intrigued by the mysteries of the radio. The same can be said of the rectifier, the filter, the resistances, the means to connect the transformer to a source of alternating current, and the filament current supplies. Each of these elements performed its old and well-known function separately from and independently of the other named elements.
The only element in any of the claims which might be said to be coacting is the metallic container, the last-named element in claims 1 and 2. It is worthy of notice that this element is entirely absent from the combination described in claim 3.
A shield thus used in a radio power supply unit was old. In the Hoxie patent, No. 1,382,914, which antedates the Edelman invention, the specifications explicitly refer to a shield. The inventor there said:
"In the construction of multistage amplifiers therefore it has been customary, as far as possible to shield the different input and output circuits from one another in such a way as to overcome both electromagnetic and electrostatic coupling between the circuits."
In the Weinberger patent, No. 1,618,319, the inventor said:
"The primary object of the present invention is to minimize or reduce the possibility of singing or self-excited oscillations by electrostatically shielding the input and output circuits and the apparatus contained therein, and also by electrostatically shielding each element within the amplifier from every other element by appropriate means."
If any asserted novelty in this element exists by adding the grounding of the shield, it is successfully met by the argument that such modification of elements or element was an obvious and old device which any electrician would adopt.
Moreover, there is no coaction between this element and the other elements enumerated in the combination covered by either claim. This shield served but one function. As its name emphasized, it shielded the radio set from the electromagnetic and electrostatic lines of the power supply unit. In the prior art such shields were used to shield one stage of a radio set from another stage, or one element of a set from other elements. Coaction it is true may result from the excision, as well as the addition, of a factor. In either case, however, the action of the element must affect another element of the combination described. It must coact with another element. If it acts separately — performs its functions separately and apart from the other elements — it adds nothing to the combination as such. As to this element then, the conclusion is inescapable that it was old and did not coact with any other element in the combination described. Each of the three claims of this patent, therefore, must be declared invalid for failure to disclose a patentable combination.
Edelman Patent No. 1,682,492. This patent covers a "Radio Battery Eliminator" for which an application was filed May 17, 1927. It is an improvement patent. The two claims in suit are 3 and 4 which read as follows:
"3. In a radio battery eliminator, a filter circuit including two choke coils having separate cores so placed that the mutual coupling between said coils causes ripple current flowing in one of said coils to be substantially opposed by electromagnetically induced ripple current supplied by the other of said two coils, and a by-pass condenser connected at the junction of said two coils.
"4. In a radio battery eliminator, a filter circuit comprised by two choke coils one of which has a core with relatively small air gap while the other has a separate core including a plurality of distributed air gaps therein, a by-pass condenser connected to the junction of said two coils, means to couple said two coils with a low percentage of mutual coupling sufficient to induce some ripple current from one of said coils into the other to oppose the flow of residual ripple current flowing in said other coil."
In his specifications, Edelman says:
"My invention relates to radio power supply and similar devices in which alternating current is converted to direct flowing current, filtered and supplied to a load circuit such as a connected radio receiving set without employing any battery. An object of my present invention is to provide a reliable battery eliminator capable of supplying direct current to standard radio sets of varying types and sizes despite various current requirements thereof which will function satisfactorily notwithstanding variations in the applied alternating current line voltage, and controllable from the regular radio set switch."
The defenses to claim 3 are invalidity and noninfringment. Judge Lindley observed in reference to a test made in his court as follows:
"Bearing upon the question of whether or not defendant infringed, * * * certain oscillograph tests were explained in open court. There were likewise certain ear phone tests conducted in which the court participated. The oscillograph tests were made by placing the testing instrument across the output of the power pack and operating the device first in the normal position and defined by the defendant's expert as in opposition, being the condition when the receiver is sold. In the second test the coil of one choke was reversed with respect to the other. The result showed less ripple or alternating current component when the coils are in the normal position, i.e., as sold, than when they are reversed. The ear test disclosed that when the coils were in normal position there was much less alternating current component heard than when in reverse. The construction of the defendant is equivalent to that of the plaintiff and infringement results."
A reproduction of a part of one of the drawings and a further quotation from the specifications will help to an understanding of this claim.
Edelman described this construction as follows:
"Choke coil 43 may be constructed similarly to choke coil 40 but preferably will have only 200 turns of wire affording less resistance to direct flowing electric current, also greater air gap preferably attained by interposing a plurality of distributed gaps 42 along the core. Such a core with distributed air gaps has the peculiar property which is very desirable of causing coil 43 to have an increased effective impedance to alternately fluctuating current as the amount of said current is increased, whereas coils with iron cores having small concentrated air gaps become less effective as the current flow increases. The effect of coil 43 therefore in part balances the effect of coil 40 with varying loads. A further novel and desirable filtering action is had by mutually coupling with a small percentage of coupling indicated by arrows 44, the two coils 40 and 43 whereby any slight residual fluctuating current flowing in coil 43 will tend to be substantially counterbalanced by the oppositely phased induced current set up therein by reason of the small coupling 44 so placed."
If infringement be avoided, it must be due to the fact that appellant's coils are so wound as to be series opposing, whereas in Edelman's illustrated drawing they are series aiding. Upon the record before us this difference can not be said to avoid infringement. This conclusion is predicated upon the fact that neither the claim nor the specifications limit Edelman in the construction of his claim to coils that are series aiding. However, if we were to restrict this claim to a structure that contained series aiding coils only, infringement would still not be avoided. The winding of the coils is not the essential thought of this invention. The locating of the coils so that opposition of the ripple currents necessarily follows is the significant thing. Opposition can be so readily obtained by differently positioning the coils so as to bring about neutralization of the fluctuations therein that the difference due to the winding of the coils is immaterial. If the claim be valid, there must be some range of equivalents allowed. The two choke coils of appellant's device are so located that that they are within their respective magnetic fields, which fact is determinative of the question of infringement.
Validity. As to the validity of claim 3, which is the question that has given us the most difficulty, the court is unfortunately in the position where it must rely entirely for enlightenment upon the experts' and counsels' explanations. There is no fact information of which we can take judicial notice. Doubtless due to their much greater knowledge, the experts have approached this question with little appreciation of the state of the mind (not art) to which their remarks were addressed. Our acceptance of the court's finding respecting the desirable results which accompanied the precise location of the two choke coils' construction, as Edelman describes, does not avoid the necessity of a study of the prior art. The court is, of course, limited to the record before it, which in turn is evidenced almost entirely by two prior patents. Inasmuch as our conclusion is based solely upon this record, it is hardly necessary to observe that should other prior art be disclosed, whether covered by patent or otherwise, which describes an apparatus wherein the choke coils were intentionally positioned as Edelman described, a different conclusion would follow.
Appellant relies upon the Houck patent, No. 1,684,703, and the Johnson patent, No. 1,613,952. The latter, it seems to us, is the more pertinent citation, but as appellants' first reliance is upon the Houck patent, it will be first considered. The District Court's observation respecting this Houck patent seems to be justified by its specifications.
The court said, speaking of the Houck patent:
"He says there must be no mutual inductance between the coils, but that mutual inductance must be avoided because it will cause hum and noises in the receiving set. Edelman taught the opposite, i.e., that there must be a small percentage of a coil inductance between the coils to effect the result hereinbefore described."
Appellant points to Figure 11 of this patent. Houck says in his patent:
"As indicated in Figure 11, the coils 49 and 52 must be wound in such a way that one will have no inductive effect upon the other. * * *"
Surely Houck did not, therefore, point the way for Edelman.
The Johnson patent, No. 1,613,952, issued January 11, 1927, upon an application filed December 31, 1920, relates to an "electric wave filter." The applicant states in his specifications:
"The invention relates to electric wave filters adapted to be employed in a line such as a telephone line, to prevent the passage of currents of certain frequencies and permit the passage of currents having other frequencies. It has for its object to produce filters of various types which have the characteristics of types of filters in general already known, but which may be manufactured from a less amount of material, thus reducing the cost and size of the filter.
"In accordance with the invention, filter sections are employed which have similar impedances spaced apart in the direction of current propagation but having mutual impedance therebetween, and which have one or more impedances located between the mutually related impedances. The effect of mutual impedance in general is to add an impedance to the common portion without the use of additional apparatus."
It is thus to be noted that Johnson did not apparently intend to cover, and further examination of the specifications shows that he did not knowingly cover, the combination covered by claim 3 of Edelman's patent. Johnson's specifications are replete with certain mathematical formulæ represented by equations, which we assume are sound, but all of which have no bearing upon the validity of Edelman's patent. In his drawings, however, Figure 3, Johnson discloses two closely positioned coils which he describes as follows:
"In Fig. 3, is shown a convenient arrangement for varying the self-impedances and the mutual impedance of the windings in a filter section of the type shown in Fig 2. * * * Cores 14 and 15 are mounted for endwise movement in the sleeves 10 and 11. It is apparent that the magnetic leakage, and consequently the mutual inductance, may be varied by rotating the sleeve 11 about its pivot 12, and that the self-inductances may be varied by sliding the cores 14 and 15 in and out."
Johnson's structure was for a different purpose and performed a different function than Edelman's. It had as its object the allowance of the passage of currents of certain frequencies and the prevention of the passage of currents of other frequencies. True, the Johnson combination may, under certain circumstances, anticipate Edelman even though Johnson did not appreciate its action or function when serving a different purpose. But we cannot say on the record before us that Johnson's disclosure would eliminate ripple currents of one frequency, more especially as Figs. 2 and 3 appear to be considered more or less as equivalents and neither figure discloses two separate choke coils each with a substantially closed magnetic path of its own. The patent as a whole did not deal with that problem at all. If one drawing disclosed two coils so positioned as to bring about a certain result (not appreciated by Johnson) and which disclosed action, due to the position of such coils, played no part in the problem which Johnson was attacking, it would not be an anticipation of the Edelman novel combination, unless in fact it would, and did do, what Edelman's combination did and aimed to do.
But, would Johnson's arrangement of the coils eliminate the ripple currents as Edelman's filter eliminated them? To establish this fact, the burden was upon appellant. Citing the Johnson patent as an anticipation, it had the laboring oar to establish the fact of anticipation. We are not satisfied of the existence of the fact upon which anticipation must turn.
Appellant's entire argument in support of its contention that this claim is invalid is to be found in a single sentence — "If the Edelman patent had disclosed the series opposing arrangement employed by defendant, which we respectfully deny, then it would be clearly anticipated by the Johnson patent." Obviously this sentence contains no persuasive reason for finding Johnson to have anticipated Edelman. If appellant's arrangement were dissimilar to Edelman's, but was similar to Johnson's, it could hardly be argued that Johnson anticipated Edelman. On the other hand, Edelman might have avoided Johnson, but yet have caught appellant, by showing and claiming his specific form of filter, the coacting influences of its two coils because of their position. The heart of Edelman's discovery, if we understand it correctly, is in the positioning of the two coils so as to cause ripple current flowing in one to be opposed by electromagnetically induced ripple current of the other in the manner specifically described. Johnson in no claims pointed to the positioning of the coils as means whereby the neutralizing effect of the two magnetically induced ripple currents were obtained. This is proved by his statement that "It is apparent that the magnetic leakage, and consequently the mutual inductance, may be varied by rotating the sleeve 11 about its pivot 12, and that the self-inductances may be varied by sliding the cores 14 and 15 in and out." It was not variance that Edelman was seeking, but constant and effective opposition of ripple current that he was after. We therefore reject Johnson as an anticipation.
Claim 4, we think, is clearly not infringed by appellant's structure. The core of each of the latter's choke coils has three air gaps, whereas claim 4 calls for one choke coil with a core that has a relatively small air gap while the other coil has a core including "a plurality of distributed air gaps therein." The specifications of the patent support this difference in air gaps in the cores of the choke coil. Edelman says in his specifications:
"Choke coil 43 may be constructed similarly to choke coil 40 but preferably will have only 200 turns of wire * * * also greater air gap preferably obtained by interposing a plurality of distributed gaps 42 along the core."
The drawings show three air gaps in the core of one choke coil, with only one air gap in the other. Having stressed in claim 4 a difference in the number of air gaps in the core of the two coils, appellee cannot read out of his first element this limitation in order to hold appellant as an infringer.
The decree is reversed, with directions to enter one in accordance with the views expressed in this opinion.