Opinion
Nos. 851, 852.
February 13, 1934.
Hugh M. Morris, of Wilmington, Del., and Clifton V. Edwards and Henry T. Kilburn, both of New York City, for plaintiff.
William G. Mahaffy, of Wilmington, Del., Thomas G. Haight, of Jersey City, N.J., and Abel E. Blackmar, Jr. (of Sheffield Betts), of New York City, for defendants.
In Equity. Separate patent infringement suits by John M. Miller against the National Broadcasting Company, Inc., and against RCA Communications, Inc.
Bills dismissed.
These two suits charge infringement of United States patent No. 1,758,000 for a piezo-electric oscillation generator, granted to John M. Miller April 22, 1930. Broadly, the patent relates to two specific radio signaling circuits and more particularly to circuits utilizing a piezo-electric crystal to control the frequency of oscillations generated with the aid of a vacuum tube with associated circuits including a tuned plate circuit. All claims are in suit except 3, 9, and 10. The alleged infringement consists of the use of crystal controlled oscillators employing tuned plate circuits in certain of the radio transmitting stations of the defendants.
The defenses are (1) invalidity and (2) noninfringement. The defense of invalidity is based upon the state of the prior art, particularly the work of Prof. Cady of Wesleyan University and Prof. Pierce of Harvard College.
"Piezo-Electric" (from to press) is a quality possessed by certain crystalline substances such as quartz. A piezo-electric crystal when compressed develops an electric charge on certain of its surfaces, and conversely, when charged by an electric current, the crystal is compressed and expanded.
Plaintiff specifies as objects of his invention (1) "to facilitate the generation of oscillations of constant frequency"; (2) "to select, at will, any possible mode of oscillation of the piezo-electric substance"; (3) "to permit adjustments to be made for the generation of such oscillations of maximum amplitude"; and (4) "to furnish a convenient means for connecting the oscillator to other tubes for amplification or other uses." The merit and significance of the patent is sufficiently indicated by the first of the above objects.
The claims in issue may be classified in three groups: (1) Those calling for a tuning of the plate circuit; (2) those calling for tuning of the plate circuit at a frequency approximately that of the natural frequency of the crystal; and (3) those calling for a tuning of the plate circuit at a frequency higher than the natural frequency of the crystal where the crystal is connected between the grid and the filament.
A glance at the broadcasting art may throw light upon a proper understanding of the patent. Radio communication is carried on by means of carrier waves transmitted from a radio transmitter through the ether to a receiver. In the transmitter alternating current of 10,000 cycles per second up to several million is fed into an antenna consisting of one or more wires suspended in the air above the ground. This current alternately flowing into and out of the antenna causes waves in the ether which spread in all directions, much as ripples caused by a stone thrown into a still body of water. Successive waves follow one another as long as successive impulses of electric current flow into and out of the antenna. The waves follow exactly both the timing and intensity of alternations of the electric current. The early stations produced these alternations by means of a spark gap. Then followed the use of electric arcs and mechanical generators.
In 1912 De Forest discovered that the vacuum tube could be utilized to generate alternations of electric current at radio frequencies. Thereafter the vacuum tube with its associated electric circuits was used as an oscillator to generate oscillations or high frequency alternating current for antenna, and became known as the "vacuum tube transmitter." Each radio transmitting station is assigned a definite transmitting frequency. To avoid interference, it is essential to maintain that frequency constant. When we turn the dial of a radio receiver, we make it more responsive to waves of one frequency than to those of any other frequency. This is tuning the receiver. It can be made to discriminate between many transmitting stations sending out waves of different frequencies. To enable the receiver to select between transmitting stations, it is essential that the stations transmit at different frequencies or rather at the particular frequency assigned to them. The tube and circuits of the oscillator are so constructed and interrelated that they generate alternating current at the assigned radio frequency. While the vacuum tube transmitter materially improved the stability of frequency as compared with the early forms of transmitting apparatus, the increasing use of wireless and the multiplication of transmitting stations required stability of frequency in order that sending stations would not interfere with each other. This demand for stable oscillations finally led to the adoption of piezo-electric crystals to control the frequency of the oscillations of the vacuum tube transmitter.
Long before the fall of 1923, when plaintiff entered the field, it was a fundamental axiom of radio that "wherever there are high frequency oscillations tune the circuit." For many years before 1923 the use of variable tuning in the plate circuit was fundamental. It had been common practice to employ for tuning a coil and in parallel thereto a variable condenser. In such oscillators the proper plate circuit reactance was obtained by tuning the plate circuit to a frequency either higher or lower than the operating frequency. Long before 1923, tuning by varying the inductance of a circuit was used as the equivalent of tuning by means of varying the capacity of a circuit. It was well known that no result is secured by one method which cannot also be secured by the other method. Before the present litigation, the plaintiff in another case testified that a coil in the plate circuit of a vacuum tube oscillator corresponds exactly to a coil having a condenser in parallel thereto.
Before 1923 the operation of vacuum tube oscillators controlled by electrically tuned circuits was thoroughly understood. It was understood that, where the grid-plate circuit is capacitative, the plate circuit must be tuned to a higher frequency than the operating frequency; i.e., have an inductive reactance at that frequency. Where the grid-filament circuit is capacitative, the plate circuit must be tuned to a lower frequency than the operating frequency; i.e., have a capacitative reactance at that frequency.
The method of operation of circuit controlled oscillators has been indicated above in general terms and is particularly disclosed in the work of Armstrong and Heising. The method and theory of operation of crystal controlled oscillators and of circuit controlled oscillators are the same. There is no fundamental difference between the crystal controlled oscillator in Fig. 1 of the patent in suit and tuned circuit controlled oscillators in the prior art including Fig. 3 of the Armstrong patent. The difference in detail is the substitution of the sharply mechanically tuned crystal for the less sharply tuned electrical circuits. Hogan, a recognized authority, testified: "Fig. 1 of the patent in suit and Fig. 3 of the Armstrong patent are both regenerative circuits, having the oscillation frequency controlled by a tuned element. The tuned element in Fig. 1 is the crystal shown at M in the patent in suit. The tuned element in the Armstrong Fig. 3 is the tuned circuit similarly connected between grid and filament and having its frequency determined by its effective electrical capacity and inductance. Fundamentally the two circuits are the same except for the substitution of a crystal having a definite natural frequency for an electric circuit having a definite natural frequency."
Walter G. Cady, professor of physics at Wesleyan University, began experiments with piezo-electric crystals as early as 1917, and is recognized as the pioneer in their use in this country. He was working with them actively in 1920 and 1921. In April, 1922, he published a paper in the proceedings of the Institute of Radio Engineers entitled "The Piezo-Electric Resonator." In his paper he disclosed to the art that a crystal was suitable to control the frequency of a vacuum tube oscillator; that a crystal was a sharply tuned element; that a crystal, like an electrically tuned circuit, had both capacitative and inductive reactance. He treated the crystal as though it were an electric circuit having resistance, capacity, and inductance.
Crystal oscillators employing a tunable plate circuit were used by Cady and his assistant, Van Dyke. The use of a tunable plate circuit to increase the power tube of such oscillators was described by Cady in the Physical Review for April, 1922, in an article entitled "A Piezo-Electric Method for Generating Electric Oscillations of Constant Frequency." In Figs. 2 and 3 of his patent No. 1,472,583, granted October 30, 1923, applied for May 28, 1921, Cady described two crystal controlled oscillators in which the crystal was connected between grid and filament and was an essential element in the generation of oscillations. In both Cady circuits the frequency of oscillation was controlled by a natural frequency of the crystal. In Fig. 3 the plate circuit is tunable indirectly, and in operation is tuned to a higher frequency than the natural frequency of the crystal.
Prof. George W. Pierce, Rumford professor of physics and director of the Cruft Laboratory for high-tension electrical research at Harvard College, became interested in the use of crystals through the work of Prof. Cady. In calibrating wavemeters in the spring of 1923, Prof. Pierce fabricated and utilized a crystal controlled oscillator in which the crystal was connected between grid and plate and the plate circuit contained a variable inductance coil. In May, 1923, he constructed his "piezo-electric oscillator No. 1," which he used in his experiments until the summer of 1923.
The results of the work of the first half of 1923 Prof. Pierce embodied in a paper drafted in July and published in the "Proceedings of the American Academy of Arts and Sciences" in October, 1923. The paper described a crystal controlled oscillator in which the crystal was connected between the grid and the plate of the vacuum tube and in which an inductance coil was connected in the plate circuit. Figs. 3, 5, and 6 of the article described the use of an inductance coil in the plate circuit. The article specifically states that an inductance of any value, if large enough, may be used in the plate circuit, and a coil of specific size is suggested for use with a crystal of a certain specified frequency. It was well known in the art that every inductance coil inherently has capacity, and therefore possesses a natural frequency, and that the coil has a capacity reactance. The description in the article was at the time sufficient to enable a radio engineer to secure crystal controlled oscillations. The article, in suggesting the trying of various coils in the plate circuit, described a practical method of tuning that circuit, and a radio engineer who followed the instructions of the article would tune the plate circuit. Furthermore, when a large enough coil to secure crystal controlled oscillations was employed, the natural frequency of the coil (by virtue of the inherent distributed capacity) was necessarily lower than that of the crystal, and therefore had a capacity reactance at the frequency of oscillation. Thus the plate circuit of the oscillator was necessarily tuned to a lower frequency than the natural frequency of the crystal. In this paper Prof. Pierce discloses the use of elements in the plate circuit for tuning.
In the fall of 1923, Prof. Pierce conceived of, disclosed to others and reduced to practice, the grid-filament connection of the crystal. At this season Prof. Pierce and Prof. Cady had a conference at Cambridge about their inventions in the use of piezo-electric crystals and about patenting and manufacturing their respective devices. When Prof. Cady was asked whether anything was said at this meeting about crystal connections in the circuit, he said: "I remember very clearly that Professor Pierce told me that when he first planned to make a circuit oscillate by means of the crystal, he expected to connect the crystal either between filament and grid, or between grid and plate, expecting that if the circuit did not oscillate one way, it would the other." In a footnote to a letter of October 23, 1923, after the conference, Prof. Cady wrote: "You noticed, perhaps, that my claims cover the case where the resonator is between filament grid, but not when it is between plate grid." Regarding this, Prof. Cady testified: "As near as I can remember we had compared notes in order to determine what was originated by one of us and what by the other, and I was making clear to him there that I laid no claim to the connection of a crystal between grid and plate." Eastham, president of General Radio Company and manufacturer of electrical devices, was also present at the conference. Regarding the instructions for manufacture, he said: "I know that at the time that we arranged to go ahead with it, we planned for a circuit having the two possibilities available, that is the grid-plate connection, and the grid-filament connection," and that this was planned "within a few days of that meeting." As to his instructions to Eastham to enable him to construct a commercial apparatus, Prof. Pierce said: "I had explained to Mr. Eastham that the crystal might be shifted from its connection between plate and grid to its connection between grid and filament, leaving the rest of the circuit constant and bringing out thereby two different frequencies of oscillations, one being the high frequency mode of operation of the crystal and the other the low frequency mode of operation of the crystal."
November 15, 1923, Prof. Pierce and his assistants completed the construction of his "piezo-electric oscillator No. 2," which was provided with a variable tuning coil permanently connected in the plate circuit. Prof. Pierce and his collaborators also tuned the plate circuit of his crystal controlled oscillators in other ways, such as by means of a closely coupled tunable circuit and by means of two coils used as a variometer, both of which were recognized electrical equivalents of the arrangement shown in the plate circuit of the figures of the patent in suit. In using variable coils in the plate circuit of the crystal controlled oscillators, Prof. Pierce provided the means for and utilized "rough" tuning of that circuit, i.e., a step-by-step tuning. The only difference between his plate circuit so used and the circuit of Fig. 2 of the patent in suit is that the method of tuning employed by Prof. Pierce was "rough," whereas the method indicated in the patent in suit is "fine." "Rough" tuning and "fine" tuning merely express differences in the method of arriving at the desired tuned condition of the plate circuit. This condition is the same whether secured by "rough" tuning or "fine" tuning. The difference between these two circuits is, at most, merely one of degree and not in any sense one of substance.
In November, 1923, Prof. Pierce, believing he had made an important invention, called in his patent counsel. Prof. Pierce prepared a memorandum of the principles of his invention, and delivered it to his attorney November 27, 1923, with explanations to guide him in preparing a patent application. After further conferences, an application for patent was filed by Prof. Pierce February 25, 1924. This application contained claims specifically directed to a crystal controlled oscillator in which the crystal was connected between grid and filament and which had a tunable plate circuit.
Early in 1924, crystal controlled oscillators were manufactured and sold by the General Radio Company of Cambridge, Mass., on behalf of Prof. Pierce. These oscillators were so constructed that the crystal might be connected either between grid and plate or between grid and filament. The plate circuit of these oscillators contained a coil which, in the process of manufacture, had been so adjusted or tuned with respect to the natural frequency of the crystal that the necessary reactance relations were secured and crystal controlled oscillations obtained.
It should be borne in mind in considering plaintiff's work that his application for the patent in suit was filed September 10, 1925. The application of Prof. Pierce covering the same subject-matter was filed February 25, 1924. Since the bringing of this suit, an interference has been declared by the Patent Office between the patent in suit and the pending Pierce application. The necessary legal consequence of this situation is that the law imposes upon plaintiff the burden of proving by a fair preponderance of the evidence that he was the first and original inventor. Evans v. Associated Automatic Sprinkler Co. (C.C.A. 3) 241 F. 252; Willard v. Union Tool Co. (C.C.A. 9) 253 F. 48.
From 1907 to 1919 plaintiff was a physicist in the Bureau of Standards. In 1919 he was transferred to the Navy Department. About 1923 he was transferred to the United States Naval Research Laboratory at Washington where he was in charge of the section having to do with receiving sets, vacuum tubes, and frequency measurements.
In 1921 plaintiff obtained from the Bureau of Standards a crystal resonator loaned by Prof. Cady, and experimented with it in two circuits shown in a patent of Prof. Cady, but his experiments were failures. In July, 1923, the Bureau of Engineering of the Navy forwarded to him an Italian article about frequency standardization. Plaintiff reported against the Italian suggestion, but recommended the use of crystal resonators for the purpose of checking frequency meters, saying: "These resonators are quartz crystal rods which have natural frequencies of vibration in this range and can be made to oscillate by the application of an electric field, exhibiting characteristics similar to a tuned circuit. They are described in a paper by Prof. W.G. Cady of Wesleyan University. * * * This laboratory has investigated these resonators to some extent, utilizing a borrowed crystal, and is convinced that they would be of very great value as frequency standards. Further they can be utilized to control and keep constant the frequency of the output of a vacuum tube oscillator and in this respect might be very valuable in the design of constant frequency transmitters."
Early in November, 1923, after the publication of his article on crystal resonators and oscillators, Prof. Pierce distributed many hundred copies of the article to scientists, radio engineers, and friends, including Dr. Miller. In acknowledging receipt of his copy on November 6, 1923, plaintiff said:
"Have had a number of reports with respect to the interesting experiments you have been conducting with crystal oscillators and am greatly indebted to you for sending me a reprint of your paper on this subject.
"We are planning to use these crystals in checking our wavemeters throughout the service and your paper will doubtless be very valuable assistance to us in this work."
Not only is there no suggestion in this letter that plaintiff had worked with piezo-electric crystals as controllers of the frequency of oscillations generated by vacuum tubes, but plaintiff testified that up to that time he had done no such work.
In view of the above evidence, it is difficult to credit plaintiff's statement in the bill of particulars that he made his invention, reduced it to practice and disclosed it to others on November 6, 1923, the same day he received the Pierce article. If full credit be given this statement, it follows that it was only necessary for a person skilled in the art to read Prof. Pierce's paper to immediately conceive of what plaintiff afterwards claimed to have discovered. Upon receipt of the Pierce article, plaintiff set up a circuit similar to that shown in Fig. 1 of the article. He did not succeed in getting the crystal to os cillate.
Plaintiff testified the result of his first work on crystal oscillators was entered on a loose sheet dated "Nov. 30 and Dec. 1st 1923" (Plaintiff's Ex. No. 29), that he set up and operated a crystal controlled oscillator using a "Paragon" tuned plate circuit, and that he obtained oscillations with four different crystals. He further testified that he entered on another loose sheet dated "Dec. 4th 1923" (Plaintiff's Ex. No. 31), the result of further work employing an ultraudion tuned plate circuit. On this sheet appears a circuit sketch showing the crystal connected between grid and plate with an inductance coil and a variable condenser in parallel in the plate circuit. The evidential value of these sheets is very doubtful. There was no occasion to employ loose sheets. At that time plaintiff had two notebooks and the Navy Department furnished regular log sheets on which such entries appropriately would be made. Moreover, there is no proof where these loose sheets had been during the eight years between 1923 and the time they were introduced in evidence. Certainly they were wholly unofficial, as they were not included among the certified copies of documents which the Navy Department furnished at the request of the court. Further, the two witnesses called by plaintiff out of the eight named in the bill of particulars, as persons to whom plaintiff disclosed his invention in November, 1923, did not corroborate plaintiff's testimony. Neither of them was able to say the experiments on the loose sheets had been made before December 21, 1923, which was four days after a personal visit by plaintiff to Prof. Pierce. There is no suggestion on either of these sheets of the grid-filament connection. They are only tied up to that connection by the oral testimony of the plaintiff. That testimony stands uncorroborated.
By letter of November 30, 1923, Prof. Pierce wrote plaintiff:
"I am much obliged to you for your letter of November 6 regarding the crystal oscillators.
"The General Radio Company, has, I believe, made arrangements to manufacture the crystal resonators for checking wavemeters. The crystal oscillator will probably be also supplied by the General Radio Company at some later date.
"I shall be very happy to see you at any time that you are in this neighborhood. With best wishes."
December 6, 1923, plaintiff wrote to Prof. Pierce: "Since I have received your paper on crystal oscillators, I have been experimenting with them to some extent and have had both success and failure. I would like to take advantage of the invitation in your letter of the 30th of November * * * to see you for a few days, in order to obtain some pointers on the subject. I can come at any time which suits your convenience. I will suggest Monday December 17th but would be pleased to have you set a date this month or next which would be more convenient."
Plaintiff arranged with the Navy Department for the expense of a visit to Cambridge on December 17, 1923. He spent the day in the Cruft laboratory, and was shown by Prof. Pierce his notebook and his piezo-electric oscillator. It is significant that plaintiff states the purpose of his visit was "in order to obtain some pointers on the subject" from Prof. Pierce. Regarding the visit, Prof. Pierce testified: "I have in mind that we discussed the reason for the oscillation in the circuit disclosed in my published paper. That is what I have in mind. Whether we discussed other circuits or other theories, I do not recall." Prof. Pierce also testified that he told plaintiff that he had found spectacle lenses good oscillators and advised him to try them. At variance with Prof. Pierce's recollection, plaintiff testified: "On this trip to Prof. Pierce, I explained the operation of the circuits which I had developed prior to my visit. Also Dr. Pierce told me that he did not understand the operation of the circuit which he published in his paper, and I explained the operation of that circuit to Prof. Pierce." This testimony calls for no comment. Counsel for plaintiff lays much stress on plaintiff's further testimony:
"Q92. What did you tell Dr. Pierce about your use of the tuned plate circuit? A. I explained the operation of the tuned plate circuit practically in the manner that it is explained in my patent.
"Q93. Would you mind very briefly going over that? A. That is, that in the case of the crystal connected between the grid and filament, it is necessary to have an inductive reactance in the plate circuit in order to generate oscillations; that this inductive reactance might be obtained by using a tunable circuit, but tuned to a frequency higher than the frequency generated by the crystal; that in the case of the grid to plate connection where a capacitive reactance is required, that the capacity reactance could be obtained by using a tunable circuit in the plate of the tube, and by adjusting this circuit to a frequency lower than corresponded with the frequency of the crystal."
That he made such a full disclosure of his patent on this occasion is wholly inconsistent with the statement in his letter that the purpose of his visit was "in order to obtain some pointers on the subject" from Prof. Pierce. It is also inconsistent with the entries on the first page of the notebook that he opened upon his return to Washington. Those entries are, however, a full corroboration of the information that Prof. Pierce testified he imparted to plaintiff. To the question, "Do I understand that the net result of your visit to Cambridge in the Fall of 1923, was that you gave, and did not receive, information?" plaintiff answered, "Yes, excepting in the respect that Professor Pierce advised me to try spectacle lenses."
Upon his return to Washington, plaintiff started a notebook of his own. Following Prof. Pierce's advice, he borrowed some quartz optical lenses, and the first entry in his notebook reflects the use of this special type of crystal to secure high frequency oscillations. At that time plaintiff was using the grid-plate connection shown in the Pierce article. About three weeks later the first entry relating to the grid-filament connection appears in plaintiff's notebook in the words, "crystal connected between grid and filament." The entry contains no reference to a tuned plate circuit and no sketch or circuit diagram is shown. In fact, the first drawing or circuit diagram of plaintiff showing the grid-filament connection appears in his notebook on March 24, 1924. This entry is the first disclosure of what is substantially Fig. 1 of the patent in suit. It is significant to note that on March 6, 1924, about three weeks before this entry, plaintiff had paid a second visit to Prof. Pierce at Cambridge.
There is no occasion to further discuss Miller's work. Why did he defer his application for a patent a year and a half after the entry of March 24, 1924? The answer is obvious. He never considered himself the inventor of the grid-filament connection of the crystal with a tunable plate circuit. In view of his familiarity with the work of Prof. Cady and Prof. Pierce and the literature of the art, he did not consider that the grid-filament connection with the tunable plate circuit involved invention.
In the earlier part of this opinion the development of oscillators utilizing vacuum tubes was briefly outlined. Before the piezo-electric crystal was added, the literature of the art had explained that the plate circuit must be tuned above or below the operating frequency of the tube circuit. Plaintiff merely followed this teaching in tuning the plate circuit above the crystal frequency with the grid-filament connection. If, for the sake of argument, plaintiff should be credited with this teaching, what does it amount to?
The infirmity of plaintiff's case is that he is seeking to rest it, not on practical accomplishment, but on the statement of a theory of operation; namely, that with one crystal connection the plate circuit must be tuned to a higher frequency, and with another crystal connection it must be tuned to a lower frequency. Certain claims of the patent call for a crystal connected between grid and filament and a plate circuit which is tuned to a higher frequency than that of the crystal. The explanation with respect to the plate circuit, however, contributes nothing, for the reason that the circuit will not generate oscillations unless the plate circuit is tuned to a higher frequency. A pupil of the entire work of Prof. Pierce concerning this oscillator, like Dr. Miller, would be bound to secure oscillations, and in so doing would be bound to tune the plate circuit to a higher frequency with the grid-filament connection. Whether the circuit is tuned to a frequency above or below the crystal frequency is a matter of only a small fraction of a turn on the scale of the variable condenser. In practice, the operator sets the variable element of the plate circuit to approximately the desired frequency, and then varies the element back and forth until the oscillations are secured. The operator does not necessarily know, and does not care, whether the plate circuit is tuned to a lower or a higher frequency than that of the crystal. Oscillations will be secured when the frequency relation, whatever it may be, is correct.
All the elements of the Miller circuit were disclosed by Prof. Pierce. Prof. Pierce secured oscillations by use of his crystal controlled oscillator employing a tuned plate circuit. He was bound to tune the plate circuit above or below the natural crystal frequency in order to secure the desired oscillations. What could plaintiff contribute except a statement of the theory of operation? Assuming plaintiff did explain the theory underlying the operation of the crystal controlled oscillator employing a tuned plate circuit, that is a scientific achievement of a kind that cannot be the basis of a patent. Whether Prof. Pierce knew the scientific explanation of his crystal controlled oscillator is not important. He did know and use the device and employ the methods which produced the desired results, and which are the device and methods of the patent. De Forest Radio Co. v. General Electric Co., 283 U.S. 664, 686, 51 S. Ct. 563, 75 L. Ed. 1339.
Having in mind the state of the prior art, I find that plaintiff was not the first and original inventor of the subject-matter disclosed and claimed in the patent in suit.
This opinion contains a statement of the essential facts and of the law applicable thereto in conformity with Equity Rule 70½ (28 USCA § 723).
The bills of complaint must be dismissed.