Opinion
No. 267.
May 2, 1938.
Appeal from the District Court of the United States for the Eastern District of New York.
Patent infringement suit by the Radio Corporation of America against the Mackay Radio Telegraph Company, Inc. Decree for defendant, 16 F. Supp. 610, and plaintiff appeals.
Decree modified.
Jo. Baily Brown, of Pittsburgh, Pa., and Sheffield Betts and Abel E. Blackmar, Jr., all of New York City, and Harry Tunick, for appellant.
Darby Darby, of New York City (Hugh M. Morris, of Wilmington, Del., and Samuel E. Darby, Jr., Paul Kolisch, Roy C. Hopgood, and E.D. Phinney, all of New York City, of counsel), for appellee.
Before MANTON, L. HAND, and SWAN, Circuit Judges.
This appeal involves the validity and infringement of the Lindenblad patent, No. 1,927,522, granted September 19, 1933, on an application filed December 24, 1928, and the Carter patent, No. 1,974,387, granted September 18, 1934, on an application filed June 11, 1930. Both are for short wave directional antennas composed of a pair or pairs of long wires, the two wires of each pair being placed at an angle to each other to form a V. Claims 9, 10, 19, and 23 of the Lindenblad patent and claims 15 and 16 of the Carter patent are in issue.
Directive radio transmission is of commercial importance where communication is regularly carried on between two or more fixed stations. The advantage of highly directive systems for point to point communication is that interference with other systems is reduced and, for a given amount of energy sent out by the transmitting station, a stronger or more reliable signal is heard at the desired receiving point. When radio waves are directed or concentrated upon a given receiver, freedom from interference follows. If the transmitter in New York concentrates its radio waves upon a receiver in a city in Central Europe, no waves or only feeble waves will travel in other directions and, as a result, other receiving stations to the north, west, or south points will not be subjected to interference. If the New York station transmits using a highly directive antenna (as Carter) which concentrates most of the 360 kilowatts of power into a bundle of rays confined within an angle of 20° or less, there would be substantially the same total power concentrated within one-eighteenth of the full compass circle or nearly 18 kilowatts compressed into each one degree sector instead of only one kilowatt. Thus, much of the power that would be wasted if the waves were broadcast in all directions may be brought to bear upon a particular receiving station and the intensity of the receiving signal multiplied.
The best antenna to serve these purposes appeared in 1926 as the Marconi beam antenna. It was made up of a complex network composed of a large number of short wires. It was highly directive and expensive, costing about $100,000 to produce. It was difficult to adjust and keep in repair.
Appellant's models B and C were built in 1929. They were built of pairs of long parallel, simple wires which co-operated to give highly directional effects. They were built at a cost of $30,000 and $48,000 and were made under prior Lindenblad patents.
No successful use of any other directional wireless is referred to until antennas made under the patents here to be considered were built. The first step in the directive antenna (made as model D) under these patents was developed in the invention of Lindenblad. The primary units (like models B and C) were made of long simple wires and the antenna consisted of two such wires placed at an angle with each other to form a V. Although the V-antenna described by Lindenblad had utility, it was not adapted for commercial use until Carter's improvement had been applied for. This antenna consists of one or more pair of long, straight, unmodified wires placed at a preferred angle; when a simple V is used the antenna is bi-directional. The predominant radiation of radio waves from the antenna is in two directions, the direction determined by the line bisecting the angle between the wires. By the use of two V's properly spaced with respect to each other and fed with current in proper phase relation, unidirectional action is secured. The predominant radiation is in one direction only. The direction of the bisector of the angle between the wires is either forward or backward according to the connection used. This is the Carter patent. It was an improvement over the directive antennas which had preceded it. Its cost was $5,000 as against $30,000 for the cheapest previous types. Its directivity was equal to the Marconi beam.
Although the theory of operation of a single wire had been previously developed prior to Lindenblad and Carter, no one had found how the theory could be utilized in the solution of the problem of constructing a highly directive antenna. One of the steps in the development of Carter's patented antenna is set forth in the Lindenblad patent. A long wire V antenna is shown in fig. 2. In the antennas of each of the patents energy is fed into the antenna by transmission line through an impedance matching device, in such a manner that the energy fed to one antenna wire is of opposite phase to that fed to the other wire, that is, when the potential at any point on one wire is positive, that of the corresponding point on the other is negative and vice versa.
Lindenblad says that the current on the antenna wires may be in the form of standing waves, or "as a refinement" and "a further object of my invention" is in the form of traveling waves. He says that with traveling waves the sideways radiation is less than with standing waves, but the reduction of radiation in undesired directions may be obtained by other means.
Lindenblad stated nothing as to the advantages of any specific angle between the wires of his V-antenna and as to any relation between preferred angle, wire length, and wave length. He said that the spacing between the ends of the wires, "while variable" over a long range, should be in the neighborhood of a fifth of the length which, regardless of wave lengths and wire lengths, means an angle under 12° between the wires. The wire length suggested is of the order of five or ten waves long. Carter taught 12° is not the most desirable angle for a V-antenna having wires of these lengths. He developed the theory of long wire V-antennas and taught how to construct them so as to secure maximum directivity by co-ordinating the angle wire length and wave length. He found that the use of a preferred angle for any given wire length and wave length gave an antenna which was equal or superior in directivity to any known antenna and far less costly to erect. Carter described multiple V-antennas both for preventing undesired high angle radiation, for securing further concentration in the horizontal or compass direction, and for securing unidirectivity.
Claim 10 of the Lindenblad patent is typical and will be selected in determining the question of infringement. It reads: "In combination, a transmission line, and an antenna extending longitudinally in the direction of desired radiant action connected thereto comprising a pair of open ended conductors of the order of magnitude of a number of wave lengths long which are widely spaced at the ends remote from the transmission line and energized with energy of opposite polarity, and spaced at the spacing of the transmission line at their junction therewith."
The inventor intended to get rid of standing waves and he was to accomplish it by lengthening his antenna which eliminated reflection in the antenna. It, however, made the space between the conductors too long to be practicable. His figs. 3 and 4 show the conductors bend back and forth across each other and these figures were for traveling waves only. Fig. 2 was a standing wave antenna and would need an impedance matching device unless the transmission line was very short. He expected some standing waves and thought to improve on the di-pole.
Carter's claim 15 reads: "An antenna comprising a pair of relatively long conductors disposed with respect to each other at an angle substantially equal to twice 50.9 ()-0.513 degrees, 1 being the length of the wire and the operating wave length in like units, and means in circuit with said antenna for exciting the conductors in phase opposition whereby standing waves of opposite instantaneous polarity are formed on the conductors throughout their length."
Claim 16 differs from claim 15 primarily in calling for "a similar parallel pair of conductors spaced at odd number of quarter wave lengths away from said first mentioned pair along the bisector of the angle of the conductors."
The appellee's V-antenna wires are relatively long. The angles used by it are all substantially the angles determined by the formula set forth in Carter's Claim 15.
Appellee constructs its V-antennas (Nos. 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11) utilizing in most instances substantially the exact angle specified by Carter, and places them upon poles 80 feet high and secured antennas that were effective in the identical directions that Carter described. Structurally the antennas are Carter's and secure the same results as he describes, that is, concentration of radiation on the distant receiving antenna, simplicity, and efficiency. They are used to obtain the results Carter sought. The angle between each wire of appellee's antenna No. 2 and the bisector of the angle between them is 17.5°. Carter's formula and curve call for an angle of approximately 17.8° for wires of this length. The compass bearing of the bisector of this antenna is N 49.75°E. This is referred to as the directivity of the antenna. This particular antenna is shown to have been used for communication between Vienna and Sayville, L.I., the direction of which is N 49.74°E from Sayville, which is substantially that of the bisector.
No. 1 is representative of appellee's double V antennas and it too is established to infringe. Indeed, the difference in each is trifling except No. 8, where there is a divergence of about 10 per cent.; 22½° as against 25°; No. 10 has a departure of 5 per cent away from the prescribed angle. These we think are substantially the same as the empirical angle. By appropriating Carter's idea, the substance of his discovery was taken and there is infringement.
Lindenblad in his V-shaped antenna greatly re-enforced transmission. His angle was a little less than 12° and was practicable. Claim 10 required widely spaced conductors but the appellee used an angle never less than 17.5 and cannot be charged with infringement if the claim be limited as it should. It cannot be read more broadly for it appears that on several angles such an antenna will not operate. The problem was how to proportion the conductor length to the angle and this had to be worked out by trial. The patent gave no direction to the art, but it did give a suggestion. That is insufficient to cover the operations of the appellee and it may not be held to have infringed the Lindenblad patent.
It is argued that the patent is limited to antennas utilizing the predominant radiation which goes out exactly in the plane of the wires to the exclusion of that which goes at some small angle to that plane. When the antenna is placed over ground on 80-foot poles, the presence of the ground affects radiation in certain ways but not so far as the directivity in compass direction is concerned. The unavoidable effects in the presence of ground are that radiation that has left the antenna in the exact mathematical plane parallel to the earth is, at a distance from the antenna, canceled by a radiation that left the antenna at a slight downward angle and been reflected upward from the ground. Radiation which has left in a small upward angle is re-enforced by radiation that left at a similar downward angle and has been reflected from the earth's surface. These effects always occur when short wave antennas are placed as both appellant and appellee place theirs, that is, not far above the surface of the ground.
Cancellation of horizontal radiation does not destroy the effectiveness of the antenna as a source of signals for a distant receiving station. Long distance short wave transmission is not carried on by means of signals transmitted horizontally along the surface of the earth; the signals that reach a distant receiving antenna are those which have left the transmitting antenna at a small upward angle and been deflected back to earth by a portion of the upper atmosphere that reflects radio waves due to the fact that it is conductive because of ionization. The main consideration is that Carter described a practical directive antenna which is of great commercial utility when used just as he said it was to be used. He was concerned with antennas which would concentrate the radiation in certain compass directions. These antenna structures are used by both appellant and appellee.
It is next argued that the patent is limited to antennas, the wires of which are an integral number of half waves long. Most of appellee's antennas do not use wires that are of integral number of half wave lengths but that does not avoid infringement. Carter's application stated that the invention was applicable to wires of any finite length. His original application included the empirical formula and the curve of that formula is shown and described. Appellee's argument is based upon the limited application of the prior art Abraham formulas. The effect of this argument is to show that, as to wires which are not an integral number of half waves long, the recognition that there was for such wires an angle of predominant radiation was not obvious but was covered by Carter. Carter found that interpolation between these lengths was sufficiently accurate for practical purposes and disclosed the angles to use for V-antennas using wires of intermediate lengths. Appellee followed this disclosure and accomplished the results set forth by Carter.
Another argument advanced is that the patent is limited to antennas using the exact angle recommended by Carter. The argument proceeds that appellee's V-antennas utilize a slightly smaller angle than that recommended by Carter, with the result that the upwardly tilted radiation is still more augumented by ground effect. But nine out of appellee's eleven V-antennas were found to be substantially the exact angle recommended by Carter and the other two were appreciably smaller.
We do not find that the prior art anticipated Carter's invention. The Bethenod French patent, No. 596,737, was an old type of long wave antenna — two horizontal networks of wires one placed above the other, the two being fed in phase opposition. It fails to disclose the Carter invention, for the latter was not merely for the use of long wires in antenna system but involved placing the wires in the form of a V and at the preferred angle disclosed. Bethenod failed to disclose either of these arrangements. His parallel networks taught nothing to the art in the solution of the problem here involved. It is insufficient to invalidate this patent disclosing a meritorious invention. Babcock Wilcox Co. v. Springfield Co., 2 Cir., 16 F.2d 964; Electro Bleaching Gas Co. v. Paradon Engineering Co., 2 Cir., 12 F.2d 511.
The Bethenod French patent, No. 625,293, discloses a single wire antenna. The novelty of this arrangement is in the feed system but not in the antenna. Even though it disclosed an electrically long wire, it does not disclose the two wire V-antennas of the Carter patent. He shows a single wire which he says may be placed at an angle with respect to the ground and appellee argues that this single wire with its image in the ground constitutes a pair of wires. This is referred to as the image theory. But the testimony of the experts clearly refutes this theory. The image in the ground of an actual wire above the ground is purely a theoretical consideration. After an antenna has been designed, the image theory may be used to analyze the ideal effect of ground upon its radiation. But it is quite different from taking a prior antenna and its image and attempting to use those as a new antenna. The record discloses no one has ever attempted to do that. Moreover, the actual wire and its image is never equivalent to a two-wire antenna for a two-wire antenna necessarily is more than twice as directive as a single wire antenna including its image.
The French patent to Levy, No. 593,570 (and the addition thereto of No. 30,798), shows two networks each composed of three electrically short wires, the network being fed in phase opposition. Levy's wires were short and he failed to appreciate any advantage in the use of long wires which constitute the fundamental antenna unit of Carter. He, like Bethenod, is indefinite as to the obtaining of a directive action with his antenna. He does not tell what type of directivity may be secured or how any desired kind of directivity may be secured. He uses parallel wires in each of his networks and failed to make any provision with respect to feeding the individual wires in such a way that each network as a whole will radiate substantially as though it were a single wire. The radiation of such networks is uncertain because the phasing and relative position of the wires are not definite.
The alleged prior invention of Bruce does not anticipate. He had experience in the study of short wave antenna problems and did seem in 1926 to have thoughts of a directive antenna composed of a pair of simple straight long wires arranged in common plane, but he failed to appreciate the possibilities of a long wire V-antenna nor did he recognize the possibilities of such construction. It was after the patents in suit had become known that he proceeded to obtain his result. His experiments previous to the work of Lindenblad and Carter were no more than abandoned efforts. Such abandoned experiments, even though they utilized the exact patented structure, which Bruce did not, have been held insufficient to invalidate a patent. Coffin v. Ogden, 18 Wall. 120, 21 L.Ed. 821; Deering v. Winona Harvester Works, 155 U.S. 286, 15 S.Ct. 118, 39 L.Ed. 153; The Corn Planter Patent, 23 Wall. 181, 23 L.Ed. 161.
The Abraham articles of 1898 and 1901 with the formula suggested, although 30 years old, had never been utilized. From that time was the entire period of the development of the radio art. While his formulas unquestionably had useful scientific interest to radio engineers, they failed to teach how to construct a directive antenna. Nothing in them suggests the inventions of Carter.
The problem solved here was of long standing and it was an eminently successful solution. Critical examination of prior patents and uses finds nothing invalidating the Carter invention and it shows a substantial accomplishment.
The decree will be modified holding the Lindenblad patent not infringed and the Carter patent valid and infringed.
Decree modified accordingly.