In re Pupin

299 F. 697, 55 App. D.C. 14, 1924 U.S. App. LEXIS 3480

• Court: Court of Appeals for the D.C. Circuit
• Decided: June 2, 1924
• Docket: No. 1644
• Status: Published

Opinion

SMITH, Acting Associate Justice.

This is an appeal from the decision of the Commissioner of Patents denying the claims of an application for the patenting of new and useful improvements in the electro-magnetic production of direct currents without fluctuations, which improvements supply means for eliminating fluctuations in the voltage and current caused by the action of the commutator bars of electro-magnetic generators.

The application for the patent was filed in the Patent Office on the 31st day of December, 1915, and contained the following claims, which the applicant now contends should have been granted:

1. A dynamo® electric generator producing unidirectional periodically fluctuating electromotive force and current, in series with an aperiodic pilot conductor made up of unit sections containing reactance and resistance so adjusted as to afford high attenuation power for fluctuations in the generated electromotive force and current.

2. A direct current dynamo electric generator having a commutator and collector brushes, in series with an aperiodic pilot conductor made up of unit sections containing reactance and resistance so adjusted as to have a high attenuation power for fluctuations of the frequency of commutation.

**********

5. A direct current dynamo electric generator having a commutator of relatively large number of commutator segments and designed for rotation at such a speed that the commutation frequency is relatively considerably higher than in ordinary commercial machines of similar type, in series with a pilot conductor made up of unit sections containing reactance and resistance so adjusted as to afford high attenuation power for fluctuations of the frequency of commutation.

6. A direct current dynamo electric generator of the commutator type having a commutation frequency greater than 2,000 p. p. s. in series with a pilot conductor made up.of unit sections containing reactance and resistance so adjusted as to afford high attenuation power for fluctuations of the frequency of commutation.

7. A direct current dynamo electric generator having a commutator of a relatively large number of commutator segments and designed for rotation at such a speed that the commutation frequency is relatively much higher than in ordinary commercial machines of similar type, collector brushes operatively associated with said commutator and an aperiodic pilot conductor, electrically connected to said brushes and made up of unit sections of reactance and resistance so adjusted as to have a high attenuation power for fluctuations of the frequency of commutation, each unit section including a resistance electrically connected in series relation with said brushes and a reactance electrically connected in parallel relation with said brushes.

8. A dynamo electric generator producing unidirectional periodically fluctuating electromotive force and current, in series with an aperiodic 'pilot, conductor made up of unit sections of reactance and resistance so adjusted as to afford high attenuation power for fluctuations of the frequency present in the generated electromotive force and current, each unit section comprising a resistance connected in series relation with said electromotive force and a reactance connected in parallel relation with said electromotive force.

The quoted claims were rejected by the Examiner on the ground first, that the combination of the dynamo electric generator producing unidirectional periodically fluctuating E. M. E. with a piloting conductor in series, having a high attenuating power for fluctuations, was old on reference to Steinmetz, and that the large commutator with many segments used by the applicant to produce a smooth uniform current, was disclosed by Van Deventer and differed from that of Steinmetz only in the particular, that the former had a higher commutator frequency which was a difference soléiy in degree; second,, that there was no invention in substituting for the pilot conductor in Steinmetz the several sections of attenuating conductor in Rudenberg,. inasmuch as Rudenberg states that several of the sections may be connected in series, and that his device will annihilate small disturbing waves and will not permit free oscillation “because the damping resistance must in itself be of such a value that any natural vibration is suppressed.”

The Examiner therefore held that the substitution of several sections of the attenuating conductor of Rudenberg for the pilot conductor of Steinmetz, in order to eliminate fluctuations in the work circuit of a direct current generator or in any circuit in which fluctuations were objectionable, was obvious and therefore not invention. The Examiners in Chief affirmed the decision of the Examiner, and on appeal to the Commissioner their decision was affirmed.

Applicant’s device comprises a generator of the double current type having two armature windings, each of which is connected with a -corresponding commutator. One of the windings is designed to produce a relatively large current and small voltage, and the other a relatively small current and high voltage. The armatures supply separate load circuits, and between each armature and its corresponding load circuit is inserted a so-called pilot conductor or filter, which has for its purpose the suppression or filtering out of the commutator voltage ripples. On the low voltage -side of the generator the pilot conductor consists of a series of connected iron core inductance coils on the side of the supply circuit, and a series of condenser bridges connected in parallel across the supply circuit between the inductance coils. On the high voltage side of the generator, resistances are substituted for the inductances with the condensers connected in parallel between the resistances. The condensers furnish shunt paths of low impedance to the high frequency currents caused by the commutator voltage ripples. The inductance coils offer relatively high impedance to the flow of such high frequency currents on the low voltage side and the resistances accomplish a similar result for the high voltage side, thereby practically taking up the fluctuations in voltage and suppressing the commutator pulsations.

In sound-reproducing systems the desired sound is produced by intentionally causing fluctuations in the electrical current. Fluctuations other than those intentionally produced result in an objectionable hum, which blurs the communication and impairs the usefulness of such systems. To avoid the hum, electric current for telephones is usually supplied by batteries which are in large measure free of unintentional fluctuations.

The device of the applicant made it feasible to substitute for the batteries a dynamo, which, although more convenient, was not favored, because its commutator was divided into segments, and each segment in turn, produced an undesirable fluctuation. Evil as is commutator hum-in a telephone, it is accentuated in radio sets having vacuum tubes and is more difficult to suppress. Whatever may be said, therefore, as to invention, there can hardly be any question as to the utility of applicant’s combination.

Steinmetz in his 1909 edition of “Theory and Calculation of Transient Electric Phenomena and Oscillations,” published a mathematical formula tor the suppression of pulsations in direct current circuits by series inductance and shunted capacity. In relation to that formula Steinmetz made use of the following language:

“In cases where from a source of e. m. f., which contains a slight high frequency pulsation — as the pulsation corresponding to the commutator segments of a commutating machine — a current is desired showing no pulsation whatever, as for instance for the operation of a telephone exchange, a very high indvotive reactance in series with the circuit, and a condensive reactance in shunt therewith, entirely eliminates all high frequency pulsations from the current, passing only harmless low frequency pulsations at a greatly reduced amplitude.’’

The formula of Steinmetz contemplated a single inductance and a large condenser, instead of a series of smaller- inductances and condensers, which is the feature of appellant’s device. Notwithstanding the fact that Steinmetz was the recognized leader of mathematicians in the electrical engineering profession, his mathematical equations and his discussion of them indicate no actual structure whatever, unless a purely conventional figure to which the mathematical calculations relate can be so considered. Steinmetz’s mathematical calculations unquestionably opened a new field for inventive thought, by establishing that a high inductive reactance in series with the circuit and a condenser reactance in shunt therewith would eliminate all high frequency pulsations from the current and pass only harmless low frequency pulsations at a greatly reduced amplitude. Theoretically the problem was solved. Practically the solution was of no avail, inasmuch as it required for the efficient suppression of undesirable fluctuations a single inductance and a single condenser of such size and cost as to bar its use by industry.

The applicant’s device is a series of five small condensers and five small inductances, or five small resistances, which may be contained in a shoe box. The five sections, with a capacity of 15 microfarads, would cost less than $20, and would nevertheless be 40,000 times as effective as a Steinmetz device of the same size and substantially the same cost. To secure with a Steinmetz single unit filter the same diminution in commutator hum as that obtained by applicant’s five-section filter would require a single condenser costing more than $300,-000, and having an electrical capacity of 600.000 microfarads.

The applicant does not multiply the device of Steinmetz. He subdivides it, and by using five condensers and five inductances or resistances, each one-fifth the size of the corresponding element employed by Steinmetz, accomplishes an enormously greater efficiency at an enormously less cost.

The contention that the Rudenberg patent taught the applicant how to conquer by subdividing cannot be sustained. That patent.was designed to protect an alternating current line from the surges caused by sudden abnormal changes in electrical conditions, and could not be used for the purpose of eliminating objectionable regularly repeated small fluctuations or ripples from a direct current circuit.

The Rudenberg apparatus is a lightning arrester, and was devised to meet an electrical pr-'-’ - - different from that contemplated rester are necessarily costly'high voltage condensers having an eleceither by Steinmetz condensers of the lightning artrical capacity of a few hundredths of a microfarad, whereas the condensers of the appellant are cheap low voltage condensers, with an electrical capacity approximately a few hundred times greater than those of Rudenberg. The Rudenberg apparatus, moreover, admittedly has no material influence on a low periodicity working current, and as it would permit such a current to pass through it substantially unimpeded, it would apparently be of little use in removing unwelcome fluctuations from a direct current line.

Rudenberg suggests the multiplication of his device for greater protection against sudden abnormal surges. The applicant does not multiply the device of Steihmetz, but subdivides it, and thereby at an insignificant cost secures a capacity hundreds of times greater than that required for the lightning arrester, and an attenuating effect thousands of times greater than that produced by the inductance coil and condenser of Steinmetz.

The coadjutant components of appellant’s filter must be properly proportioned, in order to provide the attenuating power required by the commutation frequency, and the applicant discloses a method of determining such proportions which is not suggested by Rudenberg or Steinmetz. The method of computation taught by Rudenberg is directed to a problem entirely different from that which confronted the appellant.

Van Deventer disclosed nothing more than that the generator intended for use in telephone central stations would deliver a smoother current, if it had a larger commutator with many segments. He disclosed no filter, and it cannot be said that he anticipated the applicant, because the applicant combines his filter with a generator having a large number of segments in its commutator.

We cannot hold that, after Steinmetz’s disclosure, appellant’s apparatus became obvious to any one skilled in the art. To say that would simply mean that, faced by a crying need for a device which would remove the objectionable hum from sound-reproducing systems, inventors of electrical appliances and those skilled in the art were blind to the aggressively apparent for more than six years after the announcement of the Steinmetz formula. Problems which vex the brain for many a weary hour and many a weary year become obvious to all the world, once they are solved; but their obviousness after the fact does not necessarily prove their obviousness before the fact.

A careful study of the record on appeal leaves but little, if any,. doubt in our minds as to whether there was invention in this case; but, if we had any doubt, settled law requires that the benefit of it should be given to the appellant.

The decision of the Commissioner must be reversed.