US3466503A

US3466503A - Assisted arc a.c. circuit interruption

Info

Publication number: US3466503A
Application number: US646063A
Authority: US
Inventors: Leon J Goldberg
Original assignee: General Electric Co
Current assignee: General Electric Co
Priority date: 1967-06-14
Filing date: 1967-06-14
Publication date: 1969-09-09
Anticipated expiration: 1986-09-09

Description

SePt- 9, 1969 n.. J. GQLDBERG 3,466,503

ASSISTED ARC A.C. CIRCUIT INTERRUPTION Filed June 14, 1967 5 Sheets-Sheet 1 Q1 Pfg- 2- H2 i 2 z d f a av" F/t'g. /a

E l/ mnu/Mr ,mm/fir mma:

l g l Amal/1mi m ma: M J/ n i fr? Venca/".-

017 d Go/dberzg. Y" r/Mf Afm? www 2Mo by m la GMM/ t r'ney Sept. 9, 1969 L, J, GQLDBERG 3,466,503

ASSISTED ARC A.C. CIRCUIT INTERRUPTION Filed June 14, 1967 s sheets-sheet 2 [771/227 t: or'.- .Leon J Go/dbezjg,

y Hfs A 6% Sept. 9, 1969 L, J, GQLDBERG 3,466,503

ASSISTED ARC A.C. CIRCUIT INTERRUPTION Filed June 14, 1967 5 sheets-sheet s t t t 77M:

w fnl/encor: aan d Go/dbe/:g

H/lS Ab rney Sept. 9, 1969 L.. J. Gower-:RG 3,466,503

ASSISTED ARC A.C. CIRCUIT INTERRUPTION Filed June 14. 1967 5 Sheets-Sheet 4 /27 Vend; or: Lean rf Go/a/erjg.

y H/.s A rney.

Sept. 9, 1969 L. J. GOLDBERG 3,466,503

ASSISTED ARC A.C. CIRCUIT INTERRUPTION Filed June 14, v196'? 5 Sheets-Sheet 5 HAS AZ'J rrmf.

United States Patent U.S. Cl. 317-11 15 Claims ABSTRACT OF THE DISCLOSURE An assisted arc interrupter for A-C power circuits having physically separable contacts and associated arc controlling horn structure is provided with an auxiliary horn for creating two gaps and with at least one assisting diode or controlled avalanche diode so that one gap is un-ionized because the assister diverts or blocks current from that gap, while the other gap arcs over to a current zero and deionizes during the next half cycle. Either single phase or polyphase circuits can be interrupted, and in some applications the assisting diode performs a dual function such as rectifying. For higher voltages the assisting component may be a triggered vacuum gap or ignitron. Substantially arcless interruption is achieved by SCRs or a triac in parallel with the contacts which are triggered on as the contacts start to open and conduct to the next current zero and are prevented from 1being retired repetitively.

This invention relates to the interruption of alternating current circuits by opening physically separable contacts wherein an electric arc tends to form between the contacts -as they are opened. More particularly, the invention relates to assisted arc circuit interrupters for reducing the arc or for effecting substantially arcless interruption in alternating current circuits upon passage of the alternating current through a current zero.

When circuit interrupters or breakers comprising a set of physically separable contacts are employed to switch circuits in which large currents of hundreds or thousands of amperes are flowing, an electric arc discharge normally is established between the two separable contacts as they are opened. This arc lasts for a finite period, and requires special structures to be extinguished. Many different structures have been used for both elongating the arc and driving it into an are chute which absorbs energy from the arc. Some of the commonly used devices of this nature employ a magnetic field transverse to the arc or air blasts either axially aligned with or transverse to the arc, as well as oil baths. `In the case of the magnetic blowout interrupter, a magnetic field transverse to the are is `formed by a series coil and core and accomplishes arc extinction by driving the arc into an associated are chute structure. These structures are large for high current ratings. Additionally, they allow the arc to continue for substantial periods of time thereby introducing undesired switching transients into the power system, and further may require replaceable contacts whose opera-ting life is greatly shortened the longer an arc is allowed to continue. A more recently developed type of interrupter is substantially arcless and comprises a commutating capacitor and a gate controlled conducting device such as a silicon controlled rectifier in series relationship with the physically separable contacts, the silicon controlled rectifier being rendered conductive by the increased potential across the contacts as they open to couple the charged capacitor across the contacts in opposite sense to the load current to force a current zero and apply a reverse voltage that aids in deioninzing the space between the contacts. This arrangement it will be noted Mice requires a pair of oppositely charged commutating capacitors when used in an alternating current circuit.

Accordingly, an object of the invention is the provision of generally improved and more satisfactory assiste'd arc interrupters for alternating current power circuits which produce less arcing and which interrupt the circuit more quickly.

Another object of the invention is to .provide for new and improved assisted arc interrupters for single phase and polyphase power circuits having dual function components or which are operable in such a manner as to provide some function in addition to interrupting the circuit.

Yet another object is the provision of fast, reliable, substantially arcless interruption of an alternating current power circuit by conventional switchgear apparatus modified to employ the assisted arc interruption principle.

In accordance with the invention, an .assisted arc interrupter adapted to be connected in an alternating current power circuit by conventional switchgear apparatus of a set of physically separable contacts between which an electrical arc forms upon the opening thereof, an arc controlling structure including a pair of main arc horns and means for transferring the arc onto the arc horns. In addition there is provided auxiliary arc horn means for creating two gaps for the passage of sections of the arc, and at least one assisting diode connected in parallel circuit relationship with a first one of the gaps for diverting current therefrom following opening of the contacts until the next current zero, whereby the first gap is un-ionized for half a cycle. Upon polarity reversal, the assisting diode blocks current from passing through the second gap so that the second gap deionizes for a half cycle and interruption is achieved.

Diode assisted arc nterrupters are employed in various single phase and polyphase circuits in which the assisting component may serve a dual function such as rectification or supplying reduced starting voltage to a load. For example, a bridge rectifier circuit for delivering d-c power to a load over lines from a single phase or three-phase a-c source conventionally includes a pair of diodes in each line, one on each side of the load, poled to provide rectification. An assisted arc interrupter is combined therewith and comprises a set of physically separable contacts for each line located between the load and an adjacent one of the diodes. Current arcs across the contacts until the next current zero in those lines which are conducting current when the contacts in the various lines are opened simultaneously. They respective adjacent diode in each line serves the dual function of rectifying and as an assisting diode in the assisted arc interrupter by blocking arcing across the sets of contacts, after their opening, during the time the polarity in each of the lines is such that the aforesaid adjacent diode is reverse biased. Thus, upon opening the contacts at least one of the adjacent diodes is blocking so that its respective set of contacts deionizes, opening one of the lines, while at least lone other set of contacts arcs over to the next current zero whereupon its respective adjacent diode blocks so that the other set of contacts is deionized, opening another oil the lines to achieve interruption.

Substantially arcless interruption can be `achieved in an alternating current power circuit by connecting the load terminals of a main gate turn on, non-gate turn off solid state bidirectional conducting means such as a pair of oppositely poled SCRs or a triac directly across the set of physically separable contacts. As the contacts start to open, the rising contact voltage across them is sensed and actuates gating circuit means for turning on auxiliary gate turn on, non-gate turn off controlled conducting means. The auxiliary controlled conducting means cornprises a portion of a trigger pulse forming circuit which also includes a pulse capacitor and a pulse transformer primary winding connected across a source of electric potential. The pulse developed in the pulse transformer primary winding is applied to the secondary winding thereof in a gate-to-cathode circuit of the main bidirectional conducting means, turn ing it on to divert current from the contacts until the next current zero during which time the gap deionizes. To prevent the main bidirectional conducting means from being tired repetitively, zero current in the A-C power circuit after interruption has been achieved is detected and inhibits the gating circuit means from subsequently turning on the auxiliary controlled conducting means.

The foregoing and other objects, features, and advantages of the invention will be apparent from the following more particular description of several preferred embodiments of the invention, as illustrated in the accompanying drawings wherein:

FIGS. 1, 2, and 3 are elementary schematic circuit diagrams illustrating the basic principles of three forms of the invention utilizing diodes as the assisting components;

FIG. 4 is a graph of gap breakdown voltage versus time after a current zero for different dielectric gap materials;

FIG. 5 is a plot 0f circuit voltage versus time after the parting of physically separable contacts and further shows the rise of recovery strength of the gap for two different sizes of gaps;

FIGS. 6(a), 6(b), and 6(0) are schematic diagrams of circuit interrupters operating in accordance with the elementary diagram of FIG. l;

FIGS. 7(a) and 7(b), and FIG. 8 are schematic diagrams of circuit interrupters operating in accordance with the elementary diagrams of FIGS. 2 and 3, respectively;

FIG, 9 is a side View of a design for an assisted arc interrupter of the type shown more schematically in FIG. 6(a), which is a double break bridge contact employing a single assisting diode;

FIG. 10 is a plot of gap breakdown voltage versus time after current zero on which the circuit transient voltage is superimposed, illustrating the operation of controlled avalanche diodes as the assisting electronic components;

FIGS. 11(a) and 11(1)) are equivalent schematic circuit diagrams of a single phase circuit having a diode assisted arc interrupter for each line;

FIG. 12 is a schematic circuit diagram of a threephase circuit having a diode assisted arc interrupter in each line;

FIGS. 13(a), 13(1)), 13(5'), and 13(a') are schematic circuit diagrams and a graph of circuit voltage versus time for a single phase, full wave rectier incorporating diode assisted arc interrupters;

FIGS. 14(a) and 14(b) are respectively a schematic circuit diagram of a three-phase bridge rectifier incorporating diode assisted arc interrupters, and a graph of circuit voltage in the three lines versus time for clarifying its mode of operation;

FIG. 15 is a schematic diagram of an assisted arc interrupter using a triggeed vacuum gap as the assisting electronic component;

FIG. 16 is a schematic cicuit diagram of an arcless assisted arc interrupter employing silicon controlled rectiers; and

FIG. 17 is similar to FIG. 16 but uses triacs as the assisting component and for triggering the assisting triac.

The use of physically separable contacts in a circuit interrupter is desirable in that they offer continu-ous current carrying capability at low voltage drop, small heat generation and reasonably eriicient removal of heat, low losses, and high thermal capacity for short time overcurrent in the range of seconds and minutes. By combining physically separable contacts with electronic switching devices which divert all or a portion of the arc which forms as the contacts are opened, there is produced a marked improvement in the interrupter. The achievement of arc interruption can take place in either of two modes. In one, the arc is suppressed by making its environment such that the power source cannot supply the Voltage necessary to maintain the( arc. This is the common mode of interruption in D-C circuits. It often occurs as an incidental and usually undesirable phenomenon in the interruption of alternating current where the principal mode is to permit the arc to persist until a current zero by the alternation of the source voltage, whereupon the arc fails to reignite because the interrupter recovers dielectric strength faster than the voltage to re-establish an arc develops. Analogous to this second mode, a semiconductor electronic switching device that is conducting current which falls to a zero value due to the circuit external to itself can develop a blocking ability after the current Zero. For diodes, it is necessary that the voltage across the diode reverse at the current zero for blocking to occur. Some multilayer semiconductors do not require the polarity reversal, but in the interrupters considered here, interruption is desired at the point where the current zero and polarity reversal are coincident. As will be pointed out later, electronic switching devices other than semiconductors can be useful in certain situations.

The principles of three forms of the invention employing diodes as the assisting component or components are illustrated schematically in FIGS. 1, 2 and 3 which show the condition of interruption apparatus immediately after the apparatus has been operated to commence the process of circuit interruption. In FIG. 1, two

gaps

21 and 23 are in series and a unidirectional conducting semiconductor diode 22 is connected in parallel with one of the gaps, namely, the gap 21. In FIG. 2, there is provided in addition a second diode 24 which is poled in the opposite direction from the diode 22 and is -connected in parallel with the second gap 23. Looking first at the operation of FIG. 2, as the gaps are formed by contact parting or arc movement, one diode must conduct and the other must block. For instance, assuming that the direction of positive cur- -rent is from left to right, the diode 22 conducts while the `diode 24 is blocking. Consequently, the gap 21 cools and is prevented from ionizing for up to a half cycle. The other gap 23 conducts by arcing during that interval, 'but upon current reversal it deionizes due to the fact that although the diode 24 conducts during this rst full half cycle, the parallel unit of deionized gap 21 and reversely biased diode 22 is in series with it and blocks. At the end of the second full half cycle, both gaps are deionized and each is capable of withstanding circuit voltage. The opposing diodes .do not close the circuit and therefore interruption is achieved. Should the polarity be reversed when the gaps are formed initially, that is the direction of positive current is now from right to left, a mirror image operation occurs in that initially the diode 24 conducts while the gap 23 is un-ionized, `and thereafter upon current reversal the gap 21 deionizes for a half cycle `due to the parallel unit of deionized gap 23 and blocking diode 24. Regardless as to the relative polarity at the time the gaps are created, if at this time only a small fraction of a half cycle remains before reversal, the action described may be postponed for up to one half cycle.

The second gap deionization, whether it be the gap 21 or the gap 23, takes place `because of the blocking action of the first gap-parallel diode unit in series with it, regardless of the fact that there is a diode in parallel with this second gap connected in the direction to divert. Therefore the second diode is useful principally to make the entire unit symmetrical and independent of the polarity at the moment of initial arcing. If it were not used, the circuit would become that of FIG. 1. This circuit operates generally similar to that of FIG. 2, but if the polarity 'of the circuit at the moment of initial arcing is such that the diode 22 blocks, then both

gaps

21 and 23 arc to the end of the half cycle, whereupon the action as described with respect to FIG. 2 commences. There are other secondary differences which will be mentioned later.

An alternative parallel arrangement Iof gaps and diodes shown in FIG. 3 comprises a first gap 25 and diode 26 in parallel with a second gap 27 and diode 28, the diode 28 in the second branch being poled oppositely with respect to the diode 26 in the first branch. When the contacts part in this circuit, one gap (25, for example) arcs over until the first current zeropbecause its associated diode 26 is conducting, while the other diode 28 is blocking and its associated gap 27 is un-ionized. Upon polarity reversal, the first mentioned diode 26 is blocking and the gap 25 in series with it deionizes for a half cycle. At this time both gaps are cool and deionized and can withstand the recovery voltage. Interruption is therefore achieved. In FIG. 3 the assisting diodes are working continuously and are not limited to temporary pulse duty during interruption as is the case with the assisting diodes in FIGS. l and 2. The parallel FIG. 3 circuit, on the other hand, is advantageous in that there is no steady state stress on the diodes and there is no diode leakage current to contend with.

In order for the assisted arc interrupters shown schematically in FIGS. 1 3 to operate, it is observed that the assisting diodes must divert current from a gap for a half cycle, and the assisting diode or diodes must develop dielectric recovery after current zero more rapidly than the recovery voltage across the gap develops. As further explanation of this and of the advantages of the arrangements shown in FIGS. 1-3, the dielectric recovery in an arcing air gap can be represented by a curve (see FIG. 4) of voltage to cause ybreakover or gap breakdown versus time. Depending on the history of the arc just prior to current zero, the curve starts out at approximately the low voltage at which the arc existed prior to current zero and rises rapidly to a voltage in the vicinity of 100 to 200 volts. It then rises more slowly to its ultimate value which depends on the length of the gap and certain features of its electrodes and insulators. In a vacuum type of interrupter, the characteristic curve rises rapidly from its initial value to its ultimate value. The corresponding curve for breakover voltage for a diode or other solid state semiconductor rises somewhat less rapidly from its forward drop value to its ultimate avalanche value. The voltage which develops across an interrupter immediately following either a forced or natural current zero depends on the power factor of the circuit. A unity power circuit factor is easy to interrupt because the steady state voltage which appears at a natural current zero starts from zero and rises at the power frequency rate. The dificult interruption occurs in a low power factor inductive circuit, the limiting case which will be considered being a zero power factor inductive circuit. In this case, the circuit voltage just prior to the current zero is at its steady state crest value. Ultimately substantially this voltage appears across .the air gap or interrupter, but only after gO- ing through a transition voltage which includes a transient voltage of a frequency determined by the capacitances and inductances of the circuit. At a natural current zero, the total of a steady state and transient voltage is (for a somewhat simplified model):

where er is the total recovery voltage instantaneous value, E is the crest of a steady state voltage, f is the power frequency, and n is the ratio of .the natural frequency to the power source frequency as determined by the circuit parameters. The crest of this value is twice that of the power frequency steady state voltage. The transient recovery voltage is illustrated in FIG. 5 and is superimposed upon the steady state voltage which is shown in dotted lines. This graph also shows the characteristic curves of recovery voltage and recovering gap strength for a small air gap and for a large air gap. The small gap reignites the arc at zoint z. The large gap effects a successful interruption. Similarly, a fast recovery semiconductor device would not break down and conduct, thereby achieving interruption.

If the arc is suppressed before the power frequency zero, the recovery voltage rises higher and faster than according to the equation above for interruption at' a current zero. Suppression refers to techniques used to absorb a significant portion of the available voltage and usually increases the speed and magnitude of the recovery voltage. The dielectric must have a much faster recovery rate if it is successfully to maintain a forced current zero as compared to a natural one. If the interrupter is constructed .to work adequately at the natural zero but cannot maintain the forced zero, at interruption at the forced zero the arc will re-esta'blish, the transient voltage will collapse, and this sequence may -be repeated many times. Since this is undesirable, it is an advantage as cornpared to a suppression that cannot be maintained and sometimes as compared to one which can, to allow smooth conduction to the natural current zero and thus maintain the interruption. With assisted arc interruption as for instance in FIGS. l-3, the arc space or gap width in the basic switch is shorter than in conventional interrupters and therefore less likely to cause suppression. The use of a shorter gap results in a cost saving. The assisters considered here are all of the type not to force current suppression but to have fast recovery after a current zero. Therefore, this combination is less likely to have premature interruption and restrike cycling than a conventional air arc interrupter. In FIGS. l-3 the recovery voltage is shared by .the deionizing air gap and its series blocking diode. For example, in FIG. l, after passage of current in the direction of diode conduction, the recovery voltage is blocked by the series combination of the rapidly developing dielectric strength of the diode and the more slowly developing dielectric strength of .the deionizing series gap. The parallel gap has already developed a required dielectric strength because of its multimillisecond period on non-conduction. The slowly developing dielectric strength of the deionzing series gap is by itself inadequate to fully withstand a recovery voltage, and inherently so because for most applications the cost saving resulting from making the gap less capable is one of the objectives of the assisted arc interruption arrangement. Without the assisting diode, this series gap would otherwise have to be larger in order to effect interruption.

In considering the physical forms of interrupters which conform to the electrical elementary diagrams of FIGS. 1-3, the use of an interrupter with the conventional arcing horns onto which the arc is transferred and insulating arc chutes for containing the arc and absorbing energy is still desirable since there would be excessive burning of the contacts if the arc were allowed to remain in a fixed position of them for the one, two, or three half cycles during which the arc may exist with the instant apparatus. The difference as compared to the conventional interrupter is that, first, the arc horn assembly has one or two intermediate members so as to form two or more sections of the arc in series with provision for connection access to these sections and, second, the total length of the arc is greatly shortened. The arc space need be long enough only to hold off voltage after it has cooled some eight to sixteen milliseconds for a 60 c.p.s. power frequency. This reduction in the size of the arc horn and chute assembly provides the cost saving that pays for the assister and its assembly. These considerations appear to apply most favorably over a middle range of voltage and current ratings. It might be added that the heat generation rate, per unit length along the arc path, is about the same as or less than in a conventional interrupter. Since conventional interrupters must also tolerate the arc for several half cycles, the chute structure is similar or simplified as compared to a conventional chute except that it may be much smaller due to the reduction in length along the arc path.

Different forms of physical interrupter apparatus corresponding to the elementary electrical arrangements of FIGS. l-3 are shown respectively in FIGS. 6-8. FIG. 6 (a) employs a single diode and comprises two .sets of contacts effectively connected in series circuit relat1onsh1p provided -by a bridge double break movable Contact assembly 29 which upon separating from the fixed

contacts

31a and 31b forms two arc portions shown in dotted lines. A conventional blowout mechanism of the air blast or magnetic type shown schematically here at 33 moves the arc sections into a single long arc extending between oppositely curved

main arc horns

35a and 35b extending upwardly from the

stationary contacts

31a and 31b and between which is an auxiliary horn member or arc runner 37. A diode 39 is connected between one of the

curved horns

35a or 35b and the auxiliary horn 37. The space available for the arc to move and the arc rate of propagation must provide for the possibility of three normal half cycles of arcing after the sections form. FIG. 6 (b) is a variation of FIG. 6(a) in that two auxiliary arc horns 37a and 37b are provided in the space between

horn sections

35a and 35b. The diode 39 is connected between the two intermediate are horns 37a and 37b and has no other connection to the external terminals.

FIG. 6(6) shows a single break contact type of interrupter wherein an arc forms between a movable contact 41 and a stationary contact 42 as the contacts separate and is driven magnetically upward as indicated by the dotted line by the field of a blowout coil 33 or other means. Arc horn section 45a is connected to the stationary contact 42 while the other main arc horn section 45b is spaced from the movable contact 41 and is parallel to a shorter auxiliary horn 47 which makes connection to the movable contact 41. The diode assister 49 extends between the arc horn 45b and the auxiliary horn 47. As the arc forms, it eventually takes the position indicated for the two sections 1 and 2. Section 1 continues to move upward in the chute formed between horn sections 45a and 45b, while section 2 moves downward between auxiliary horn 47 and arc horn section 45b. It is obvious that FIG. 6(c) operates in accordance with FIG. 1.

The interrupter illustrated schematically in FIG. '7(a) employs two diodes and operates in accordance with FIG. 2. This is a bridge double break movable contact assembly type of interrupter and is identical to FIG. 6(a) except that there are two

diodes

39a and 39b each connected to one of the arc horns. FIG. 7(b) is similar to FIG. 6(c) and differs in that the arc horn section 45b is connected to the movable contact 41 while the intermediate arc runner 47 is located between the main arc horn sections 45a and 45b and provides connection for two diodes 49a and 49b- FIG. 8 shows one possible form of circuit interrupter embodying the parallel arrangement of gaps and diodes of FIG. 3. Here, oppositely poled

diodes

50 and 51 are connected between the line conductor and either stationary contact, and movable bridge double break contact assembly 52 is connected to the other line conductor.

FIG. 9 shows a less schematic view of an assisted arc interrupter of the type shown in FIG. 6(a) employing a single diode and a bridge double break movable contact assembly. Like parts are indicated by the same numerals. The main

arc horn sections

35a and 35b and the auxiliary arc horn 37 are folded to effect reduction of the length of the apparatus. The outline of the chute halves is indicated at 38. The diode 39 is connected externally between the inner end of the auxiliary arc horn 37 and a circuit connection tab 51b on which the stationary contact 31b is supported. As the arc sections are formed between the movable contact assembly 29 and the

stationary contacts

31a and 31b, they assume successive locations as indicated by the numerals 1, 2, 3, and 4. A depending button 53 on the movable contact assembly 29 acts as an intermediate arcing horn for each section of the arc, one section moving into each of the two gaps formed by the arc horns. The gap in parallel with the diode 39 is made wider than the other series gap because this gap, but not the series gap, must withstand the transient surge recovery voltage. As an example of a specific piece of equipment, an assisted arc interrupter of this type for a 10,000 ampere, l5 kv. circuit has an arc chute about 18 inches in length, other dimensions being approximately proportional thereto.

It has previously been stated that in alternating current interruption aftera current zero the circuit develops a recovery voltage which tends to break down the blocking diode and the deionized gap in parallel with it. The recovery voltage reaches instantaneous values in excess of the crest of the power source voltage. The power source instantaneous voltage is 1.4 times the RMS voltage, and the RMS voltage may equal or approach line-to-line value. In addition there is superimposed a transient voltage of higher frequency caused by energy stored in the distributed inductances and capacitances of the circuit. If the arc is not suppressed or chopped before the natural zero, the crest of the transient high frequency voltage will equal the crest of the power source voltage. The total recovery voltage is the sum of the transient high frequency voltage and the steady state power frequency voltage. If there is premature interruption by arc suppression, the recovery voltage is higher. The assisting diode must tolerate whatever inverse voltage is imposed on it. If some arc suppression voltage less than that corresponding to instant suppression of the maximum value of current is expected, it is practical to select diodes with transient peak inverse voltage ratings suicient to withstand this. A value suggested is about 1.5 times the crest that would occur without suppression or chopping, which in turn is 2.8 times the circuit RMS voltage, i.e., about 4.2 times the RMS voltage.

An alternative design to having the diode withstand the fully recovery voltage imposed upon it is to make the assisting diode of the controlled avalanche type. In this modification the different diodes in FIGS. 1-3 are replaced by controlled avalanche diodes. A controlled avalanche diode blocks the passage of current in the reverse direction only for voltages less than a prescribed value for which it is designed. When the voltage equals or exceeds this design value the diode conducts in the reverse direction7 at a voltage which is more precisely equal to its avalanche rating plus the product of the diodes dynamic resistance and current at the instantaneous temperature and current. This results in clipping the reverse voltage applied to the diode. The manner in which interruption takes place can be analyzed with the help of the curve shown in FIG. 10. For a zero power factor inductive circuit, the circuit voltage at a current zero is at its crest value and the transient recovery voltage has an oscillatory form as shown which is described by the equation given previously. When the voltage required to spark over or break down the gap develops, after conduction to a current zero, as fast as shown by curve a there is no further breakdown and interruption is achieved. If, however, it is slower as shown by curve b, then at the point c the gap sparks over and continues to arc. Since an objective of the assisted arc interrupter is to make the gap lsmaller than would be the case with a conventional interrupter, reignition at point c would occur when the assisting diode is blocking.

The controlled avalanche diode shunting the gap is selected to have an avalanche voltage which is higher than the crest value of the power frequency voltage, so as not to require the diode to conduct in the reverse direction under normal steady state conditions. The avalanche voltage is further selected to be less than twice the crest value voltage, which it will be recalled is the peak of the transient recovery voltage curve (FIG. 5). Then for conditions in which the interrupter gap recovery characteristic were like curve b in FIG. l0, and where in the absence of the avalanche diode reverse conduction the gap would spark over at point c, the presence of the avalanche diode causes it to start to conduct at point d and to hold the voltage down to the avalanche voltage level. After having conducted to the current zero, the result is that the main gap does not arc over at all again. Consequently, the interruption takes place more quickly and more reliably and with less consumption of material on the contacts and other parts of the main interrupter. This is accomplished with no added cost or size of the main interrupter, simply by making the diode of suitable avalanche characteristics.

Although not here illustrated, a conventional assisting diode may be protected from the recovery voltage pulse by means of a shunt connected thyrector or thyrite. A thyrector comprises two back-to-back PN junctions and passes rapidly increasing current above rated voltage in either direction. When shunting an assisting diode which in turn is in parallel with a gap, the use of the thyrector enhances the interruption capability of the gap to some extent by reducing the peak value of the reapplied transient recovery voltage. Thurs, referring to FIG. 10, the first peak f the transient recovery voltage curve is reduced in value such that the peak then lies below the point c whereby a gap having a breakdown voltage characteristic as shown in curve b is effective to provide interruption. A thyrite is a non-linear resistance Whose resistance falls rapidly and non-linearly with increasing voltage, resulting in effectively low resistance at high voltages. The thyrite protects the diode in the same manner as the thyrector inasmuch as they have characteristics resembling each other.

The use of a diode assisted arc interrupter in a single phase A-C circuit is illustrated in FIGS. 11(a) and 1l(b). In FIG. l1(a), a source winding 55 on the left is connected in a circuit by means of

line conductors

57 and 59 with a load winding 61 on the right. A contact 63 in line 57 is shunted by a diode 65, while the other line 59 is likewise provided with a contact `67 'shunted by a diode 69. Electrically equivalent to this is the arrangement shown in FIG. 1l(b), wherein contact 67 is in series with the contact 63, the contact 63 being shunted by the diode 65 and the contact 67 being shunted by an oppositely poled diode 69'. The interrupter structure shown in FIG. 1l(b) will be recognized as being the same as that illustrated in FIG. 2 and operates in the `same manner in that when the contacts are opened under load the polarity relations necessarily require one contact to be shunted by a diode in the conducting direction so that its corresponding gap does not ionize, while on current reversal this diode blocks because of its inherent characteristics with respect to polarity and its parallel gap blocks because it is not ionized. The interrupter structure shown in FIG. 7(a) can be used or two conventional interrupters each having its contacts shunted by a diode can be used, it being understood that the conventional interrupter while having arc controlling structure does not have an auxiliary arc horn.

The principle of assisted arc interruption as herein taught can also be applied to the commercially important three-phase circuit. Referring to FIG. 12, Y- connected three-phase source winding 71 is connected to a similarly connected load winding 73 in the conventional manner by lines A, B, and C. In line A are contacts 75a shunted by diode 77a and lines B and C have similar interrupter structures. These may be conventional interrupters. In the polyphase circuit in general, continuity of `current ow in any line is achieved by completion of a circuit through one or more of the other lines. Therefore, in the conventional arrangement of one interrupter structure or switch pole in each line, there are inherently two arc gaps in series without special provision for splitting the arc or for more than a single break per pole.

In the three-phase circuit shown in FIG. 12 for supplying Iconstant voltage power to the load 73, simultaneous opening of the contacts 75a, 7517, and 75e normally would produce arcing in at least two of them if the respective diodes 77a, 77b, and 77awere not present.

With the rectifiers present, one or two of the lines will be carrying current of polarity such that the current is diverted from the contact gap to the diode. Upon reversal of polarity in these lines, respectively, the diodes block because of the reversal and the respective associated gaps block because they have deionized for several milliseconds. lf only one line was of polarity to permit conduction at the moment of contact parting, a second line would come into this polarity in 60 or less. When two lines block, the three-phase circuit is interrupted because the third line can complete a circuit only through one or both of the other two. In like manner, after the arcs are extinguished, the circuits remain open because at least one diode prevents continuity of a circuit through one `or two of the others that may have voltage on them in the conduction direction. It can be shown by the use of the diagrams, symmetrical current being assumed, that total interruption on the three-phase line occurs in 150 to 270 after contact parting. If the line currents are not symmetrical, as occurs when for example there are offsets or D-C components immediately after switching on, at least two lines will have asymmetric current. This means that the diodes in one or two lines will carry more than steady -state current for more time than for the case of symmetrical current. The longer time for deionization of the associated contact, following reduced ionization of that contact because arcing was in the minor loop, makes interruption easier in that line. The line carrying the opposite polarity D-C component will cause more heating in its respective contact, but the third line which will provide the nal interruption also has more deionization time and therefore easier interruption.

The interrupters for the three-phase circuit of FIG. 12 can be used as a motor or other control by closing them one or two at a time. With all three contacts a, 75b, and 75C, open, there is no power ow from the source to the load -but there is an application of source potential to the load. This may be interruptedvby opening a disconnect switch in each line (not here shown). With only one set of contacts closed, 75a for example, three-phase power flows for 1/6 of each cycle or single phase power for 1A cycle. This results in an average power flow of 33.3 percent of full three-phase power. Two lines would carry an average current of 45.9 percent of that of a full three-phase connection. If the load were an induction motor, its starting torque would be 16.7 percent and its running torque would be 33.3 percent of that in the full three-phase connection. When only two sets of contacts are closed, three-phase power ilows -for one-half of each cycle, and single phase power for one-half of each cycle. The average power is 75 percent and the average current in two lines is 93 percent. Starting torque on an induction motor is 50 percent and running torque is 75 percent. Consequently, this equipment provides for reduced voltage application of power to a load as well as assisted arc interruption.

In the application of the diode assisted arc interrupter to a bridge rectiiier and load, it will be seen that some of the rectiiiers have the dual function of also being the assisting diode. The single phase, full wave bridge rectier circuit will be examined first. FIG. 13(a) shows a single phase source winding 79 which in one half cycle passes current through diode 81 to a load 83 and back again through a diode 85. On the other half cycle the current path is through diode 87, load 83, and diode 89. An assisted arc interrupter comprising, respectively, contacts 91a and shunt diode 93a and contacts 91b and shunt diode 93h is connected in each of the lines from the source winding 79. When the line switch is open the eifective circuit as viewed at the source comprises a diode in each line as shown in FIG. 13(b). The same effective circuit when the switch is open is produced by the arrangement shown in FIG. 13(0), wherein the shunt diodes 93a and 9317 are eliminated and the contacts 91a and 91b are moved to positions at the side of the bridge rectifier between the load 83 and the respective

adjacent rectifier diodes

85 and 89.

In the steady state the operation of the circuit shown in FIG. l3(c) is completely symmertcal, and upon random parting of both contacts the contact which is carrying current will arc over until the next current zero. At this time associated diode becomes reverse -biased and stops conducting. In the meantime the contact which was not carrying current has parted to a substantial gap without arcing because its associated diode was reversed biased. Therefore this open contact blocks current in its line, the diode blocks current in the other line, and at the next current zero the contact associated with the latter diode is deionized and capable of blocking. This achieves circuit interruption. If only one of the contacts 91a or 91-b is opened, rather than both at the same time, then half wave rectification is achieved since the other circuit can continue to conduct. For either full interruption or half wave rectification, the contacts are not required to suppress the arc nor to :be inherently capable of withstanding recovery voltage immediately after a current zero in the conventional manner. They must however tolerate arcing current for the normal or modified half cycle. Under some fault conditions the current in the rectifier may become asymmetrical as illustrated in FIG. 13(d). If both of the contacts part at time t1, only one will carry current at that instant, and if it is in the branch which carries the major loop (the shaded area is arcing current) it may not deionize sufficiently in the reverse half cycle. In this case it will require more than one half cycle to interrupt. On the other hand, the contact carrying the minor loop has more opportunity to open successfully. It will assure half wave rectification if the other contact does not arc for more than one half cycle, or it will assure immediate interruption at the first current zero if it happens to be the contact carrying current at the moment of switch opening.

The same principles can be applied to the three-phase full wave rectifier circuit. In FIG. 1-4(a), three-phase source winding 93 supplies power to the load 95 through the full wave rectier comprising diodes 97-102 connected as shown. Respective contacts 103a, 10317, and 103e are connected in each circuit at one side of the rectifier bridge between the load and the adjacent set of diodes in the same manner as for the single phase version. In this polyphase circuit, the diodes of the main rectifier assist the contacts by blocking and commutating the current from one line to another rather than diverting or bypassing current from the contact directly. The effect is that successively each of the

diodes

100, 101, and 102 conducts for 120 and is reverse biased for 240. Thus, under normal loads each contact has an opportunity to deionize for 240. One pair of contacts may part during its non-conducting interval, and unless it has separated by too small a gap by the instant when otherwise its diode would commutate on, it will fail to ignite an arc at this commutation instant. This will constitute interruption on this pole. If it parts during the conduction interval of its associated diode, it will arc until its diode is commutated orf, and if the circuit has not been interrupted on one or the other poles, this pole will be deionized for 240 and will fail to reignite when otherwise it would be commutated on. This constitutes interruption on this pole by a different mode. When one pole interrupts by failure to ignite an arc across its open gap, the pole from which the current would transfer normally continues to conduct until the natural current zero in that pole. This occurs 30 after the normal commutation instant. At that time the circuit is interrupted. This occurs because the diode that had conducted up to that moment becomes reverse biased; of the other two lines the one with polarity in the direction of its associated diode conduction has had its contacts parted without arcing for 30 to 270, and in the remaining -line its diode is reverse biased and its contacts have deionized for up to The operation of the circuit of FIG. 14(z) may be clearer by an example and the use of the voltage versus time curve of FIG. 14(b), wherein the voltages appearing on the positive and negative D-C buses are accented. If the three-pole switch containing contacts 103a, 103b, and 103C is open at random time t1, diode 97 (see numbers above or below curve portions) is conducting voltage to the positive D-C bus, and diode 102 which had been conducting from the negative D-C bus will continue to do so because contact 10301 arcs as it opens under load. At time t1 there is no conduction through contacts 103m` and 103b so they part without arcing. Normally (with the switch closed) when the voltage on line C became less than that on line A, at the point 105, the conduction would transfer to diode 100. But with the contact 103C open and un-ionized, this circuit cannot act to cause transfer. Therefore diode 102 continues to conduct until the current through it and the voltage across it reverse, at'

time t2. At this moment line VCl does not have negative voltage; line A cannot conduct on the negative leg. Line B does not have negative voltage on the negative leg, is open circuited on this leg by its open contact and could not by itself provide continuity anyway. Therefore, the negative bus is open circuited and interruption is achieved. Had the random opening occurred at time t0, the gap at contact 103a might have been small enough to arc over at the normal commutation point. If so, this sequence would have continued as described, and if not, a similar sequence would have completed interruption at time t3.

On inductive circuits interruption takes place at a moment when energy is stored in the inductance of the D-C circuit. Therefore, it is desirable to provide a discharge or coasting diode across the load as indicated in dotted lines at 107. The voltage requirements of the rectifier diodes are not different than for the circuit using a conventional interrupter, providing a discharge or coasting rectifier is used. When the poles of the interrupter switch are independently actuated, it is possible to provide three levels of D-C power by closing one, two, or three poles, much in the same manner as for the single phase circuit. In both cases, the reduced power is supplied at the expense of increased ripple.

Assisted arc interrupters employing diodes or controlled avalanche diodes, possibly protected by thyrectors or by thyristors, have been described up to this point; however, other electronic devices can be used to advantage as the assisting component. When triggered devices are used instead of diodes, the assister can for some devices or combinations become independent of polarity reversal, but not independent of a natural current zero unless added circuitry to force a current zero is used. For circuits having currents and voltages in the order of tens of kilovolts at kiloamperes or higher, the characteristics of the triggered vacuum gap suggest that its compact size and high pulse current rating, high voltage rating and fast reproducible operation and fast recovery of dielectric strength may make it a Idesirable assisting component. The triggered vacuum gap is described for instance in Patent No. 3,087,- 092 to Lafferty which is assigned to the same assignee as the present invention. It has three characteristics that makes its application different from that of the diode assisters previously discussed. The first and second are that it conducts for either polarity and must be triggered on, and the third is that its voltage drop is significantly difierent. Since it conducts by carrying current through an ionized metal vapor, it has an arc drop that may have the same magnitude of voltage as a short gap between mechanical contacts or arcing horns. The characteristic arc drop voltage rises with current. The arc drop between arcing horns may rise or fall, depending on the construction, but there is a possibility that when the triggered vacuum gap is in parallel with a gap between arc horns in air or gas that the arc will not positively transfer from the arc horn gap to the triggered vacuum gap over the full range of current of interest. This uncertainty can be avoided by arranging the design so that when the arc forms on the arcing horns, it automatically inserts a resistance or reactance in that brench of the circuit. This should be so designed that the voltage of the arc between the horns plus the IR drop in series with it is always greater than the drop across the triggered vacuum gap. This situation is of course unstable and the arc transfers to the triggered vacuum gap, where it extinguishes at the next current zero with much the samf capability as in a vacuum switch.

The arrangement is shown in FIG. 15 in a single break type switch. Stationary contact 111 is connected to input conductor 113 and is associated with a iirst arcing hornl 115. Movable contact 117 is coupled through a flexible lead 119 to the output conductor 121 and is associated with the second arcing horn 123. A three-electrode triggered vacuum gap device 125 has its two load electrodes connected respectively to the input line 113 and the output line 121. The central triggering electrode is energized from a firing circuit 127 which is connected at either side in the output line 121 and which has another connection 129 to the input line 113. A resistor 131 is coupled between the second arcing horn 123 and the output lead 121. It is seen that after the contacts 111 and 117 lhave parted, the arc shown here by dotted line 133 transfers to the arcing horns 115 and 123 and is now in series with the resistor 131. The firing circuit 127 has in thevmeantime fired the triggered vacuum gap 125 which conducts in one direction or the other depending on the polarity at which the contacts are parted. Since the voltage drop across the series combination of the arc 133 and the resistor 131 is greater than the drop across the triggered vacuum gap 125, the arc transfers to the triggered vacuum gap where it extinguishes at the next current zero. Both the triggered vacuum gap 125 and the dionized air gap between the contacts 111 and 117 block at the next current Zero, the triggered vacuum gap because of its inherently fast dielectric recovery capability, and the air gap because it has deionized lfor an interval of up to about one half cycle (60` c.p.s. basis). The interruption is normally complete at the first current zero or at the second current zero if the contacts part too close to the rst current zero to develop a sufficient almost cold gap to hold off the recovery voltage. A pair of oppositely poled ignitrons may be substituted for the triggered vacuum gap device 125 in FIG. 15, two being required because the ignitron conducts only in one direction. A possible overriding advantage is that ignitrons may be available with especially low arc drops, and these promote simplicity by removing the need for circuit components such as the resistor 131 for assuring the transfer of the arc to the ignitrons.

Substantially arcless interruption is achieved by the use of solid state gate turn on, non-gate turn off multilayer devices such as silicon controlled rectifiers and triacs as the assisting component to divert current in an assisted arc interrupter. As used in this context, the term arcless means that there is no significant arcing that requires containment by arc chutes or insulation barriers nor is there sufcient arcing to cause significant erosion of the metal parts. However, the interrupter to be described should not be considered as being absolutely free of arcing since there will probably be an arcing of a few mils of length that will last a few microseconds. It is desirable that the interrupter contacts `have a simple arc controlling structure that is provided by the construction of the contacts themselves.

The use of silicon controlled rectiliers (SCRS) as the assisting components is illustrated in FIG. 16. This circuit comprises a pair of high voltage alternating

current supply terminals

135 and 137 across which are connected in series circuit relationship a load 139 and the interrupterx main contacts 141. The main contacts 141 are ordinary single break or double break contacts, the movable portions of which are separated from the fixed portions substantially simultaneously with a single unitary motion to create a gap. Auxiliary contacts which require mechanical synchronization are not required. A pair of oppositely poled

main SCRs

143 and 145 are connected in parallel circuit relationship with the main interrupter contacts 141 by having their respective load terminals connected directly across the main contacts, and together provide a gate turn on, non-gate turn off solid state bidirectional conducting means. To render the SCR 143 conductive when the polarity of the terminal is positive and the terminal 137 is negative, its gate electrode is energized by a pulse lfrom a pulse transformer having a secondary winding 147s connected between the gate and the cathode. The SCR becomes conductive when the terminal 137 is positive and the terminal 135 is negative upon the application to its gate electrode of a triggering pulse from a pulse transformer whose secondary winding 149s is connected in its gate to the cathode circuit.

The operation wil be described brieiiy before proceeding to a detailed description of the other portions of the circuit. When the main contacts `141 open sufliciently to develop a few volts across them in the order of about 5-20 volts, this rising contact voltage is sensed and causes a trigger pulse to be applied to the pulse transformer secondary winding 147s or 149s, according to the polarity in the high voltage A-C lines, to fire the SCR 143 or the SCR 145. The SCR which is rendered conducting remains conductive until the next current zero, and during this time the gap between the main contacts 141 deionizes sufficiently to withstand the recovery voltage. Upon polarity reversal the SCR which had been conducting becomes blocking and is prevented from retiring at the next subsequent polarity reversal by circuitry which stops the generation of repetitive pulses by the open circuit voltage across the main contacts after the current has been interrupted. The other SCR, which was originally in blocking condition at the time that the main contacts 141 opened, remains nonconductive throughout this entire sequence of events. The interruption is substantially arcless since, as has been pointed out, only a very small arc develops across the opening contacts 141 before one of the SCRs is triggered on, which occurs within about 10 microseconds after sensing a relatively low contact voltage across main contacts 141 and operates to divert current from the main contacts before appreciable contact voltage develops.

The energy for the pulse supplied to the pulse transformer secondary winding 147s associated with the SCR 143 is derived from a low voltage D-C source having a positive terminal 151 and a negative terminal 153. Connected between these two D-C supply terminals is the series combination of a charging resistor 155 and a pulse capacitor 157. To complete this trigger pulse forming circuit, one end of the primary 'winding 147p of the pulse transformer is connected to the junction between the resistor 155 and the capacitor 157, while the other end is connected through the anode to cathode path of an auxiliary SCR 159 to the D-C negative terminal 153. With the SCR 159 in blocking condition, the capacitor 157 charges to the value of the low voltage D-C supply in the indicated polarity, and is discharged through the primary winding 147p of the pulse transformer when the auxiliary SCR is turned on to provide a triggering pulse for the main SCR 143.

A gating pulse is applied to the gate of the auxiliary SCR 159 when a relatively low contact voltage is sensed across the main contacts 141 as they open, assuming that the contact voltage has a polarity such that point 161 is positive with respect to point 163. To accomplish this, a voltage dividing network comprised of the series combination of a resistor 165 and a Zener diode 167 is connected across main contacts 141 between

points

161 and 163. With the main contacts closed, the difference in potential between

points

161 and 163 is very small,

but rwhen the main contacts are opening a contact voltage appears between the

points

161 and 163 Kwhich starts from close to zero volts and rises rapidly. The Zener diode 167 protects the gating circuitry for the auxiliary SCR 159 from the high A-C supply voltage. To provide the gating circuit, the base of a PNP transistor 169 is connected to the junction between the resistor 165 and Zener diode 167, and its collector is connected to the parallel combination of a capacitor 171 and resistor 173, this parallel coupling network in turn being in series circuit relationship with a protective blocking diode 175 connected to the gate electrode of the SCR 159. Assuming for the moment that the transistor 169 is conducting due to the appearance of an appropriate voltage at its Ibase electrode, current flows out of the collector through the coupling network comprised of the resistor 173 and capacitor 171 and through the diode 175 to apply a gating signal to the SCR 159 turning it on. This circuitry thus effectively serves to sense the rising contact voltage appearing across the main contacts 141 as they start to open and to rapidly apply a gating signal to the SCR 159.

The auxiliary SCR 159 is Ifurther prevented from being tired repetitively due to sensing an open circuit voltage across main contacts 141 after the main SCR 143 has diverted current to the next current zero and has been -in blocking condition for a hal-f cycle because of polarity reversal of the high voltage A-C supply. To do this, zero main line current is detected and the gating circuit is inhibited, For this purpose, the transistor 169 is actually a switch which is rendered nonconductive when there is no current in the main alternating current power circuit through load 139. To sense the load current, there is connected in series with the load 139 and main contacts 141 the primary winding 177p of a current transformer. The terminals of the secondary winding 177s of the current transformer are coupled through the cathode to anode paths of a pair of

diodes

179 and 181, one of which is conducting while the other is blocking on alternate half cycles, to the junction between the resistor 165 and the Zener diode 167. A resistor 183 is connected between the emitter of the transistor -169 and the center tap on the secondary winding 177s. When the main contacts 141 are separated and the contact voltage begins to rise, the voltage at the junction between the resistor 165 and the Zener diode 167 appears at the base of the transistor 169 and at the anodes of the

diodes

179 and 181, and there is current in the emitter-base junction of the transistor 169 to turn on the transistor amplier and switch 169, thereby energizing the gating circuit of SCR 159. At the current zero there is no current in the current transformer secondary winding 177s and transistor 169 turns 0E. It is also nonconducting after interruption has been achieved, when the auxiliary SCR 159 could be turned on if a gating pulse appeared, to prevent repetitive retiring of main SCR 143.

In the event that the main contacts 141 are separated when the voltage across the contacts is negative at point 161 and positive at point 163, the main SCR 145 is turned on to divert current from the main contacts to the next current zero and is prevented from being retired repetitively. The circuitry is similar to that for the other main SCR 143 and will be reviewed only briefly. Across the low voltage

D-C supply terminals

151 and 153 are connected the series combination of a charging resistor 185 and a pulse capacitor 187 which normally charges to the supply voltage with the polarity indicated. Upon discharge of the capacitor 187 by turning on the auxiliary SCR 189, a triggering pulse is developed in the pulse transformer primary winding 149p. The auxiliary SCR 189 and the auxiliary SCR 159 together comprise auxiliary gate turn on, non-gate turn off conductivity controlled conducting means for the trigger pulse forming circuits which are responsive to 'both polarities of the alternating current power circuit. In the cathode-to-gate circuit of the SCR 189 are a blocking diode 191 and a protective diode 201, their junction being tied to the negative D-C supply terminal 153.

The rising low contact Voltage across the separating main contacts 141 is sensed at the junction between a series connected resistor 203 and a Zener diode 205 coupled to the

points

161 and 163. The gate circuit protecting Zener diode 205 is connected in reverse direction as compared to the corresponding Zener diode 167. A triggering pulse is supplied to the cathode-to-gate circuit of the auxiliarly SCR 189 by means including a diode 207 connected between the -cathode and the blocking diode 191, the diode 207 further being in series circuit relationship with the parallel combination of a resistor 209 and capacitor 211 which are in turn connected to the collector of an NPN transistor switch 213. The base electrode of transistor 213 is coupled to the junction between the resistor 203 and the Zener diode 205, and its emitter is coupled through resistor 215 to the center tap position of another secondary current transformer winding 177s whose terminals are each coupled to the junction point between the resistor 203 and the Zener diode 205 by the anode to cathode paths of respective diodes 217 and 219.

When the main contacts 141 separate, the developing contact voltage while still at a low value is sensed and applied to the cathode of auxiliary SCR 189 so as to depress this cathode Voltage below the gate voltage of the SCR 189. The circuit is completed through the transistor switch 213 so long as there is current in its base electrode and through the parallel connected resistor 209 and capacitor 211. After the main SCR 145 is tired and conducts to thenext current zero, it blocks for the next half cycle, and at the second current zero following separation of the main contacts 141, the transistor switch 213 is nonconductive due to its connection to the secondary winding 177s' of the current transformer. When the primary winding 177p of the current transformer senses a current zero or when there is no current in the main circuit after interruption, transistor 213 is off and main SCR 145 cannot be turned on. Thus, the SCR 145 is prevented from retiring repetitively after having initially conducted to the tirst current zero following separation of the main contacts. During this time, of course, the gap between the main contacts 141 has had time to deionize and since the SCR 145 does not retire, interruption is achieved.

In the modification shown in FIG. 17,v a main gate turn on, non-gate turn olf solid state bidirectional conducting device in the form of a triac 221 has its load terminals connected directly across the main contacts 141. The triac 221 is rendered conductive by a triggering pulse applied to a pulse transformer whose secondary winding 223s is located in the cir-cuit between the gate electrode and one of the load terminals of the triac. Regardless as to the polarity of the half cycle of the supply A-C voltage between the terminals and 137 at the time that the main contacts 141 are separated, the triac 221 is triggered on and conducts to the next current zero and is prevented from being retired at this next current zero. The circuit for producing a trigger pulse in the pulse transformer secondary winding 223s is similar to that described in FIG. 16 except that a triac is also used in this circuit in the place of the pair of SCRs. Thus, the series combination of charging resistor 225 and pulse capacitor 227 are connected across the low voltage

D-C supply terminals

151 and 153, and the series combination of the pulse transformer primary winding 223p and an auxiliary triac 229 are connected across the capacitor 227. Upon tiring the triac 229, the capacitor 227 discharges and develops a pulse in the pulse transformer primary winding 22311.

To actuate the gating circuit of the auxiliary triac 229, the developing contact voltage across the main contacts 141 as they begin to separate is sensed by a resistor 231 connected in series with two oppositely poled

Zener diodes

233 and 235 which protect the triac gating circuit from the high voltage A-C supply regardless as to the polarity at which the main contacts 141 separate. The gating circuit for the triac 229 comprises the parallel combination of a resistor 237 and capacitor 239 connected between the gate electrode of the triac 229 and the juncture of the resistor 231 and a ydiode 241. The diode 241 is connected in series circuit relationship between the resistor 231 and one of the Zener diodes 233. Another diode 243 has its cathode connected to this same juncture and its anode coupled to the A-C terminal 137. The gating circuit further includes a pair of back-to-

back diodes

245 and 247 connected across the oppositely poled

Zener diodes

233 and 235, and the juncture between

diodes

245 and 247 is coupled to one terminal of the base resistor 249 of an NPN transistor 251, which functions as a switch in the gating circuit to be turned olf when zero current is detected in the load circuit. To this end, the collector of transistor 251 is coupled to the juncture between the cathode of triac 229 and the negative D-C terminal 153, and the emitter is connected to either end of the current transformer secondary winding 177s through respective similarly poled

diodes

253 and 255. The current transformer secondary winding 177s is center-tapped to a limiting resistor 257 which is also connected to the junction between the pair of back-to-

back diodes

245 and 247.

In the operation of the gating circuit for the auxiliary triac 229, assuming that the terminal 135 of the high voltage A-C source is positive with respect to the terminal 137, the rising contact voltage across the main contacts 141 as they start to open is sensed by the series combination of resistor 231 and oppositely poled

Zener diodes

233 and 235. Upon sensing a relatively low contact voltage in the order of 20 volts, a gating pulse is applied through the coupling network comprising resistor 237 `and capacitor 239 to the gate of triac 229, the circuit being completed through the conducting transistor 251, diode 253, one-half of the current transformer secondary winding 177s, resistor 257, and the now forward biased diode 247 to the negative terminal 137 of the high voltage A-C supply. Upon tiring the auxiliary triac 229, the pulse capacitor 227 discharges through the pulse transformer primary winding 223p, thereby producing a pulse in the secondary winding 223s of the pulse transformer which gates on or renders conductive the main triac 221. This occurs rapidly, within about a period of time on the order of ten microseconds or so. The main triac 221 conducts until the next current zero to divert load current from the main contacts 141, and when polarity reversal occurs, the main contacts are deionized and both the main contacts and the main triac 221 are blocking. At the current zero, the zero load current condition is detected by current transformer 177p, and there is likewise no current in secondary winding 177s so that transistor 251 is biased to render it nonconductive. In this manner the reapplication of gating pulses to `auxiliary triac 229 is prevented. If at the moment the main contacts 141 part the point 163 is positive with respect to the point 161, the gating pulse is applied through diode 243 and the resistor 237-capacitor 239 combination to the gate of triac 229, and the gating circuit is completed through conducting transistor switch 251, diode 255 and half of the current transformer secondary winding 177s, Iand thence through resistor 257 and diode 245. In the same manner, transistor 251 turns off when a current zero or zero load current condition is detected to prevent the reapplication of gating pulses to the auxiliary triac 229 which in turn would gate on the main triac 221 connected across the matin contacts 141.

Although not here illustrated, a variety of circuits for pulse forming and detection of current zero is possible drawing upon the already developed state of the art as applied by anyone skilled in solid state circuits. The significant point of the operation of the arcless A-C interrupters illustrated in FIGS. 16 and 17 is the diversion of current from the mechanical contacts within a few microseconds after the contacts part to a solid state device that conducts and interrupts arclessly at the next current zero and is kept from being retired repetitively so as to conduct after the current has reached zero. Its advantage for .application to interrupters'for alternating current circuits is that the system s simpler than others and therefore likely to be less costly and more adaptable to use within the housing of an associated equipment. Another advantage is that the diversion of current from the contacts is timed to within a few microseconds after contact parting independent of mechanical pilot devices and wear on the contacts. This circuitry can 'be added to any conventional switchgear without further modification of the switchgear to have it perform substantially arclessly. The use of two oppositely poled SCRs to divert the line current as illustrated in FIG. 16 performs successfully on faster recovery voltages than the systems employing a single triac as shown for example in FIG. 17. This is because the triac must stop conducting in the face of a recovery voltage in the direction in which the triac could conduct if its gate were excited, whereas the individual SCRs have an interval of half a cycle in which to recover before voltage in their respective forward directions is reapplied.

While the invention has been particularly shown and described with reference to several preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention.

What I claim as new and desire to secure by Letters Patent of the United States is:

1. In an assisted arc interrupter adapted to be connected in an alternating current power circuit and comprising at least one set of physically separable contacts between which an electrical arc forms as said contacts start to separate, and arc controlling means associated with said contacts including at least two main arc horns and means for transferring the arc onto said arc controlling means, the improvement wherein said arc controlling means further includes an auxiliary arc horn for creating two gaps for the passage of sections of the arc, and

at least one assisting diode connected in parallel circuit relationship with a first one of said gaps for diverting current from said rst gap following opening of said contacts until the next current zero, whereby said rst gap is un-ionized while the second of said gaps arcs over,

said diode then assuming a blocking condition to block current from passing through the second of said gaps whereby said second gap deionizes Iand the circuit is interrupted at the next polarity reversal.

2. A construction as defined in claim 1 wherein there are two of the sets of contacts connected in series circuit relationship, and

wherein two of the assisting diodes are provided each connected between one of said main arc horns and said auxiliary arc horn.

3. A construction as defined in claim 1 wherein there is only one set of the physically separable contacts, and

4. A construction as defined in claim 1 wherein said set of contacts comprises a pair of stationary contacts and a bridge double break movable contact assembly having a pair of mating movable contacts, and

wherein said auxiliary arc horn is located intermediate said main horns but unequally spaced therefrom so that said first gap is larger than said second gap,

there being only one said assisting diode connected between one of said main arc horns and said auxiliary arc hom in parallel across said larger first gap.

5. A construction as defined in claim 1 wherein the assisting diode is replaced by a controlled avalanche diode for clipping the applied reverse voltage so that the first gap can withstand the recovery Voltage.

6. In a bridge rectifier circuit for delivering direct current power to a load over lines from a single phase or polyphase alternating current source and including in each line thereof a pair of diodes, one on each side of the load, poled to provide rectification, the improvement which comprises the combination therewith of an assisted arc interrupter comprising a set of physically separable contacts for each of the lines located between the load and an adjacent one of the diodes and between which an electrical arc forms until the next current zero in those of the lines which are conducting current when said sets of confacts are opened substantially simultaneously to create a gap between each set of contacts,

the respective adjacent diode in each of the lines serving the dual function of rectifying and as an assisting diode in said assisted arc interrupted by blocking arcing across each of said sets of contacts after the opening thereof during the time that the polarity in each of said lines is such that the respective adjacent diode is reverse biased,

whereby upon opening the sets of contacts at least one of the adjacent diodes is blocking so that the gap between its respective set of contacts is 11n-ionized, opening one of the lines, while at least one other of the sets of contacts arcs over to the next current zero whereupon its respective adjacent diode becomes blocking so that the gap between said other set of contacts is deionized, opening another of the lines and thereby achieving interruption.

7. A construction as defined in claim 6 wherein said alternating current source is a single phase source and the bridge rectifier comprises two pairs of similarly poled diodes connected at either side of the load,

the assisted arc interrupted having two of the sets of physically separable contacts operable independently to achieve half wave rectification when only one of said two sets of contacts is opened.

8. A construction as defined in claim 6 wherein said alternating current source is a three-phase source and the bridge rectifier comprises three similarly poled diodes connected at either side of the load,

the assisted arc interrupted having three of said sets of physically separable contacts,

said circuit further including a coasting diode connected across said load.

9. In an assisted arc interrupter for connection in an alternating current power circuit to effect substantially arcless interruption, said interrupter comprising a set of physically separable contacts and main gate turn on, nongate turn off solid state conductivity controlled bidirectional conducting means having its load terminals connected directly across said separable contacts for diverting current therefrom as they start to open until the next current zero, at which time said main bidirectional conducting means inherently turns olf and the gap between the opened contacts is un-ionized thereby establishing interruption, the improvement wherein the set of physically separable contacts comprises movable contact portions which are separated from fixed contact portions substantially simultaneously with a single unitary motion, and further including trigger pulse forming circuit means responsive to both polarities of the alternating current power circuit and comprising auxiliary gate turn on, non-gate turn off solid state conductivityA controlled conducting means connected in circuit relationship with at least one pulse transformer primary winding and at least one pulse capacitor across a source of electric potential,

means for sensing the rising contact voltage across said separable contacts as they start to open,

gating circuit means responsive to the rising contact voltage for turning on said auxiliary conductivity controlled conducting means when a relatively low contact voltage is sensed to develop a trigger pulse in said pulse transformer primary winding,

said trigger pulse being applied to a secondary winding of said pulse transformer connected in a gate-to cathode circuit of said main bidirectional conducting means, thereby turning it on to divert current from said opening set of contacts before appreciable contact voltage develops, and

means for detecting zero current in said alternating current power circuit and for inhibiting said gating circuit means from subsequently turning on said auxiliary conductivity controlled conducting means to prevent retiring of said main bidirectional conducting means after interruption has been achieved.

10. A construction as defined in claim 9 wherein said set of physically separable contacts are connected in series circuit relationship with a load, and

said means for detecting zero current in said alternating current power circuit comprises means for detecting current through the load, and

said means for inhibiting the gating circuit means for said auxiliary conductivity controlled conducting means comprises switch means for completing the gating circuit which is conductive when load current is detected and non-conductive when zero load current is detected by said means for detecting current through the load.

11. A construction as defined in claim 9 wherein said means for detecting zero current comprises a current transformer having its primary winding connected in series circuit relationship with said set of physically separable contacts, and

said means for inhibiting said gating circuit means comprises atleast one transistor having its base electrode coupled to said contact voltage sensing means and its emitter electrode coupled to the center tap of the secondary winding of said current transformer, said transistor being rendered nonconductive when there is zero current in said alternating current power circuit.

12. A construction as defined in claim 9 wherein said main bidirectional conducting means comprises two oppositely poled silicon controlled rectifers each having a pulse transformer secondary winding in its respective gateto-cathode circuit.

, 13. A construction as defined in claim 9 wherein said auxiliary conductivity controlled conducting means comprises a pair of similarly poled auxiliary silicon controlled rectifers each connected in series circuit relationship with a sad pulse tarnsformer secondary winding, the series circuit thus formed being connected in parallel with a said pulse capacitor which is in turn operatively coupled across said source of electric potential to thereby provide a separate trigger pulse forming circuit means for each polarity of said alternating current power circuit,

there being a separate one of said sensing means, said gating circuit means, and said means for detecting zero current in said altern-ating current power circuit and for inhibiting said alternating current power circuit and for inhibiting said gating circuit means after interruption has been achieved, for each of said auxiliary silicon controlled rectifiers.

14. A construction as defined in claim 13 wherein each of said sensing means for sensing the contact voltage across said separable contacts as they start to open comprises a voltage dividing network including a/resistor and a Zener diode, and wherein 2'1 22 said main bidirectional conducting means comprises a FOREIGN PATENTS pair of oppositely poled silicon controlled rectiers each having a pulse transformer secondary Winding 511702 5/1944 Great Brftafn in its respective gate-to-cathode circuit. 775,620 5/1957 Great Bfltam' 15. A construction as defined in claim '9 wherein said main bidirectional conducting means is a triac. 5 JOHN F- COUCH Primary Examiner J. D. TRAMMEL, Assistant Examiner U.S. C1. XJR.

References Cited UNITED STATES PATENTS 1,755,324 4/1930 Jacobs 317-77 10 3,273,018 9/1966 Goldberg 317-11 3,309,570 3/1967 Goldberg 317-11