US2421148A - Electromagnetic relay circuit - Google Patents

Description

May 27, 1947. M HADFlELD 2,421,148

ELECTROMAGNETIC, RELAY CIRCUIT Filed May 22, 1943 FiGi ae. IA

INVENTOR. BERTRAM MORTON HADFIELD T ATTOR N EY- Patented May 27, 1947 UNITED STATES PATENT OFFICE ELECTROMAGNETIC RELAY CIRCUIT ware Application May 22, 1943, Serial No. 488,026 In Great Britain June 12, 1942 Ciaims.

The present invention concerns improvements in or relating to electromagnetic relay circuits and has for its object the improvement of the circuit conditions under which electromagnetic relays function, with particular reference to the case where the relay is intended to release slowly, and without detriment to the normal require ments as regards ampere-turns for operation or maximum release lag after operation. A further object is to enable the steady state flux to be obtained in a period of time shorter than that required to operate other magnet systems commonly used in automatic telephony, so that the breakdown of such systems becomes only dependent on the operate and release times of such other magnets.

According to one feature of the invention, the buildup of flux in the relay on application of the battery is made dependent on a circuit comprising the relay and a series resistance, whilst the decay of flux in the relay on removal of the battery is made dependent on a circuit comprising the relay and a shunt resistance. The series resistance may be permanently connected in series with the relay and the operating contact and battery, whilst the shunt resistance may be connected to the relay via a non-linear device which only becomes effective in completing the shunt circuit when the operating contact opens.

According to a further feature of the invention, the steady state ampere-turns or flux in the relay is maintained at normal values by reducing the number of turns on the relay and increasing the current from the battery, by comparison with a relay operated directly from the battery, without necessarily increasing the current density in the wire or the permissible wattage dissipation of the relay.

According to a further feature of the invention, the time for the flux to attain the steady state upon closure of the relay circuit is made substantially equal to the time constant of the relay circuit including the series resistance, by shunting the latter with a suitable condenser. The value of the condenser may be such as to give a minimum of overshoot on the steady state flux, or may be such as to give a greater overshoot with the object of ensuring that magnetic saturation always takes place even when the closure time of the circuit is as short as the above time constant, so that the release-flux-time-decay curve is at least maintained. In this manner the static release lag is maintained for all circuit closure times down to that which is necessary to produce the steady state flux, and which time as has been stated, is substantially equal to the time constant of the relay circuit including the series resistance.

The invention will be better understood by reierring to the accompanying drawing, in which:

Fig. 1 is a schematic diagram illustrating the use of a series resistor and a shunt rectifier to control the waveform of current flowing through the holding relay of an impulsing circuit,

Fig. 1A is a modification of Fig. 1 wherein the shunt rectifier is replaced by additional contacts on the impulsing relay,

Fig. 2 is another modification of Fig. 1 wherein a condenser is bridged across the series resistor to accelerate the build-up of current in the holding relay during impulsing,

Fig. 3 is an extension of Fig. 2 illustrating the use of an additional rectifier to enable the holding relay to be held over another circuit without affecting the operation of a magnet controlled by the impulsing contacts, and

Fig. 4 illustrates the application of the foregoing basic circuits to a two-motion stepping switch.

In order that a better appreciation of the invention may be obtained it is first necessary to establish the essential conditions for ensuring that the normal operating ampere-turns and release time may be obtained, and secondly to establish the conditions for ensuring a substantial reduction in operate time without conflicting with the former.

It can be shown that the release time of a relay with a short-circuited winding, such as a slugged relay, where the release is initiated by a disconnection of the battery from the operating coil, is due entirely to the time constant of the short-circuited winding. It can also be shown that the time constant of a winding on a relay of a given magnetic circuit, is proportional to the Wound area divided by the product of the specific resistance of the metal and the length of the mean turn (by wound area is meant the total cross-sectional area of the winding, exclusive of insulation or air space). If the latter product be taken as constant for both operating and short-circuited windings, then the time constants of such windings are proportional to the wound areas. Hence the release time of a relay having an operating winding and a short-circuited winding or slug is proportional to the wound area of the latter.

The winding space factor for enamel wire is approximately 0.6 over a wide range of commonly used gauges, and hence a coil fully wound with such wire will have a time constant equal to 0.6

that of a solid slug of the same metal as the wire and having the same shape and total volume. As the slug, or a short-circuited winding, can never fill all the available winding space and in fact in a typical relay at least 0.4 of the space must be left for the operating winding, it follows that the release time of a fully wound relay when short-circuited is at least equal to that of a normal slugged relay and is greater than that of an auxiliary relay with a short-circuited winch ing. Hence if the relay be fully wound and effectively short-circuited on release, the release time will be at least as long as for a normal slugged relay, and independent of the number of turns or resistance of the relay provided the normal operating ampere-turns are obtained.

As regards the steady state ampere-turns, it can be shown that they are equal to the product of the current density in the wire and the wound area. normal for a given class of relay, the normal ampere-turns can be obtained without any increase in current density, and independently of the number of turns or resistance.

As regards the permissible wattage dissipation of the relay it can be shown that the watts are equal to the product of the square of the ampereturns and the resistance of a single turn winding filling the wound area. The wattage dissipation will therefore be normal if the wound area is normal and the ampere-turns are normal, and is independent of the number of turns or resistance.

To sum up therefore, it can be seen that provided the relay is wound normally, and preferably fully, then the normal ampere-turns and time GUEST/all! can be obtained with normal requirements as to current density and dissipation and independently of factors such as steady state current or voltage or number of turns or resistance. It follows that the manner in which the necessary ampere-turns are obtained on operation will not 'aiiect the normal requirements during the steady state period of operation nor the release lags obtainable by short-circuiting on release.

It can be shown that theoperate time of a relay is proportional to the product of the ampere-turns and number of turns divided by the battery voltage. For a given, normal value of ampere turns, therefore, the operate time can be reduced by increasing the battery voltage or reducing the number of turns, or both. Generally speaking the battery voltage is fixed by other considerations, and in what follows it will be assumed that only the turns may be reduced.

It can be shown that the number of turns is equal to the steady voltage drop on the relay divided by the product of the ampere-turns md the resistance of a single turn occupying the wound area. Hence for a given wound area and given ampere-turns, the voltage drop is proportional to the number of turns and if the latter are reduced then the difference voltage must be absorbed by a series resistance.

It can therefore be stated that the operate time of a relay with a given wound area and ampereturns, can be reduced by a series resistance and Thus, provided the wound area remains I in the same ratio as the resulting voltage drop on the relay to the battery voltage. It has already been shown that the required normal ampere-turns can be obtained without any increase in current density in the wire or wattage dissipation of the relay, for a given wound area, so that there is thus no limit to the degree of reduction of the operate time by reducing the number of turns and adding series resistance, apart from mechanical considerations, such as that the minimum number of turns is one and limitations set by the additional watts drawn from the battery and absorbed by the series resistance. A specific embodiment will now be described incorporating the above statements.

Referring to Fig. 1 a relay B of normal type is fully wound so as to produce the ampereturns normally required, with a voltage drop whose ratio to the battery voltage Eb is the same a the required ratio of reduction in the operate time (i. e., the time constant of the relay alone as compared with the time constant of the relay with the series resistance). A resistance Rs is connected in series with relay B, the battery Eb and the operating contact Al, and absorbs the diiference voltage. Connected across the relay is a shunt circuit comprising a further resistance R7 and a non-linear device MR, the latter of which only becomes effective in completing the shunt circuit when the operating contact-opens. The non-linear device MR may be replaced by a further contact A2, as illustrated in Fig, 1A, which closes when the operating contact AI opens and vice versa, and is conveniently an additional contact on the line relay A. In Fig. 1 rectifier MR is connected so as to be non-conducting to the operating voltage applied to the relay, but it automatically becomes conducting when the operating contact opens by virtue of the back E. M. F. generated by the decay of flux in therelay. In both cases the time constant of the decay may be controlled, up to the maximum permitted by the relay itself, by alteration of the shunt resistance ET. This permits of adjustment of the release lag without alteration of the mechanical adjustments. If a contact such as A2 is used instead of MR and both Al and A2 are contacts of the same relay, it is readily possible to employ a single make and break contact arrangement as will be readily apparent to those versed in the art.

It should be noted that although the above described method of reducing the operate time for given ampere-turns increases the watts drawn from the bat ery by comparison with a similarly wound relay giving the same ampere-turns operated directly from the battery, the watts required for the relay itself are less than that required by a normal slow release relay having a shortcircuited winding. This is because the watts reqtured are inversely proportional to the wound area for a given value of ampere-turns, and in the above method the relay is fully wound as compared to the present partially wound relay with an auxiliary short-circuited winding.

It is well-known that slow release relays of the type having a short-circuited winding or slug sometimes fail to remain operatedor fail to 0perate when the operating contact has an impulsing role, as it does in automatic telephone systems. Both types of failure are due to the same cause, in that the buildup flux time constant is at least as great as the decay flux time constant release; indeed it can be shown that the buildup time constant is the sum of the time constants of the operating and short-circuited coils, whereas the decay, on disconnection of the battery, has a time constant due only to the short-circuited winding. Thus, after a certain number of impulses a steady state of successive increments and decrements of flux is set up, whose mean value is proportional to the mean Value of the input impulses, i. e., the make percentage of the operating contact. It follows that release of such a relay will be governed rather by the minimum make percentage of the contact than by the break time period. Similarly failure to operate can occur if the make percentage is too small, or at least failure to operate over the first few impulses. Both types of failure can be avoided from the point of view of the system as a whole, if the steady state flux on operation of the contact is attained in a time less than the minimum necessary for operating (or release) of the remaining relays or magnets in the system. It is then only necessary that the release lag of the relay shall exceed the longest permissible break period, during impulsing.

Although adequate reduction in the time constant of the relay circuit by the method given above will, if carried far enough, enable the steady state flux to be substantially attained upon closure of the relay circuit in the desired minimum time, it will be seen that if the word substantially is defined as 95%, then the time constant of the circuit must be not greater than one third of this desired minimum time, (since the current in a relay attains 95% of the steady state in a time equal to three times the time constant of the relay circuit). though attainment of the steady state to this order is very desirable, this process would be very uneconomical in battery drain. If however the steady state could be attained substantially in a period equal to the circuit time constant, then reduction of the latter as described above would be both beneflcial and economical. It can be shown that the connection of a condenser Cs as shown in Fig. 2 across the series resistance Rs has this effect and this addition to the embodiment before described forms the preferred embodiment of the inventlon.

It can be shown theoretically that the time to attain the steady state value (i. e. independently of whether the subsequent relay current waveform is oscillatory about the steady state value, or tends thereto without oscillation), is equal to the time constant of the relay inductance and the series resistance, if the time constant of the condenser with the series resistance is made equal to that of the relay alone. In general the time required to attain the steady state value tends towards the time constant of the relay circuit without the condenser, as the value of the latter, and the ratio between the series resistance and the relay resistance, increases. In order therefore to attain the steady state value of relay current within a time substantially equal to the time constant of the relay circuit without the condenser, it is necessary that the latter shall be large and the ratio between the series resistance and relay resistance shall be large.

In practice it is found that if the ratio between the series resistance and the relay resistance is not less than 25:1, and the time constant of the condenser shunting the series resistance with the latter is of the same order as the relay alone, then the desired effect is obtained. This value of condenser/series resistance time constant also gives a reasonable amount of overshoot in the current waveform, which has been found useful in ensuring that the release flux/time waveform is maintained even when the circuit closure time approaches the design minimum. A typical relay circuit design using a well-known type of telephone relay normally operated on 50 volts, consists of fully winding the relay to a resistance of ohms, whilst the series resistance has a value of 400 ohms and the condenser shunting the latter a value of 40 microfarads. The minimum time to ensure attainment of the steady state flux with this arrangement, is less than 10 milliseconds, which is a satisfactorily low value for automatic telephone systems and will ensure that breakdown is not now due to this relay on impulsing. The release circuit is of course as described iormerly, for instance, a rectifier connected across the relay in such a manner as to be operative only to the back E. M. on release, and of course the release lags obtainable in these circumstances are normal since the relay is fully wound.

Provided the condenser is large enough to give a time constant with the series resistance which is much greater than that of the relay circuit without the condenser (i. e., the reduced time constant due to the addition of the series resistance), theory shows that the precise value of the con denser is immaterial, and this is borne out in practice. Using the design quoted above, it was found permissible to alter the value of the condenser between 20 and 100 microfarads without materially altering the time to attain the steady state current in the relay. The overshoot of current of course, was greatly altered, being negligible for the lowest value and increasing for higher values. The circuit of this embodiment of the invention is therefore amenable to the use of large commercial tolerances on the condenser value, so that an electrolytic type can be employed, whilst the possibility that the value may be sometimes large enough to produce considerable overshoot is of no disadvantage in automatic telephone systems, since this only means that the release lag becomes larger as the operating pulses approach the specified minimum time.

The value of the condenser should not, of course, be so large that the voltage on it due to a previous operation of the contact has not appreciably fallen, when the next operation ensues. For instance, if the condenser were of infinite value it follows that a subsequent operation would only build up the relay current at a rate dependent on the relay, time constant. However this can readily be avoided by making the time constant of the condenser with the series resistance substantially the same as that of the relay itself. when the condenser and relay currents will decay at the same rates and from the same values on the opening of the operating contact. As was shown before this is also a suitable condition for obtaining the steady state value upon closure of the circuit in the minimum time period.

In applying the invention to present equipment it will be seen that the relay circuit may replace any present slow release type which has only one operating winding, since by using a rectifier to short-circuit the relay during release, only one operating contact is required. In the case where it is required to hold the relay by a further contact or contacts, without also holding or operating other relays also connected to the original operating contact, this can be arranged by inserting an additional rectifier in the lead from the latter to the present relay circuit and connecting the further holding contact directly to the relay circuit. If the rectifier be connected so as to pass current from the normal operating contact to the present relay circuit, and other relays energised also from the operating contact be directly connected thereto, then when the latter contact is opened no current can flow from the holding contact for the present relay circuit to these other relays, because to do so it would have to fiow through the rectifier in the reverse direction. Such an arrangement is illustrated in Fig. 3 which shows a preferred embodiment with the addition of a series rectifier MR2, which permits of the holding of relay B by an additional contact such as H, whilst not interfering with the separate operation of other relays such as X from the main energising contact Al. Rectifier MR2 prevents the flow of current from earth through contacts H to relay X thus causing relay X to be deenergized when contact Al opens, so that X may be a repeating impulsing relay operated by A.

In addition to allowing holding of relay B, rectifier MR2 also permits the pie-operation of relay B by a contact such as H, without operating X. This is illustrated in Fig. 4, which shOWs the application to a C relay in the well-known automatic telephone selector circuit. The figure also shows the use of the invention in connection with the normal slow release B relay, and other relevant parts of the selector circuit.

In Fig. 4, relay A is the line relay which is controlled by a calling device at a subscrlbers station over lines LI and L2 in the usual manner. Connected to the make contact of relay A is the B relay circuit according to the invention as disclosed in Fig. 2. Connected to the break contact of relay A in series with contacts B! and Cl is the C relay circuit and stepping magnets in accordance with th invention as disclosed in Fig. 3. A pre-operating circuit for relay C is provided in series with the vertical oiT normal contacts VON and rotary off normal contacts RON of the selector switch in the same manner as with contacts H in Fig. 3 although in Fig. 4 the preoperating circuit is controlled by contacts on relay A rather than by a separate relay. The selector switch is seized by the closure of lines LI and L2 which causes the operation of relay A. Relay A operates, closes a circuit to relay B, and closes a circuit to relay C through the VON contacts. Relays B and C operate and prepare a circuit to the vertical magnet at contacts Bi and Cl. Rectifier MR2 is poled so as to prevent the operation of the vertical magnet V at this time. Operation of the calling device at the subscribers station interrupts the circuit to relay A a number of times corresponding to the first digit of the called subscribers number. Each time that relay A restores it opens the circuit to relay B at its make contact, and closes a circuit to the vertical magnet V and to relay C through rectifier MR2 at its break contact. The first operation of the vertical magnet causes the operation of the VON contacts which remain operated for the duration ot the call. Each time that relay A operates during the series of impulses it closes the circuit to relay B at its make contact and opens the circuit to the vertical magnet and to relay C at its break contact. Relays B and C both remain operated during th series of impulses due to the slugging effect of rectifiers MR and MRI. At the end of the first series of impulses relay C restores and closes a circuit to relay E from ground through operated contacts A3. operated contacts VON, contacts C2, and contacts E3 to relay E. Relay E operates, closes its locking circuit to contacts B2 and opens its operating circuit at contacts E3, transfers the pulsing circuit from the vertical magnet V to the rotary magnet R at contacts El, and closes a circuit to relay C at contacts E2 from ground at operated contacts A3 through the rotary oii normal contacts RON. This latter circuit does not affect the rotary magnet R due to the uni-directional conductivity of rectifier MR2. Relay C again operates and prepares a circuit to the rotary magnet R at contacts CI. The second series of impulses causes the rotary magnet R to be operated one step for each impulse in the same manner as for the vertical magnet during the first series of impulses. The first operation of the rotary magnet R causes the operation of the rotary oii normal springs RON which remain operated for the duration of the call. Relay C restores at the end of the second series of impulses and opens its holding circuit at contacts Cl. Since relay C cannot be re-operated the opening of contact Cl prevents any further operation of the stepping magnets. When the circuit to relay A is opened at the end of the call it restores and opens the circuit to relay B which restores and opens the locking circuit to relay E. The restoration of relay B also closes a circuit to a release magnet (not shown) which allows the selector switch to restore to normal.

The circuit also possesses the advantage that the C relay and the stepping magnets operate in independent circuits, which makes for greater efiiciency. The circuit of the invention is also independent of the back E. M. F.s of the other relays or magnets connected in parallel, since in the case of Fig. 4, MR2 takes the back E. M. F. of the magnets in its non-conducting sense, whilst the back E. M. F. of the slow release relay is absorbed within itself owing to the conducting shunt rectifier and produces the slow decay time constant.

The invention may, of course, be applied to equipment other than automatic telephone sys tems, and where the release lag of a relay is desired to be independent of the operating time down to a specified minimum value which is much less than the desired value of release lag.

With regard to the design and use of the rectifier which automatically short circuits the relay winding when the contact opens. it might be thought that some difficulty would result from the reduction of the resistance of the relay. This is not so. and in fact the ratio of the forward resistance of the rectifier to that of the relay can be kept constant at a value of about 10%, so that no material reduction of the maximum decay time constant is suffered. Since the watts in the relay constant for given ampere-tums, it follows that as the voltage on the relay is reduced by the addition of series resistance, the current is correspondingly increased. Now the initial value of the decay current equals the steady state value, and hence the cross-sectional area of the rectifier must be increased with the current. Hence as the voltage on the relay falls owing to the lower relay resistance, the number of elements of the rectifier required to withstand this voltage also fal s, and the area of the element increases with the increase in relay current, so that the forward resistance of the rectifier falls at the same rate as the relay resistance, and their ratio is constant.

I claim:

1. In combination, a relay having an energizing winding, a rectifier, a source of direct current, a first circuit path including said winding, said rectifier, and said source connected in series a second circuit path including said winding and said source connected in series but excluding said rectifier, means for completing said second circuit path to operate said relay, means for intermittently completing said first circuit path to maintain said relay operated, and electromagnetic means bridging said rectifier and said relay winding operated in response to the completion of said first circuit path to disable said second circuit path, said rectifier preventing operation of said electromagnetic means in response to the completion of said second circuit path.

2. A combination as claimed in claim 1 including a second relay operated in series with said first means and contacts on said first relay in response to the restoration thereof upon the termination of the intermittent completion of said first circuit path for again completing said second circuit path.

3. In a switching system, an impulsing relay, a second relay controlled by make contacts on said impulsing relay, a rectifier connected in parallel with said second relay to retard its release, a third relay, a second rectifier connected in parallel with said third relay to retard its release, a third rectifier, a circuit path for operating said third relay from said make contacts prior to impulsing, a second circuit path for holding said third relay during impulsing comprising break contacts on said impulsing relay, make contacts on said second relay, make contacts on said third relay, and said third rectifier, a stepping magnet connected to said third rectifier so as to be controlled by said break contacts, and means for increasing the ratio between the mean value during impulsing and the steady state value of the magnetic flux in said second and third relays comprising a parallel combination of a condenser and a resistor connected in series with the operating circuits of said second and third relays.

4. In a switching system; an impulsing relay; a first slow release relay controlled by make contacts on said impulsing relay; a second slow release relay; a circuit path for operating said second slow release relay from said make contacts prior to impulsing; a second circuit path for holding said second slow release relay during impulsing comprising break contacts on said impulsing relay, make contacts on said first slow release relay, make contacts on said second slow release relay, and a rectifier; and a stepping magnet connected to said rectifier so as to be controlled by said break contacts.

5. In combination, a relay having an energizing winding, a rectifier, a source of direct current, a first circuit path including said winding, said rectifier, and said source connected in series, a second circuit path including said winding and said source connected in series but excluding said rectifier, means for completing said second circuit path to operate said relay, means for completing said first circuit path to maintain said relay operated, and electromagnetic means bridging said rectifier and said relay Winding operated in response to the completion of said first circuit path for disabling said second circuit path, said rectifier preventing the completion of said second circuit path from operating said last means.

BERIRAM MORTON HADFIELD.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 1,742,367 Nettleton et al Jan. '7, 1930 2,182,637 Marbury Dec. 5, 1939 2,279,849 Van C. Warrington Apr. 14, 1942 2,299,941 Townsend Oct. 27, 1942 2,001,494 Jones May 14, 1935 2,128,063 Peters July 5, 1938 1,693,124 Stehlik Nov. 27, 1928 1,758,255 Hudd May 13, 1930 FOREIGN PATENTS Number Country Date 812,133 France Jan. 27, 1937

Images