THE COLLIERY GUARDIAN AND JOURNAL OF THE COAL AND IRON TRADES. Vol. CX. FRIDAY, AUGUST 13, 1915. No. 2850. A New Battery Signalling Bell.* By W. M. THORNTON, D.Sc.; D.Eng.f The total energy of the break-spark is |Lz2=-—g—■- It therefore varies as the square of the amperes and turns on the coils for a given magnetic circuit. For a Nature of the Spark, and Limits of Igniting Power.— The recent important Home Office report by Dr. R. V. Wheeler, on the risks of ignition of methane by battery signalling bells,! has proved beyond reasonable doubt that the spark in the bell or on the wires is a danger that must be dealt with. Before, however, this can be done effectually, it is necessary to know the nature of the spark, and the limits to its igniting power. Dr. Wheeler has given in his report several ways of improv- ing the safety of bells,’and much useful information for the guidance of bell manufacturers. The present paper is a statement of the chief points • which arise in the working of such bells, and affect sparking; and an arrangement is described by which danger from the spark is prevented. In the first place, the spark at the trembler contact is not continuous. If viewed in a revolving mirror, it with the period of the bell. Measurements of curves taken show that it is sometimes as low as a fiftieth of the period, so that in a bell ringing twenty times a second the time of high voltage is a thousandth of a second. So short a duration is less favourable to ignition than an ordinary non-inductive break-spark, and this may account to some extent for the rarity of accidents in coal mines which might have arisen from bells. The power w of the bell-spark—that is, the product Lz^ of volts and amperes in it—is given in fig. 5, and is seen to be even less uniform than the voltage. As shown in fig. 4, the ease of ignition of a given mixture by break-sparks, as measured by the product Li, is approximately constant. So long as w does not reach the critical value, the bell is safe. In order to keep w given number of ampere turns it varies inversely as R = f- Ap, where, since the air-gap has much greater magnetic resistance than the core of the coils, I is the double length of the air-gap, A the mean area of the magnetic field in the gap, and/z = 1, when Z and A are in centimetre measure. The product Li is the number of lines of force N linked with the coil windings. N only affects the spark when changing, for ___d N~________-r- dz ' dt dt Lb= R and is, therefore, proportional to the current and to the square of the turns. The voltage __4 7T T2 di Bg~ R-’ dt The rate of break depends so much on the mechanical is seen to be regularly intermittent; and if the current is observed by an oscillograph, it is found to have the general form shown in fig. 1. In the case of an electric circuit of resistance r and inductance L,§ the curve OA has the form E z = ~ r where E is the battery voltage, and e = 2’718, the base of the natural logarithms. ' E/r is the final steady value of the current i if the circuit is closed for a long time. In bells this curve is modified by the movement of the armature and by eddy currents in the solid parts of the magnetic circuit of the windings, and may become a straight line, as in fig. 2. The voltage e8 which causes the spark at the trembler contact is produced by the rush of magnetism out of the coils when the current is broken. It is, in fact, di L , where L is the self-induction and di/dt the rate of dt break of the current. When L is in henries and di/dt in amperes per second, Ldi/dt is the voltage on the spark. At the moment of break the battery voltage is all absorbed in the resistance, and does not affect the Fid 'I - VOLTAGE ACROSS Sl’ARK-GAl’ WHEN RESISTANCE IS CO.NNECIED, AS Figs. 7 ok 8. £ Oh - h-----/w/tf Fu - /Wtf MAKE Voltage across Spark < BREAK BREAK * BREAK & BATTffVr VOLT Abe MAKE T 1R EM BLER-VON TAC'J TlHEt Fig 6 O MAKE MAKE- a * BREAK BATTERY VOLT Abt ONLY MAKE-- time t Fig 2—Form of Current-li b\ i- <><• aMonai.ii F<>un Fig 1 Cur’e of Current in a Batteha bi ll Circuit in Full Kinging. X I _ BREAK J jl Fig. 5.—Volt-amperes of Trembler Spark. iunt Resistance wross Spark-gap diminishes Spark at a, but INCREASES THAT ON LlNE AT b "4 i BREAK * initial value of es. The gap voltage as observed is that shown in fig. 3; after the spark has ceased, the voltage across the spark gap is that of the battery alone. In good bells the time of break is found to be about half that of make, and the maximum voltage v is about 10 times that of the battery e. The spark at the trembler contact is a series of ordinary’ break-sparks of short duration, in which both current and voltage are greatest at the instant of break. Now, the igniting power of an electric break-spark depends on the voltage with which it is associated. It has been shown that the product of these is roughly constant,|| as in fig. 4. When the voltage is high, a small current becomes dangerous, and the degree of danger increases with 'the inductance in the circuit. With single continuous-current break-sparks on a 20-volt CIRCUIT VOLTAGE b Fig 7 - Resistance across Magnet-coils suppresses Sparks both «t Fig 4.—Charge of least Igniting Current with Voltage in a Non-inductive Continuous-current Circuit Fig 8 - -Resistance connected at c just before Break at a battery circuit, 0’5 henry causes ignition of the most inflammable mixtures of illuminating gas with a current of 0-5 ampere.U Another factor of great importance is the duration of the spark. It is clear from the curves of figs. 1, 2, and 3 that the duration of the high voltage is small compared * A paper read before the North of England Institute of Mining and Mechanical Engineers. f Professor of Electrical Engineering in Armstrong College, Newcastle-upon-Tyne. J Report on Battery-bell Signalling Systems as Regards the , Danger of Ignition of Firedamp Air Mixtures by the Break- flash at the Signal Wires, by R. V. Wheeler, D.Sc., 1915. § The inductance of an electric circuit is defined as the number of lines of force which are linked with the circuit when unit current is passing through the wires. It is measured in henries. l'| Sec “ The Ignition of Coal Gas and Methane by Momen- tary Electric Arcs,” bv Prof. W. M. Thornton, Trans. Inst.M.E., 1912, vol. xliv., p. 145, figs. 2 and 3. 11 Ibid., fig. 5. low, it is advisable to make -the product Lj as small as possible, for the rate of break depends to a great extent on the. mechanical design and setting of the bell. Measurements of L have been given in the Home Office Report, and have since been found to reach the high value of 1 henry. There is an upper limit to the current, depending on the kind of battery used. It may be taken that signalling currents range on an average between 0-1 and 0-5 ampere, and the value of Li from 0-05 to 0’1. There is no limit to the number of turns on.the bell, that is, to the self-induction L. Expressed in terms of dimensions,L=4ttT2/R,where T is the number of turns on the . windings and R the reluctance of the magnetic circuit, which is equal to the average length of the magnetic lines of force divided by the sectional area of the core and by its magnetic permeability. The air- gap has a greater reluctance than the iron of the magnetic circuit. adjustments that it is the least definite of all the important factors in sparking. For a fixed rate of break it is nearly proportional to the current broken. Writing di ■ . 4ttA:T2. the result is es = —--------1, following the same (Lt Jtt law as Lt. The Influence ■ of Resistance.—How does sparking depend upon the working resistance in the circuit? The greater the resistance r, the greater is the drop of voltage in the wires for a given current, and the less of it there is for the gap while the current lasts. The equation of the current in the bell at break may be written i = — e , where r. is the resistance after break, r including that of the spark. From this = -1, and JS v the resistance of the spark rs, which is - r = r (e“ O’ The resistance of the spark is not independent of the