Electical Engineering at the Inventions Exhibition

Messrs. Elwell-Parker Limited, of Wolverhampton, show a large display of their dynamos, ranging from machines for an output of 1,500 watts up to 50,000 watts. With exception of the large machine, which has four poles all the others are of the Gramme type with two poles, and are now being manufactured in almost every detail according to the original patents taken out by Messrs. Elwell and Parker in 1883.

It is not often that we find the whole process of manufacture of any machine covered by the original specification. On the contrary, in most cases the first patent is merely the starting point for the practical development of the invention, and manufacturers find it necessary to continue patenting details until the final shape taken by the invention is, on the one hand, shorn of a good deal that was considered most important at first, and, on the other hand, contains some salient features which were not even hinted at in the final patent.

Messrs. Elwell-Parker’s specification, 1883, No. 770 is an exception to this rule, and possesses, therefore, considerable interest. It would be perfectly correct if we were to describe their dynamos by simply quoting parts of their original specification; but as this does not contain any numerical data, we prefer to give the description in our own words and to add such figures and practical information as, by the courtesy of the firm, we are able to lay before our readers.

It might be said that the Elwell-Parker dynamo was invented to order. The firm commenced their electrical work by the manufacture of an improved form of Planté accumulator; and as in the “forming” of the cells, dynamos were required, they used at first, what machines they could obtain in the market. This was two years ago, and our electrical readers who have had experience in similar work will not be surprised to hear that none of the machines obtainable at that period could satisfactorily be used for charging accumulators. Sparking at the brushes, worn out commutators, waste of power, reversal of polarity, burned up armatures, and other like evils were the order of the day. All this has been changed since then, and perfectly satisfactory machines can now be obtained from dozens of makers; but then it was not so, and Messrs. Elwell-Parker resolved to satisfy their wants by making their own machines.

The first machine was finished in October, 1883, and has been used continually since that time in the company’s works, partly for lighting, and partly for charging. Its armature can be described as a compromise between the ordinary Gramme ring and the Bűrgin hexagon. It shares with the former the circular shape and the winding which completely covers the external surface of the core, and at the same time it has the mechanical advantage of the latter, the way the core is supported on metal arms. In the Bűrgin machine the arms are forced into the ring after the same is wound, and they take a bearing in the corners of the hexagon.

Since there can only be six distinct coils on each ring, it is necessary to employ a number of these rings so as to get an even current and to avoid sparking. This limits the width of each ring to about 2 to 3 inches. Now in the Elwell-Parker armature any desired number of coils may be wound on to the core to ensure an even current without the necessity of employing several rings on the same spindle, and the width of the ring is not limited by any other consideration but that of obtaining the desired output.

Since the forcing of the arms into a circular ring would necessarily deform the ring, Messrs. Elwell-Parker have introduced the improvement of coiling the wire direct over the arms, and not into a dummy as was done by Bűrgin. They insert segments of wood between the arms, so as to preserve to the ring a true circular shape, although the wire be coiled with a considerable amount of tension. We illustrate in Figs. 1 and 2 the method adopted for coiling the core. The supporting arms a a are keyed to the spindle C, and after having been insulated with tape and fibre, the latter to prevent metallic contact between the core and the arms, the wood blocks A^I are inserted and held in place by discs DD, into which they are morticed. Three bolts b keep the whole frame together. To prevent the small lugs or projections e coming in metallic contact with the iron of the core, washers d d of vulcanised fibre are placed one at each end, and the space between them is now completely filled with iron wire coiled on as evenly as possible. In the process of coiling a bituminous compound is applied, which is intended to insulate to a certain extent the wires from each other.

In our article of last week we had occasion to speak of the value of that kind of insulation, and need not reopen the subject. The application of bitumen has, however, this important advantage, that it renders the core more like a solid mass of metal, which can be turned up true in the lathe. In fact, that is what Messrs. Elwell and Parker do. After a few layers of wire are coiled on, the surface is tested, and if found out of truth a cut is taken over it. We need hardly mention that the depth of a cut never equals the diameter of the wire, and usually is only a fraction of it. A few more layers are then coiled, and if necessary a cut is taken gain, and so on until the core is completed to the right radial depth. In this way a true cylindrical core is obtained.

After removing the blocks of wood and insulating the core, it is wound with double silk covered copper wire in the usual Gramme fashion, but only one layer of wire is used on the outside of the core; and this rule is not departed from even if the electro-motive force required be 1,000 volts. No such high tension machine is, however, shown at the Exhibition, the maximum electro-motive force reached by any of the exhibited machines being 300 volts. This is a 6,000 watt machine used in conjunction with B.T.K. accumulators for lighting the East Quadrant. It has already been remarked in a previous article that most of the modern dynamos contain only one layer of wire on the armature; the Crompton, Edison-Hopkinson, Mather and Platt, and several other machines are constructed in this way; but Messrs. Elwell-Parker have pushed the principal to the utmost limit by making this layer exceptionally thin; in other words by allowing a high density of current in their armature coils.

Most makers regard 3,000 amperes per square inch of conductor a sufficient strain, but in the dynamos under consideration the density is frequently as high as 6,000 amperes. Roughly speaking we may say that the thickness of the single layer in the Elwell-Parker machine is about one half of that found in other dynamos, and in this sense the firm have pushed the single layer principle further than other maker. They maintain that by making the core of small radial depth, so that even the inner coils do not completely fill two layers, the carrying capacity of the wire is considerably increased.

The interior of the core is not ventilated, but by a peculiar arrangement on the commutator a current of air is kept flowing into the cavity of the armature at the pulley end, and out between the wires which join the armature to the commutator. This serves to cool the inner wires whilst the outer turns are kept cool by being whirled through the surrounding air at considerable speed. How far this object is obtained we are unable to say; the point can only be settled by the actual experiment of working a machine with a full current, say, for fifty hours continuously. It should be remembered that doubling the density of current means quadrupling the heat generated, and that would require four times the cooling power. If Messrs. Elwell and Parker are able to do this we see no reason why other makers should not do the same by the application of a fan or other special device, and thus double the output of their machines. But we must return to the general description of the Elwell-Parker dynamo.

In Figs. 3 and 4 we illustrate this machine by a longitudinal section and an end view. The supporting arms or “spiders” are keyed to the spindle C by the hubs A A. Part only of the core is shown, and the armature coils are left out for clearness of illustration. The dark lines represent the insulation between the arms and core. No special description of the commutator is required, as its construction is clearly shown in our drawing. We must mention however, that the lugs are considerably extended, forming a large disc, to the front end of which is attached a fibre disc l, and this arrangement serves partly to prevent copper dust from being drawn into the armature, and partly acts as a fan by which a current of air is sent through the armature in the direction of the arrows.

The field magnets consist of flat wrought iron slabs K L bolted together in the shape of a rectangle, with cast iron pole pieces pinned on in the middle of the top and bottom slabs. A large number of Tins K K is used, these pins being driven tightly into holes bored out in the pole pieces so as to be in close contact with the metal. By this arrangement it is intended to obtain the magnetic conductivity of wrought iron whilst still retaining the cheaper material, viz., cast iron for the bulk of the pole pieces. In small machines the frame M M is made of Parker’s metal, but in larger machines it is made of cut iron, and packing pieces of non-magnetic metal are inserted between frame and field magnets.

It is of interest to note that the thickness of these pieces need not be very great; at any rate much less than would be required in an Edison machine of equal size. In the latter the pole pieces themselves have to be kept off the frame by non magnetic packing, and since the free magnetism is greatest at the poles the magnetic insulation must be correspondingly high. In the Elwell-Parker machine, however, the two portions of the field magnets which come nearest the frame both belong to the lower pole, and are therefore magnetically at a comparatively low difference of potential. Even if the packing were omitted the machine would still continue to work. The only effect would be a very slight drop of electromotive force due to the partial muffling of the two lower half coils by direct contact with the frame. The rest of the magnetic circuit would, however, not be altered. The arrangement of magnets adopted by Messrs. Elwell-Parker is certainly very compact but we think that it might be improved by keeping the vertical magnets a greater distance from the polar extensions N S. As is well known, magnetism is always densest at the corners, and a considerable number of lines of force must be wasted by leakage from those corners to the core of the magnet. Some slight improvement might also be effected by shaping the extensions as shown in dotted lines.

The first armature made by the firm is exhibited in the East Arcade, and is a very creditable piece of workmanship. Although it has been continuously at work from October, 1883, until the opening of the Exhibition, and has been used hard, it does not show any signs of this except a slight bulging of the external wires, which is doubtless due to the fact that they are held in place by friction only, and at some time or other, probably when giving an excess of current, have slightly shifted on the core. The core is 7 inches in diameter, 9inches long, and l inch thick. The effective area of iron is therefore: π/4 x 9 x 1 = 7.07 square inches. The field magnets are wrought iron slabs 8 inches by 1¼ inches, having a cross sectional area of 10 square inches, or about 30 percent in excess of that of the armature.

The latter is wound with 225 turns of 0.83 inch wire, total length about 130 yards. At a speed of 1,600 revolutions per minute the external electromotive force is 60 volts, being at the rate of 2.16 yards per volt. The field magnets are shunt wound, and contain 46 ohms of 0.058 wire. This corresponds to about 4,700 yards, and since the mean perimeter of magnet coils can be taken as about 0.83 yards, we find the total number of turns = 4,700 : 0.83 = 5,660.   The exciting power on one half of the machine is therefore 2830 x current through the shunt wire. The latter is 60 : 46 = 1.3 amperes and the exciting power 3,680 ampere turns. This is a remarkably low figure, and is probably due to the fact that the clearance between the armature and pole pieces has by careful workmanship been reduced to the very lowest possible limit.

From the figures given above we calculate the electrical efficiency of the machine as follows: resistance of the armature, 0.152 ohms; with an external current of 0.65 amperes, the current through the armature is 66.3, and the loss of pressure about 10 volts; internal energy, 70 x 66.3 = 4,640 watts; external energy, 60 x 65 = 3,900 watts; efficiency, 86 percent. The resistance of the armature is calculated at ordinary temperature, but with a density of 6,150 amperes, which corresponds to 65 amperes output, and it is probable that the temperature of the wire will rise considerably, increasing at the same time the resistance of the armature. This would of course slightly lower the efficiency. There are 6.8 lb. of copper on the armature, and 142 lb. on the magnets; total 148.8 lb. With a speed of 1,600 revolutions the output is therefore at the rate of 26.2 watts per pound of copper. With our standard speed of 1,000 revolutions the output would be at the rate of 16.3 watts per pound of copper.

A word of explanation is necessary as regards the number of yards of armature conductor per external volt. Some makers, and amongst them Messrs. Elwell-Parker argue that since at any moment both halves of the armature lying on either side of the diameter of commutation are coupled parallel, the electromotive force is produced on one half of the wire, and it is therefore right to take only one half of the total length when figuring out the number of yards per volt.

This argument is perfectly logical, and can be upheld on scientific grounds. But since an armature only half wound with wire is practically useless, we prefer to adhere to our plan of always counting the whole of the wire when the machine has two poles. With a four-pole machine the case is different. Here we are quite justified in counting only half the total wire, because it is practically possible, though not advisable, to work the machine by allowing only one-half the armature to be active. The current will thereby be diminished to half its normal strength, but the electromotive force will remain the same. Similarly in a six-pole machine, we would only count one third of the total length of wire coiled on the armature, and so on.

Messrs. Elwell-Parker exhibit a very interesting four pole machine, which we illustrate in the annexed engraving. The armature is constructed precisely as that of their first machine described above, but is of much larger dimensions. The core is 16 inches in diameter, 36 inches long, and 2 inches deep. It is wound with 186 turns of double 0.120 wire, the perimeter of each turn being 2.14 yards. This gives a total of 398 yards of double conductor. At a speed of 450 revolutions per minute, the external electromotive force is 110 volts, being at the rate of 1.81 yards per volt.

We give in Fig. 5 a transverse section through armature and magnets, omitting, however, all constructive details, as our sketch is only intended to show the grouping of the, pole pieces. To each side of the armature is fixed a complete horseshoe magnet, which is supported at the yoke in the cast iron framework, as will be seen from our perspective illustration, and at the poles by gunmetal bolts tapped into a central girder. The distance between the girder and any part of the magnet is at least twelve times the clearance between the surface of the armature core and the face of the pole piece. The exciting coils, of which we only show one, are 17 inches long, and each contain 2 cwt. of 0.058 wire having a resistance of 72 ohms. The four arms are coupled parallel. Each contains about 3,430 turns, and is traversed by an exciting current of 1.53 amperes. The exciting power in one of the magnets is therefore 2 x 3430 x 1.53= 10,500 ampere turns.

The resistance of the armature, cold, is 0.0278 ohms. We must mention that this figure is not from measurement, but is obtained by simply dividing the resistance of 398 yards of double 0.120 wire by l6. The effective area of iron in the armature is 56.4 square inches, whilst the cross sectional area of the magnets, which are slabs 4 inches by 32 inches is 128 square inches, or more than double the former figure. This is, however, perfectly correct and as it should be, for with the arrangement chosen each magnet must supply sufficient lines for filling twice the area of the armature core; and if the lines shall not be throttled in any part of the magnetic circuit, the area of the magnet cores must be at least twice that of the armature core.

The machine is classified as a 50 unit dynamo; electromotive force, 110 volts; current, 460 amperes; current in shunt coils, 6.12 amperes; loss of pressure in armature,13 volts; internal electrical energy, 466.12 x 123 = 57,200; external energy, 50,600; efficiency, 88.3 percent. The density of current is 5,150 amperes in the armature coils, and 580 amperes in the shunt coils. The weight of copper used is 104lb. in armature, 896lb. in field; total l,000lb. At 450 revolutions the output is therefore 50.6 watts per pound of copper; at our standard speed of 1,000 revolutions per minute it would be 112 watts per pound of copper.

If the coils in the field magnets be joined up in such way as to make both poles above the centre line N and both below it S, the machine will work like any ordinary two-polar dynamo. Two brushes only will in this case be employed. Messrs. Elwell-Parker have made that experiment, and have found that the electromotive force is thereby only increased by about 30 percent, whilst, of course, the current is halved. On purely theoretical grounds we would expect that the electromotive force should be doubled; but in practice a point of saturation, as regards the density of lines in the armature core, is soon reached, and that prevents the full advantage being obtained from the two-polar arrangement. To make our meaning, clear, let us assume for sake of argument that in the four-polar arrangement there are 112,000 lines emanating from each polar surface. We choose this figure simply by way of illustration, as it is a convenient multiple of the area of armature core.

Through each square inch of it there will, therefore, be 1,000 lines. In order that the electromotive force may be doubled by the two-polar arrangement, the total number of lines emanating from the four poles must remain the same, and to make this possible the density of lines would have to be 2,000 per square inch of core. This is probably more than the core can carry; or, in other words saturation sets in and prevents the full number of lines from being created.

Messrs. Elwell-Parker have supplied two dynamos like the one described to the Blackpool Electric Tramway Company, as well as seven motors for propelling the cars. A similar motor is exhibited in Mr. Holroyd Smith’s car on the South Promenade. We hope to give a full description of this system of electrical propulsion shortly, and for the present must content ourselves to lay before our readers a few data regarding the motor only. It is shunt wound, and intended to work with a potential of 200 volts. The current is supplied by an Elwell-Parker dynamo, driven by a high speed engine also made by that firm.

With a current of 30 amperes through the armature of the motor, it exerts a statical torque of 56lb. on a radius of 16 inches. We can, with sufficient approximation, assume that the dynamic torque is the same, the field magnets being shunt wound and the strength of field therefore independent of the current through the armature, and in that case the motor would give off about 8 horsepower at a speed of 600 revolutions per minute. The resistance of the armature is 0.75 ohms, that of the field magnets 200 ohms. The motor weighs 6 cwt. Its field magnets resemble somewhat those of a Siemens machine, which form has been adopted in order to be able to get the motor into the confined space under the floor of the car.

In all there are eleven dynamos and one motor exhibited by Messrs. Elwell-Parker; their list is as follows: one 50,000 watt machine, driven by Tower engine, and lighting the Club dining room; one 20,000 watt machine as a reserve to the above; one 12,000 watt machine for charging accumulators, lighting the Royal Pavilion; one 6,000 watt machine, charging B. T. K. accumulators, for lighting the East Quadrant; one 6,000 watt machine, driven by Willans' engine, and lighting the Subway; one 6,000 watt generator, driven by Elwell-Parker engine, and supplying current to electric tramway; one 3,000 watt machine, driven direct by three-cylinder Elwell-Parker engine, intended for train lighting; one 3,000 watt machine, driven by 3½ horsepower Otto gas engine, charging 26 cells at Mr. Taylor Smith's stand; one 3,000 watt machine, driven by Beachy gas engine, and lighting Messrs. Woodhouse and Rawson’s stand in East Arcade; one 1,500 watt machine, driven by Gramme motor; one 6 horsepower motor, driving electric tramcar.