DISTRIBUTION AT CONSTANT CURRENT.

155

connection between the speed of a motor and the current which flows through its armature. There is a direct connection between speed and electro-motive force, and, therefore, self-regulation is possible without the use of any external appliance in the shape of a mechanical governor or other apparatus which controls the power. But where the current is constant, some kind of external governor is necessary. This follows also immediately from M. Marcel-Deprez's experiments cited in Chapter III., page 90. We have seen that the speed was totally independent of the current, the latter remaining throughout the range of each experiment practically constant, whereas the speed was in some cases increased fivefold, by simply increasing the electro-motive force of the source. When a number of motors are coupled in series, as would be the case in a general system of distribution, the difficulties are much increased. To test this matter experimentally the author has placed three precisely similar motors (series-wound) in series into the same circuit. The current was supplied by a dynamo, and the three motors were loaded by brakes to as near as may be the same amount. It was then found quite impossible to keep all three motors going for any length of time at the same speed. The least irregularity in the current, or the least variation in the friction of the brakes, would cause first one and then the other motor to come to rest, whilst the speed of the remaining motor increased to a dangerous extent.

Professors Ayrton and Perry have in the paper above mentioned proposed to make motors self-regulating if worked by a constant current in the following way: The field magnets, Fig. 65, are wound differentially with a fine wire coil, which is a shunt to the armature only,

and a thick wire coil which is in series with the armatur and main current. The armature and shunt coil consti tute a shunt motor, the armature and main coil a brak generator which is intended to absorb any surplus powe if the load is thrown off. As far as the author is awar the system has not been tried in actual practice, and there are theoretical reasons for expecting that it would not work. From equation 7) it will be evident that the field must be strongest when the load is greatest. suppose that the differential winding could be so propor

Fig. 65.

Now

N

tioned that for a given load the field is exactly of the right strength to produce the normal speed. Now let a very slight additional load be thrown on. The immediate effect will be to slightly reduce the speed, and in consequence of the reduction in speed the magnetizing current in the fine wire coil will also be reduced. The field will thus be slightly weakened. This will further reduce the speed and again weaken the field, and so on, until the armature comes to rest. At that moment the magnetizing influence of the main coils, which is in the opposite direction to that of the shunt coils, will alone exist, and the

SELF-REGULATING CONSTANT CURRENT MOTOR. 157

field magnet instead of presenting a N pole to the armature, as shown in the illustration, will present a S pole to it. The tendency must therefore be to reverse the motion, and thus the slight addition of load has not only brought the armature to rest, but actually caused a tendency to run backwards. Whether it will run backwards depends on the relative magnetizing power of the main and shunt coils.

An arrangement devised by the author, and which

Fig. 66.

a

seems to promise somewhat better to fulfil the condition of constant speed, is shown in Fig. 66. A is the armature of a series-wound motor mounted upon a spindle, to which is also attached the armature a, of a small serieswound dynamo which has no other work to do but to supply current for demagnetizing the field magnets of the motor. The main current magnetizes them in the direction, say, from B to C, the auxiliary current from the dynamo acts in the direction f to g, and tends to demagnetize them. bc is the field magnet coil of the dynamo. Now for each dynamo working on a closed

circuit of constant resistance, as in the present case, there exists a critical speed at which it will begin to give a current of some strength. Below that speed it gives hardly any current, and above that speed it gives almost at once the full current. The motor should be so geared as to run at the critical speed of the little auxiliary dynamo. If now an additional load be thrown on, the immediate result will be to reduce the speed of the motor, thereby causing the armature of the dynamo to run below its critical speed. The dynamo will thus partly or entirely lose its current and the demagnetizing influence which previously has kept the field below its full strength, will to a greater or lesser degree be withdrawn. The strength of the field will thus be increased, and an additional magnetic pull will be brought to act on the armature, by which it can overcome the increased load. In case the load be entirely thrown off, the motor will have a tendency to race, but this tendency will be immediately checked by the auxiliary dynamo, the current from which increases considerably with a very slight increase of speed. Its demagnetizing influence is thus enormously increased, and the field of the motor is weakened to such an extent that there is just power enough left to drive the dynamo but no more. To make this arrangement successful it is necessary that the field magnets of the auxiliary dynamo be made of very soft iron, so as not to retain any considerable amount of permanent magnetism, which would alter the critical point as between an increasing and decreasing speed. The more sensitive and unstable the dynamo can be made, the better. For this reason it is also necessary to place the two armatures a considerable distance apart on the same spindle, so that the field magnets of the motor may

DISTRIBUTION AT HIGH PRESSURE.

159

not induce magnetism in the field magnets of the dynamo, and thus disturb the critical point. In practice it would probably be found necessary to place a bearing between the two armatures, and that could easily be so shaped as to act as a screen between motor and dynamo.

The importance of having self-regulating motors of this description cannot be over-estimated. If we would distribute electrical energy over a considerable area it is absolutely necessary to work at high pressure, otherwise the cost of copper in our mains becomes prohibitive. On the other hand, we cannot, under the provisions of the Electric Lighting Act, 1882, employ in our houses a higher electro-motive force than 200 volts, and even that is seldom employed, since for incandescent lighting in parallel connection we are limited by the electro-motive force of the lamps, which as yet has not much exceeded 100 volts. Lamps of greater voltage can be made, but are as a rule too delicate for ordinary use. In a system of general supply we would thus be practically compelled to distribute electricity at the exceedingly low potential of 100 volts. As Professor G. Forbes, in his Cantor Lecture at the Society of Arts, has pointed out, the transmission of electric energy at so low a potential would require the use under our streets of copper conductors half an inch thick and many yards wide. This is quite out of the question, and some means of economizing copper must be discovered before we can attempt to transmit electric energy to any distance. Professor Forbes, in the course of lectures above mentioned, has shown several methods. by which a reduction in the cost of mains can be attained. Of this question more is said in Chapter VII. For the present it will suffice to point out that the system of transmitting energy by so-called secondary generators is the