"in over-compounded machines the regulation is the ratio of the maximum difference in voltage from a straight line connecting the no-load and rated-load values of terminal voltage as function of the load current, to the rated-load terminal voltage." That is, the less the curvature of the curves, the better is the regulation.

Field Compounding. The number of series-field turns to be used is determined by a test yielding the data in Fig. 89. While the armature current is varied, the terminal e. m. f. is maintained constant at 575 volts by varying the shuntfield current. The curve shows that it requires in this case an increase from 12,300 to 14900 or 2600 ampere-turns to

FIG. 89. Field-compounding curve, d. c. generator (G. E. Co.), showing field ampere-turns necessary to maintain 575 volts e. m. f. with varying load.

maintain e. m. f. against the armature resistance drop and the magnetic effect of the armature upon the field. The curve dips on account of the field saturation, more field m. m. f. being required proportionately with large than with small

armature current.

Saturation. The saturation curve shown in Fig. 90 is plotted between field ampere-turns and armature e. m. f., which is proportional to the flux. The curve shows that the iron in the magnetic circuit begins to saturate at about 350 volts and the steepness of the curve decreases rapidly above 500 volts. This accounts for the bending of the compounding

curves already referred to. Fig. 91 exhibits the corresponding core loss (in hysteresis and eddy currents) for various

2000 4000 GOOO 8000 10000 12000 14000 16000 18000 20000 22000 Ampere, Turns

FIG. 90. Saturation curve of d. c. generator (G. E. Co.).

The efficiency curve, Fig. 92, gives the relation between per cent efficiency and per cent load and shows that above 50 per cent load the efficiency is practically constant at 95.5

Watts Loss

per cent. On light loads the efficiency is low because field loss, armature-core loss and friction are proportionately large, amounting at 5 per cent load, or 85 amperes, to 30 per cent of the input or about 21.5 k. w. This indicates the import

FIG. 91. Curves between core loss and e. m. f. (proportional to flux) in d. c. generator (G. E. Co.).

ance of operating these generators, and in fact all electrical machines, at as large a percentage of rated load as possible.

Handling Direct Current Generators.

The shunt field circuits of all d. c. generators are provided with resistances or "rheostats" for the purpose of permitting the variation of the current and thus adjusting the e. m. f. When shunt machines are operated in parallel the load is distributed by means of these rheostats or by varying the speed of the driving prime movers. Compound-wound generators

cannot be operated in parallel directly for the following reason. If one machine for any reason takes more than its share of the load its e. m. f. will automatically increase. It will take still more load, and in fact will almost immediately begin to operate the other generators as motors. To prevent

FIG. 92. Efficiency curves, d. c. generator (G. E. Co.).

this the series coils are all connected in parallel by a heavy bus bar known as an "equalizer" bar or bus. This insures the uniform distribution of the line current among the series coils, and hence if an increase of line current occurs all of the machines will increase their e. m. f's. alike and each will take its proper share of the increased load.

Series Transformers - Constant Current Transformers.

Constant Current Transformer Constant Current Regulator.

Transformer Characteristics.

Constant Potential Transformer.

Regulation Losses Efficiency Heating

Series Transformer.

Constant Current Transformer.

Transformer Installation

Transformers on Polyphase circuits.

Transformer Construction.

Principles Underlying Transformer Performance.

FARADAY'S ring, illustrated in Fig. 16, contains the principle of the transformer. Faraday found that when he connected one circuit to a battery a momentary current was produced in the other. A similar momentary secondary current was produced when the primary circuit was disconnected. If alternating current had been available at the time, Faraday would undoubtedly have applied his coil to the transformation of e. m. f., but it was not until fifty years later that this was actually done. In the intervening period all development was along the line of continuous current. Faraday's ring would