Journal of the Society of Motion Picture Engineers (1930-1949)

M2 ± a few volts. In either initial or subsequent balancing, it will be found that both Ml and M3 vary in the same direction with each other and with the Variac. However, Ml, across phase 1 and 2, varies more rapidly than M3, and M2 does not change except with input voltage. In operation, the converter may be located at the camera and the "line" switch Dl used as the camera operating switch, or the unit may be positioned at the recorder with any individual or common switching desired. In mild weather where temperature and camera load do not vary greatly, the initial load adjustment of the Variac might well remain the same for several days shooting. In case of cold weather, however, some cameras will warm up enough in a long take to change load by a factor of 2 or 3. Adjustment of the Variac to meet this condition may be made during a take without disturbance to either picture or sound. It should be noted, however, that precise adjustment is not essential to operation and under average conditions the unit can be forgotten unless the camera motor becomes noisy or lacks power. Where recorder and camera are both operated from the converter, either motor may be dropped off at will. The phase balance will be materially upset and the remaining motor will be noisy but the recorder is usually too far from the microphone to cause trouble, and if the recorder is cut off, the camera noise does not matter. Performance It is obvious that the voltmeters used as indicators of phase balance actually show only voltage across the three phases and indicate phase relation indirectly if at all. Thus, this method of indication may well be questioned. Since the voltages indicated are each a resultant of two voltages to the mid-point of the load, which in turn are the resultant of a vectorial addition of both voltage and phase relation, the theoretical reasons why such indications are of value become involved and tedious and will be omitted in favor of measured results. A true picture of the voltage conditions in the load can be obtained by measuring the voltage from each phase lead to a mid-tap on a star connected load. In addition, a method was devised which indicated voltage phase relations across the same points, and across Tl and T2 primaries, to an accuracy of db£°. The currents in each leg of the load are not in phase with the voltage but bear the same relation to each other as the voltages since the load is electrically symmetrical. With the above arrangement, it was possible to obtain an accurate picture of phase-voltage relations under varying load conditions. A motor fairly typical of three-phase synchronous camera motors was operated through the converter from a 115-v single-phase source and adjusted for capacitance balance at no-load in the manner previously outlined. The accuracy of balance in each case was probably of the order of ±2 v; about what would be expected in normal use. Figure 3 shows the phase-voltage relations without correction for load change. In Fig. 3A the primary of T2 is compared to the primary of Tl which latter is also the line input and does not change with load. The solid line shows T2 at 140 v displaced 89° from Tl when the motor ran no-load. When the load was increased to just under pull-out, the voltage dropped to 112 v across T2 and the phase relation to Tl became 75° as shown by the dotted line. In Fig. 3B of the same figure is shown the resultant conditions existing in each phase winding of the motor. At no-load as again shown by the solid lines, the voltages to the mid-point were between 131 and 134 v with phases 121, 121 and 118° apart. With 100-w load on the motor, phase 1 has A. L. Holcomb: Three-Phase From Single-Phase 37