| Readings on standard meter | Readings on meter tested | Constant |
| 150 | 149.2 | 150 ÷ 149.2 = 1.005 |
| 125 | 125.0 | 125 ÷ 125.0 = 1.000 |
| 100 | 98.9 | 100 ÷ 98.9 = 1.011 |
| 75 | 73.6 | 75 ÷ 73.6 = 1.019 |
| 50 | 50.0 | 50 ÷ 50.0 = 1.000 |
| 25 | 24.8 | 25 ÷ 24.8 = 1.008 |
| 6.043 | ||
Average constant for six readings, 6.043 ÷ 6 = 1.007. | ||
Fig. 2,906.—Transformer winding or ratio test. The object of this test is to check the ratio between the primary and the secondary windings. For this purpose a transformer of known ratio is used as a standard. Connect the transformer under test with a standard transformer as shown. Leave switch S2 open. With the single pole double throw switch in position S1B, the voltmeter is thrown across the terminals of the standard transformer. With the switch in position S1A, the voltmeter is thrown across the terminals of the transformer under test. The voltmeter should be read with the switch in each position. If the winding ratio be the same as that of the standard transformer, the two voltmeter readings will be identical.
It may be stated in general that before taking the readings for this test, the zero position of the pointer on the meter tested should be noted, and if it be more than two-tenths of a division off the zero mark, the case of the meter should be removed and the pointer straightened.
Furthermore, it will be noticed from the readings here recorded that the test is started at the high reading end of the scale; this is done in order that the pointer may gradually be brought up to this spot, by slowly cutting out of circuit the adjustable resistance, and thus show whether or not the pointer has a tendency to stick at any part of the scale. If the meter seem to be defective in this respect, it should be remedied either by bending the pointer or scale, or by renewing one or both of the jewels, before the comparison with the standard is commenced.
It is obvious from the readings recorded for the 150 scale voltmeter, that as compared with the corresponding deflections of the standard, the former are a trifle low.
In order to determine for each observation how much too low they are, it is necessary to divide each reading on the standard by the corresponding reading on the meter tested. The result is the amount by which a deflection of this size on the meter tested must be multiplied in order to obtain the exact reading. This multiplier is called a constant, and as shown, a constant is determined for each of the six observations.
The average constant for the six readings is then found, and this is taken as the constant for the meter as a whole; that is, whenever this 150-scale voltmeter is used, each reading taken thereon must be multiplied by 1.007 in order to correct for its inaccuracy.