The efficiency or life of an electric lamp depends partly upon the temperature rise at its glass wall. The temperature rise in vacuum lamp is very small but it becomes remarkably high in gas-filled lamp. It, therefore, becomes necessary to determine the exact temperature at the glass wall in order that the proper materials be selected for use in the construction of such lamps. In a recent report the Japanese Electrotechnical Committee and the Illuminating Engineering Society gave a specification about the allowable range of the temperature rise, but the temperature of the glass wall is indefinite giving different results by different methods. In this report possible errors in measuring temperature by various methods are discussed and the means to get true temperature are suggested.
Thermo-couple method. The causes of possible errors in this methods are as follow:
(1) Contact of the hot junction to the glass wall;
(2) Radiation received by the junction;
(3) Heat lost by conduction through the junction wire. When the hot junction of a thermo-couple is attached by some cement to the glass wall the measured temperature is much higher than the true value, and when the hot junction is only put on the glass wall lightly pressed by its elasticity the result is much lower, the former is due to the absorbtion of heat radiation by the cement and the latter to the poor contact to the glass wall. Thermo-couple of thicker wires absorbs much more radiating energy and discipates much more heat by its good heat conductivity and the result is more errorneous. When Pt-PtRh 13% junction of 0.1mm in diameter is taken and its hot junction sealed in glass wall of a lamp it gives very good result owing to its good contact to glass and very little amount of heat received in and scattered out by radiation or conduction. Experimentally we can eliminate this error caused by radiation or conduction and obtain a correct result.
Electrical resistance method. The cause of possible error in this method is only radiation energy received by the resistance wire, variation of its amount depends upon:
(1) Number of turns of resistance wire wound around the lamp;
(2) Color of the insulator on the resistance wire;
(3) Size of the resistance wire.
The resistance wire receives heat from the glass wall by conduction and from the lamp filament by radiation. The latter depends on the color of the insulator on the resistance wire (black or white, clean or dirty) and on the size of the wire. Now in every case temperature is a linear function of number of turns of the resistance wire wound around the lamp and the more the number of turns the higher the temperature. By exterpolation we can easily obtain a definite temperature corresponding to zero turn, therefore, independent of any kind of the resistance wire. This temperature coincide very well with that corrected result obtained by the junction method.
View full abstract