The principal parts of the apparatus used to test for flame resistance of electric cables, signaling cables and splices shall include#:

(a) Test chamber. A rectangular enclosure measuring 17 inches deep by 14 1/2 inches high by 39 inches wide and completely open at the top and front. The floor or base of the chamber shall be fabricated or lined with a noncombustible material that will not extinguish burning matter which may fall from the test specimen during testing. The chamber shall have permanent connections mounted to the back wall, sides, or floor of the chamber which extend to the sample end location. These are used to energize the electric cable and splice specimens. They are not used, but may stay in place, when testing signaling cables.

(b) Specimen holder (support). A specimen holder (support) consisting of three separate metal rods each measuring approximately 3/16 inch in diameter (nominal) to support the specimen. The horizontal portion of the rod which contacts the test specimen shall be approximately 12 inches in length.

(c) Gas ignition source. A standard natural gas type Tirrill burner, with a nominal inside diameter of 3/8 inch, to apply the flame to the test specimen. The fuel for the burner shall be natural gas composed of at least 96 percent combustible hydrocarbons, with at least 80 percent being methane.

(d) Current source. (For electric cables and splices only). A source of electric current (either alternating current or direct current) for heating the power conductors of the test specimen. The current source shall have a means to regulate current flow through the test specimen and have an open circuit voltage not exceeding the voltage rating of the test specimen.

(e) Current measuring device. (For electric cables and splices only). An instrument to monitor the effective value of heating current flow through the power conductors of the specimen within an accuracy of ±1 percent.

(f) Temperature measuring device. (For electric cables and splices only). An instrument to measure conductor temperature within an accuracy of ±2 percent without the necessity of removing material from the test specimen in order to measure the temperature.