and constantan thermocouple wires (0.0025 cm. diameter) through the holes from the 

 bottom side of the Teflon insert, from which about a 4-cm. length of each wire is 

 allowed to protrude. About 0.5 cm. of the ends of the copper lead wires (Beldon #8640, 

 26 gage) are scraped clean of all insulation and inserted into the holes from the top 

 side of the Teflon insert. The positive copper lead is tightly wedged into the hole 

 with the chromel wire and the negative lead with the constantan wire. This provides 

 a tight contact between the copper and thermocouple wires. An alternative method of 

 attaching the thermocouple wires to the copper lead wires is to solder the chromel wire 

 to the positive copper lead and the constantan to the negative copper lead. A high 

 grade of silver solder should be used. 



The chromel and constantan wires protruding from the bottom of the Teflon insert 

 are twisted together with an "L"-shaped piece of wire (26 gauge, 3 cm. long with an "L" 

 bend of about 3 mm.). The "L"-shaped tool is held between the thumb and forefinger, and 

 is twirled slowly around the two thermocouple wires, twisting the two wires tightly 

 together. The wires are twisted together until their junction is about 3 mm. away from 

 the Teflon, then the "L"-shaped tool is inserted between the chromel and constantan 

 wires and the junction pulled out to exactly 4 mm. from the Teflon insert. This tightens 

 the junction and provides a uniform length to the thermocouple assembly. The excess 

 twisted wire can be cut away with a scissors, leaving about 3 mm. of twisted wire below 

 the junction. The thermocouple is now ready to be welded or 'soldered; however, welding 

 is preferred because soldered thermocouples tend to age rather quickly and do not have 

 a linear response throughout the range of vapor pressures of interest (Campbell and 

 others 1968) . 



Welding can be accomplished as suggested by Campbell and others (1968). Also, a 

 somewhat simplified technique may be used wherein the negative lead from the arc-welder 

 is attached to the ends of both copper leads from the thermocouple. (A diagram of the 

 electrical circuit used in my arc-welder is given in Appendix 5.) The positive lead 

 from the arc-welder is attached to a sharpened graphite rod (4H graphite lead), which 

 is used to provide the arc for welding. Welding is accomplished by touching the graph- 

 ite rod to the end of the twisted thermocouple wires . The twisted thermocouple wires 

 should be welded slowly (by adjusting the voltage output from the welder) at first, 

 building up a bead of fused metal. Welding should continue until about one twist re- 

 mains beyond the beaded junction. The bead should be about three times the diameter of 

 the thermocouple wires, although the most successful psychrometers will have as large a 

 bead as possible. The largest junction diameter possible with this technique appears 

 to be about five times the wire diameter (Campbell and others 1968) . The best results 

 will be achieved if the entire thermocouple assembly process is accomplished with the 

 aid of a binocular scope with a 7X to 10X magnification. 



A legitimate argument can be raised concerning the advisability of welding thermo- 

 couples in normal air rather th#n in nitrogen gas or oil. In the presence of relatively 

 high oxygen partial pressures, the welding process leaves a deposit of oxidized material 

 on the thermocouple junction. This deposit may influence thermocouple output and psy- 

 chrometer calibration. A number of workers have used some rather elaborate devices to 

 hold the thermocouples in place during construction and welding. In general, a jig is 

 used to hold the thermocouple securely, which is then placed in a sealed enclosure and 

 flushed with nitrogen gas (Lopushinsky and Klock 1970). A' somewhat simplified yet 

 equally effective method is to perform the welding operation in a small pool of light- 

 weight optically clear oil such as 10-weight motor oil or even commercial salad oil. 1 

 The thermocouple assembly is completely immersed in a petri dish of oil, and the junc- 

 tion is welded by touching the graphite rod to the junction until the bead is formed. 

 In this manner a greater degree of control can be exercised over the rate of welding, 

 and incidences of overheating and breaking the junction are much less frequent than 

 when welding is done in a gas. The thermocouple should be dipped into acetone and 

 slightly agitated to remove all traces of oil. 



Eric Campbell, personal communication, 1969. 



12 



