Dave's points are correct. You will also find on the tube a filament current and you should be sure that the filament supply is delivering that current at the "zero" pressure.
With regard to the McLeod gauge and condensables, because the McLeod works by compression (compressing the gas by a known volume factor to provide a corresponding increase in pressure so that the height of the mercury column is measureable, a nice application of Boyle's law) the compressable gases (water vapor in most applications) will condense and not be measured. Since in unbaked systems at low pressures water vapor is the largest constituent, the reading will be off by a large margin from the true pressure. The TC gauge however is measuring the pressure of the permanent and condensable gases. By putting a cold trap in the system to get rid of the condensables, both gauges will be measuring the same mixture of residual gases.
Today the accepted calibration practice is to compare the TC gauge to a NIST traceable capacitance manometer (aka capacitance diaphragm gauge or CDG). CDGs are available with guaranteed accuracies of 0.5% down to 0.05% of reading accuracy. TC gauges, out of the box, are usually in the 15% to 25% of reading accuracy over their range. Accuracy tends to be worse at the low end (<10 microns) and high end (toward 1 torr).
Steve
Created on Tuesday, April 17, 2001 9:37 PM EDT by Steve Hansen