Measured Noise Temperature Versus Theoretical Electron Temperature for Gas Discharge Noise Sources

In the past, measured noise temperatures T/sub n/ of a few commercially available gas discharge noise sources were indicated as agreeing with the predicted electron temperature T/sub e/ of the positive column based on the von Engel and Steenbeck relationship. Data were taken over the past 2 years on argon tubes over a pressure range of 5 to 40 mm and on neon tubes at 20 mm, with current variations from 100 to 300 mAdc. These data were compared against predicted electron temperatures. For the argon tubes at pressure-radius products greater than 20 mm/spl dot/cm there appeared to be reasonable correlation between the measured noise temperature and the predicted electron temperature although it is suggested that this correlation was fortuitous. For argon pressure-radius products less than 20 mm/spl dot/cm the measured noise temperature was as much as 15 percent lower than the predicted electron temperature. For neon tubes at 20-mm pressure, with the same variation in tube radius, and for pressure-radius products less than 24.0 mm/spl dot/cm, the measured noise temperature differed even more than for argon from the predicted electron temperature. A difference of as much as 30 percent at a pressure-radius product of 3.0 mm/spl dot/cm was observed. A qualitative explanation for argon is presented based mainly on the fact that these discharges do not have a Maxwellian distribution of electron velocities nor a velocity independent electron collision frequency. For neon the wide variation was not understood.