Single-Port Technique for Adaptor Efficiency Evaluation

NFRAD - Review of the New Nist Noise Measurement System

“…In previous systems, the mismatch was measured with a six-port reflectometer, and the asymmetry was determined from measurements of the same sources on two different ports [4,5]. In the present system, the mismatch factors are calculated from measured reflection coefficients and the efficiencies are measured directly on the VNA [6,7]. The asymmetry is just the ratio of the efficiency of the cryogenic path over the efficiency of the DUT path.…”

“…The asymmetry is defined as the ratio of efficiencies s / x appearing in the radiometer equation. With NFRad, the efficiencies are typically measured with a reflective-termination technique [6,7]. The uncertainty analysis for this method [7,8] was applied to the measurements made on the various switch paths (common to ports 3-6).…”

“…With NFRad, the efficiencies are typically measured with a reflective-termination technique [6,7]. The uncertainty analysis for this method [7,8] was applied to the measurements made on the various switch paths (common to ports 3-6). It resulted in an uncertainty in T x due to the asymmetry that is typically on the order of 0.3 percent.…”

Design and Testing of NFRad- :

“…The correlated and uncorrected uncertainties are given by [8] The third method for measuring the asymmetry is to measure the two efficiencies directly, using the reflective termination (RT) method developed by Daywitt [6,7]. This method computes the intrinsic efficiency of a passive two-port from the reflection coefficients measured when one of the ports is terminated in two different reflective terminations.…”

“…Previous uncertainty analyses for NIST radiometers [1,8] (25) To develop the general expression for u^we model [12] the variable Ty as T,j=ir-^M,^R^j, (26) where ris the true value for the noise temperature of the device being measured, M^is a random variable representing variations from measurement to measurement (7), and Ry is a random variable that varies with the readings and the measurement /. Our estimate of the true value is just the mean of all the readings, 1 rK A '^''T= -^-22 X^tn (27) which is also equal to the mean of the Nj^separate measurements.…”