Quantifying the error contribution of noise parameters on Y-factor noise figure measurements

Low noise receivers are critical in telecommunication systems. Lower the noise, the less transmission power is required to receive a signal within a given distance. This has a major impact on energy consumption and hence longer battery life. Noise contribution of a receiver is often specified by its noise figure. It is the ratio of the signal to noise power ratio at the input, Si/Ni, to the signal to noise power ratio at the output, So/No. The Y-factor method is the most common technique for determining the Noise Figure (NF) or effective input noise temperature (Te) of a Device Under Test (DUT). One of the reasons for its popularity is its simplicity. It is easy to do. An important assumption was made to allow this simplicity, the measurement system is well match or very low reflection interactions. When the assumption of the Y-factor measurement method is not met, significant error will occur. This paper will introduce a way to estimate the mismatch errors and in particular the noise parameter effect of the noise source output match. A new model for predicting the added uncertainty caused by the reflection coefficient of the noise source will be presented.