Noise Figure Measurement—Overrange Method

The noise figure analyzer is usually used as an optional function of the spectrum analyzer. It is used with a standard noise source and uses the Y factor method to measure and analyze the noise figure of the DUT under test.

For DUTs with small gain or even negative gain, and the noise coefficient is greater than 20dB, the noise source needs to have an ENR much larger than the standard noise source shelf product. This article explains the integrated production of this noise source and its calibration and measurement methods.

Measurement units commonly use attenuators to measure and calibrate the accuracy of noise figure analysis. When the attenuation value is greater than 20dB, it is necessary to integrate a self-made and standardized ultra-high ENR noise source.

Noise Figure

The noise coefficient describes the degree of internal noise of the DUT under test and the degree of deterioration of the signal-to-noise ratio, that is, the ratio of the input signal-to-noise ratio and the output signal-to-noise ratio.

• F=SNR_in/SNR_out

When defining the noise figure, the input noise Nin = N0 is the electronic thermal noise power of an ideal device at 290K -174dBm/Hz.

Attenuator Noise Figure

According to the above definition, the attenuation amount att (dB) of the attenuator is equal to its noise coefficient.

• The output noise of the attenuator is equal to the input noise

• Nout=Nin=N0;

• Output signal power Sout=Sin-att;

• Therefore NF=Sin/N0-Sout/N0=att

Therefore, the noise figure of a negative-gain DUT is usually equal to its attenuation.

When cascading DUT at multiple levels:

• F = F1+(F2-1)/G1+...

• When G1>>1, the total noise figure depends on the first-stage noise figure F1

• When the first stage is attenuation (negative gain) and G1<1, it is considered that G1=1/F1

• F = F1+(F2-1)F1 = F1·F2

• The total noise figure is greater than the first stage noise figure

Standard Method for Y Factor Analysis of Noise Figure

Use the spectrum analyzer with an external noise source to perform noise figure testing:

• The control end of the noise source is connected to the power supply and control interface of the spectrum analyzer;

• The excess noise ratio ENR of the noise source is used as the reference standard value and is input in the noise figure measurement function of the spectrum analyzer;

• The ENR at each frequency of the noise source is the difference between its hot state output noise power density and its cold state noise power density. For example, a noise source is 3GHz, the temperature is 290K, the cold state output is -174dBm/Hz, and the hot state output is - 159dBm/Hz, ENR=15;

• Calibration: The noise source RF port is directly connected to the spectrum analyzer RF port, and calibration is performed to obtain the Y factor and the spectrum analyzer's own noise;

• Measurement: Insert and connect the DUT between the noise source and the RF port of the spectrum analyzer, perform scanning measurements, and obtain the noise coefficient and gain of the DUT.

• Gain range: -20~60dB

• The noise figure range is limited by the excess noise ratio ENR of the noise source:

• ENR5dB noise source: 0~20dB

• ENR15dB noise source: 0~30dB

• ENR21dB noise source: 0~35dB

• In the above noise figure range, when the noise figure is greater than ENR, the DUT gain is required to be greater than the difference between its noise figure and ENR.

Measurement and Calibration of Noise Source Extra Noise Ratio ENR

Use a standard noise source with known traceable ENR to perform ENR measurement calibration of the target noise source to be calibrated:

• The standard noise source can be a diode noise source calibrated and traceable by the superior measurement unit;

• The standard noise source can also be a pair of hot and cold resistors (such as a normal temperature resistor and a liquid nitrogen low temperature resistor), traceable to the thermodynamic temperature;

• Spectrum analyzer noise figure analysis, pass-through calibration with standard noise source;

• Replace the noise source being calibrated, measure and record ENR.

Noise Figure Analysis Overrange Method

One of the decisive factors for noise figure measurement uncertainty is the excess noise ratio ENR of the noise source. The optimal measurement conditions are:

• ENR – (NF of SA) > 3 dB

• ENR – (NF of DUT) > 5 dB

• (NF of DUT) + (Gain of DUT) – (NF of SA) > 5 dB

When the DUT noise figure is large and the gain is small:

• It is necessary to integrate the noise source of the amplifier on the basis of the standard noise source to achieve ultra-high ENR noise output;

• The integrated noise source ENR is greater than the attenuation loss.

Test steps:

• Perform pass-through calibration using a standard noise source;

• The standard noise source is connected to the amplifier and matching attenuator as an overall homemade noise source;

• The control power supply of the noise source controls the switch of the amplifier power supply, or switches to manual mode and manually switches between the cold state and the hot state for scanning measurement;

• The self-made noise source is passed through the spectrum analyzer for ENR measurement;

• Use the above measured ENR values ​​as the standard ENR values ​​for self-made noise sources;

• Use homemade noise sources instead of standard noise sources to test the noise figure of the DUT;

• The test layout is as follows:

Summarize

This solution performs measurement calibration on the ENR of the noise source;

Based on existing standard noise sources, the amplifier integrates high ENR noise sources;

Pay attention to the synchronization of the power on and off of the noise source and the amplifier, ensuring that the power supplies of both are turned off when they are cold and turned on when they are hot.

This solution is especially recommended when using an attenuator greater than 20dB for measurement calibration of the noise figure analyzer .