overview
提示：这里可以添加技术概要
Residual phase noise measurements remove the effects of external noise sources such as power supplies or input clocks, while absolute phase noise measurements include the noise from these sources. The residual phase noise setup isolates and measures the additive phase noise of a device. Using this information, designers can select individual components in the signal chain to meet the phase noise requirements of the overall system. Phase noise plots of clocked devices are included in this article to highlight the properties of the residual phase noise device. Additionally, it shows how additive phase noise can be used to identify sources of noise problems in the signal chain.
overall architecture process
提示：这里可以添加技术整体架构
The setup for measuring the additive phase noise of a device under test (DUT) is shown. Note that two DUTs are used; each DUT is connected to a common power supply and an input clock. The phase noise generated by these common noise sources is correlated at each DUT output. By simply modeling the phase detector as an analog multiplier with gain KPD, the output phase noise can be derived:\
Explanation of technical terms
提示：这里可以添加技术名词解释
technical details
提示：这里可以添加技术细节
：
Where E1 is the amplified DUT1 output signal, E2 is the amplified and delayed DUT2 output signal, EC1 and EC2 are the signal power, θM1 and θM2 are the amplitude of the phase noise, ωC is the carrier frequency, and ωM is the offset frequency. Superposition is applied so that the phase noise inherent to the DUT is negligible when considering phase noise from external sources. If DUT1 and DUT2 have the same excess phase transfer function, then the portion of θM1 generated by the clock source and power supply is equal to the portion of θM2 generated by the common clock source and power supply. This phenomenon is called power pulling and can be simply described by the following equation:
Thus, the magnitude of the phase modulation is given by the voltage noise on the power supply multiplied by KP - power supply pulling gain (rad/V). If DUT1 and DUT2 have equivalent power-supply pull gains, these noise sources at the output of the phase detector can theoretically be canceled, leaving only the unrelated noise of the two DUTs for measurement. Intrinsic DUT noise can be determined with some additional assumptions. Since the rms phase error due to device noise is typically very small, we can use a small angle approximation to change the expression for the output carrier to:
The output of the phase detector is demodulated so it can be called the "baseband signal". Once the phase detector gain and input signal power are determined, the actual phase noise can be calculated (assuming negligible phase noise from the amplifier). The inherent noise of each DUT is uncorrelated, so the noise they generate is the same, and the rms sum is the measured output phase noise. To do this, we subtracted 3dB from the phase noise (unit: dBc/Hz) measured by the spectrum analyzer to determine the noise produced by each DUT. This expresses the phase noise power relative to the signal power:
The noise generated by the amplifier can be significant when making very sensitive phase noise measurements. The residual phase noise of the amplifier is measured by removing DUT1 and DUT2 from the circuit and applying the power splitter output directly to the amplifier. The amplifier input signal power must be similar in amplitude and slew rate to the actual DUT output signal. Using the above procedure, the accurate DUT phase noise can be obtained by subtracting the measured amplifier phase noise from the measured DUT phase noise. Again, the key is to make sure the amplifier's gain and noise figure are as close to each other as possible.
summary
提示：这里可以添加总结
Note that a DUT that requires a clock input signal will have a somewhat noisy front-end amplifier installed. Therefore, a clock source with a low slew rate may inadvertently increase the phase noise generated by the DUT due to threshold uncertainty at the amplifier input. When using a sinusoidal clock source, use the maximum allowed amplitude to maximize the slew rate.