Phase Noise Cross-Correlation Measurement System Based on NI Vector Signal Analyzers

  Print Print

"Due to NI off-the shelf technology, we reached our final goals with the highest efficiency and got the research positive outcome. We will continue to evaluate the approach, making measurements faster and decreasing the system price and size. "

- Davit Zargaryan, 10X Engineering LLC

The Challenge:
Developing an amplitude/phase noise cross-correlation measurement system based on two vector signal analyzers (VSAs).

The Solution:
Using NI VSAs to develop a more precise amplitude/phase noise measurement system based on the cross-correlation technique.

Davit Zargaryan - Find this author in the NI Developer Community
Levon Grigoryan - Find this author in the NI Developer Community

About 10X Engineering

Founded in 2013, 10X Engineering is an NI Alliance Partner with sufficient knowledge and experience in engineering complex solutions for RF product (units, devices, components) quality and line testing, custom-designed ATE system assembly, verification, and development. Our offering includes solutions for a variety of industrial sectors such as RF software defined radio, radiolocation, spectrum monitoring, and more. Every request received follows project technical and software functional requirements clarification, RF measurement methodology selection, and hardware configurations. In some cases, if needed, we take responsibility to prove a concept and demonstrate the flexibility and reliability of our solution.

Problem Background and Solution

Microwave oscillators, amplifiers with low residual phase, and amplitude noises are popular and actively contribute in RF hardware development. Phase/amplitude noises are the result of small random fluctuations or uncertainty in the phase/amplitude of an electronic signal, which limits a product’s technical performance. The most popular device under test (DUT) is the oscillator, in which the short term frequency instability is considered as a phase noise PM (frequency domain-phase noise, time domain-jitter) and amplitude one as amplitude noise AM [1]. AM noise does not peak around the carrier as phase noise does, so is negligible close to the carrier. Far from the carrier, at offsets typically greater than 1 MHz, AM noise starts to dominate.

S(t) = A(1 + A’(t))sin(2πf + ϕ0 + ϕ’(t)) [1]

A = amplitude of the signal
A’(t) = amplitude fluctuations (amplitude noise)
F = frequency of the signal
ϕ0 = phase offset
ϕ’(t) = phase fluctuations (phase noise)

Accurate measurement of phase/amplitude noise is challenging for radio engineers. Methods include direct spectrum technique, frequency discriminator method, phase detector techniques, residual method, two-channel cross-correlation technique, and more. However, the cross-correlation method is considered the most productive in providing the highest measurement sensitivity up to -180 dBc/Hz. This case study is not focused on comparing the advantages and disadvantages of the measurement methods, but presents a cross-correlation measurement system implementation on two vector signal analyzers (VSAs) using 10X Engineering software based on the LabVIEW development environment.

We used two NI PXIe-5668R 14 GHz VSAs and an NI PXIe-5653 RF synthesizer as a DUT. All NI analyzers (NI PXIe-5660/5661/5663/5665/1470) were compatible. We used the NI hardware and software configuration to create the phase noise measurement cross-correlation technique through two VSAs. We used NI hardware and software combinations to develop the final solution within a short time period.

Figure 1. Phase Noise Cross-Correlation System

Our company currently offers software solutions for the most popular method of AM/PM measurement: direct spectrum methods (DSMs). The AM/PM DSM installation package includes a driver (fully integrated in the user program) and a user friendly software front panel for fast and easy access. 

Figure 2. Experimental Data—Phase Noise Cross-Correlation System Measurement Comparison with Datasheet

Figure 3. Experimental Data—Amplitude Noise Cross-Correlation System Measurement Comparison with Datasheet

We decreased development time and costs due to NI off-the-shelf technology and increased efficiency for the development of our solution.

AM/PM DSM Features

  • Simultaneous phase and amplitude noise fast measurement
  • CW, pulsed modes support
  • Compatible with NI VSAs
  • User friendly interface
  • Data reporting

Such systems are relatively low in cost, but due to the hardware limitations, the current method can cover the measurements when these conditions are considered:

1. VSA’s phase noise > DUT’s phase noise [A]
2. VSA dynamic range > DUT phase noise (for defined offset) [B]

Figure 4. Direct Spectrum Method Challenges

These points are overcome (partially) in the cross-correlation method. With this method the architecture includes two independent running receivers with separate local oscillators. The final measurement uses the correlation output of FFT_ 1st, FFT_2nd channel. This works because the noise from the DUT is common between both paths, but the noise contributed from each internal reference oscillator is independent. Thus, over time, the noise contributions from the independent sources show a zero cross correlation. However, the noise from the DUT correlates and ultimately dominates the output measurement.

In aforementioned measurement methodology, the cross-correlation function has a dramatic impaction on system performance in term of sensitivity. The sensitivity of the system increases with the increase of number of cross-correlation operations [2].

Sensitivity improvement=5log10N,
Where the N is a cross-correlation number.

Here is a small calculation: to improve sensitivity of the system up to 20 dB the 10,000 number of correlation is required. This approach dramatically increases measurement time, what becomes a bottleneck for proposed method. The offered architecture's potential, is researched in below (phase/amplitude noise cross-correlation measurement technique implementation through two NI VSAs) .

With our experimental system architecture (two NI VSAs) and special software, NI PXIe-5653 LO 1-5.4 GHz, LO 3-800 MHz phase and amplitude noises were measured, that were impossible to do with the direct method. The result shows that current configuration satisfy the criteria of measurement AM/PM noise for both CW and pulsed modes up to -160 dBc/Hz.

Figure 5. Direct Spectrum and Cross-Correlation Methods Architectures

Due to NI off-the shelf technology, we reached our final goals with the highest efficiency and got the research positive outcome. We will continue to evaluate the approach, making measurements faster and decreasing the system price and size. Two VSAs measurement technique still can be actual for automated test system ATE where for specific parameters measurement MIMO configuration is required (phase difference of channels). Additionally including simple switching the system can cover AM\PM noise measurements along with the others, instead of buying special standalone hardware for concrete AM\PM noise measurement.

Author Information:
Davit Zargaryan
Find this author in the NI Developer Community

Bookmark and Share

Explore the NI Developer Community

Discover and collaborate on the latest example code and tutorials with a worldwide community of engineers and scientists.

‌Check‌ out‌ the‌ NI‌ Community

Who is National Instruments?

National Instruments provides a graphical system design platform for test, control, and embedded design applications that is transforming the way engineers and scientists design, prototype, and deploy systems.

‌Learn‌ more‌ about‌ NI