This article presents a new network analyzer sharing (NAS) system that enables four users to simultaneously access a two-port vector network analyzer (VNA) operating up to 8.5 GHz. To enable four simultaneous users, each port of the VNA is split using an SP4T RF switch. Each channel has two ports and performs high speed sequential switching. The VNA screen is divided into four and displayed synchronously on each user’s monitor. Image processing techniques are utilized to set the optimal high speed switching time automatically.

With the telecommunications industry rapidly advancing, there is a notable increase in demand for devices and components. This is leading to a greater need for measurement instruments and the demand for high performance VNAs is increasing significantly. The VNA is an expensive piece of equipment that analyzes the electrical characteristics of electronic devices. It sends and receives electrical signals to measure and analyze the response of the device under test (DUT). This allows for the examination of parameters such as frequency response, reflection and insertion loss. The VNA is widely used in the design and testing of wireless communications, radar, antennas, high frequency circuits and more.

Typically, these VNAs are designed to measure a single DUT and there are limitations when measuring multiple DUTs or having multiple simultaneous users. To address this issue, advances now enable multiple measurement ports in a single VNA.¹ Currently, there are four-port, six-port, eight-port and other multiport VNAs available on the market. However, simply increasing the number of ports in one VNA increases the complexity of the system and increases the price. Alternatively, to improve price competitiveness, the number of ports can be increased with a switch matrix.^2,3 The primary objective of port expansion is to facilitate multiport device measurements or to analyze multiple S-parameters. Port expansion does not allow independent users to use one VNA at the same time.

Recently, several authors of this article proposed a NAS system that allows multiple users to access a VNA independently and simultaneously.⁴ To demonstrate feasibility, the study implemented a switch matrix for port expansion and presented a de-embedding method to calibrate each channel. However, this study did not address high speed switching methods. In addition, the switch matrix was implemented with an SPDT switch operating up to 6 GHz. However, this was insufficient to achieve the performance of existing commercial VNAs. The article presents an efficient switching module and its optimal high speed switching method to enable four users to simultaneously access a commercial two-port VNA operating up to 8.5 GHz.

OVERALL SYSTEM

Figure 1 Overall schematic diagram of the NAS system.

Figure 2 Switching module.

The NAS system primarily consists of switching hardware to expand the VNA port, along with software that switches it rapidly and provides each measured screen to individual users. Figure 1 shows the general schematic diagram of the proposed NAS system. Here, four users connect to a single two-port VNA through a switching module and measure the DUT while viewing their respective measurement screens.

The switching module includes two SP4T switches to split each VNA port into four. Port A of the VNA is split from A1 to A4, and Port B is split from B1 to B4, as shown in Figure 1. Each user is assigned two ports, for example, A1 and B1, for the first user. The SP4T switch operates up to 8.5 GHz to achieve the full performance of a typical commercial 8.5 GHz VNA. In addition, the switching module includes a microcontroller unit (MCU). This controls the high speed on/off switching of the SP4T switch, allowing all the users to access a single VNA simultaneously. With this technique, determining the switching time is a critical issue because incomplete measurements can lead to errors.

SWITCHING HARDWARE

The switching module schematic of Figure 1 expands each port of the VNA to four and the MCU turns the ports on and off sequentially according to the appropriate switching time. The fabricated switching module is shown in Figure 2. It has two ports on the front for connecting a VNA and four ports on each side to provide two ports per user.

In previous work,⁴ 6 GHz SPDT switches were connected in two stages to achieve the 1:4 split of the VNA port. In this work, an 8.5 GHz SP4T switch is used to connect one input to any one of four outputs or vice versa. The use of an SP4T reduces the size of the switching module and the cable lengths between ports. These improvements are expected to decrease transmission losses. The operational frequency range of the SP4T switch extends to 8.5 GHz, ensuring full compatibility with commercial VNAs.

The manufactured switching module is interconnected with a semi-rigid RF cable to minimize cable loss and increase ease of assembly. Since the magnitude and phase of the transmitted signal will change with different RF cable lengths, it is recommended to match cable lengths to reduce electrical errors between channels. These errors make channel calibration difficult and can lead to measurement errors.⁵

The S-parameters were measured to determine the channel characteristics of the manufactured switching module.⁶ Performance assessment of the switching module involves measuring path-specific insertion loss, return loss and isolation. Insertion loss reduces the signal power that passes through the cables, connectors and switches of the switching module. Minimizing insertion loss is important in RF system design to ensure efficient signal transmission and reception. In the insertion loss measurement, only the specific path of the manufactured switching module is activated, while the others are turned off. S₂₁ is measured on the activated path with a VNA. These results are shown in Figure 3a for Port A to all the associated output combinations and in Figure 3b for Port B to all the associated output combinations. These values are similar for all eight paths, decreasing from approximately -0.7 dB to -4 dB at the maximum operating frequency of 8.5 GHz.

Figure 3 Insertion loss characteristics of the switching module.

Figure 4 Return loss characteristics of the switching module.

Figure 5 Isolation characteristics of the switching module.

Reflection loss measures the amount of signal power that is reflected toward the source due to impedance mismatches or discontinuities in the transmission path. When a signal encounters an impedance change, such as at the interface between two different transmission lines or at a connector, a portion of the signal is reflected rather than transmitted. These results are shown in Figure 4a for the Port A output paths and in Figure 4b for the Port B output paths. For the measurement, only one specific path was activated and S11 was measured at one end of the path. The characteristics of all paths are similar and the worst case is approximately -12 dB.

Isolation refers to the degree of mutual interference between two different paths in the switching module. It indicates the extent to which a signal input to one path is transmitted to other paths as noise or external signals. Higher isolation minimizes unwanted signal interference within the system, improving the accuracy and reliability of the NAS system. The results shown in this section examine the amount of isolation between branched ports. For example, after activating the path between ports A and A1, S₂₁ is measured between Ports A1 and A2, Ports A1 and A3 and Ports A1 and A4, respectively. The same is true for Port B. The results are shown in Figure 5a for the Port A combinations and Figure 5b for the Port B combinations. Not unexpectedly, the isolation performance deteriorates as the frequency increases, but the value is below approximately -55 dB at 8.5 GHz, showing excellent performance.

SWITCHING SOFTWARE

The software operating the NAS system allows four users to measure their respective DUTs simultaneously. This software controls the SP4T switches according to a series of switching schedules to activate four channels sequentially. In conjunction with this, the VNA measures the S-parameters of each activated channel and displays them on the primary screen. The primary screen is divided by channel and the measured S-parameters display is transmitted to each user monitor. Performing this operation repeatedly at a high speed allows four users to use one VNA simultaneously.

With this technique, it becomes very important to determine the appropriate switching timing to achieve the highest measurement speed. This must be done to ensure sufficient time for the VNA to measure each channel. A VNA’s measurement time is variable depending on not only the measurement conditions (e.g., IF bandwidth, measurement points and the number of markers) but also the commercial VNA product used.

Figure 6 Switching schedule of NAS system.

Figure 6 shows the operational timing of the overall NAS system. This includes the switching times between channels and the one-cycle time. When a switching command is issued, Channel 1 is activated and the VNA performs a measurement. When Channel 2 is activated, the VNA performs another measurement and this process is repeated to Channel 4. Then, the measurement results are displayed on the primary screen, divided and transmitted to each user screen. As mentioned, it is difficult to know the measurement time of a VNA. Suppose the switching time between channels is too short. In that case, the measurement data will be incomplete because the measurement of the new channel starts before the measurement of the previous channel is completed.

Figure 7 Differential image processing concept for setting optimal switching time.

Figure 8 NAS system setup.

Figure 9 Measurement results of DUT sample.

The measurement results screen for the case where the switching interval is too short is shown on the left side of Figure 7. Conversely, if the switching interval is too long, the user’s screen refresh time will be inconveniently long. To eliminate these issues, this investigation developed a method to set the optimal switching time to achieve the highest switching speed. This technique relies on the fact that the measurement data image varies between channels when the switching time is not sufficient. The right side of Figure 7 shows the concept for setting optimal switching times based on differential images between channels.

As an automatic setting environment, the two ports of each channel are directly connected by an RF cable without a DUT. Then, the same measurement data image can be obtained for all channels, thereby improving the accuracy of the difference image. In this work, the initial switching delay time was set to 10 ms to obtain the measurement data images shown on the left side of Figure 7. To obtain the significant differential images, regions of interest (ROI) were set instead of the entire measurement image.⁷ Differential images are obtained by subtracting the ROI of each channel from the reference image of Channel 1. The resulting differential images are illustrated on the right side of Figure 7. The differential images for Channels 2 and 3 show a difference in the pixels. This difference arises because of the short switching time. Therefore, it is necessary to increase the switching delay time. The delay time is adjusted based on the number of pixels in each differential image. In this work, the threshold was set to 100 pixels. This value is obtained experimentally by setting the switching time to a sufficiently long value to ensure that the four measured images are almost the same. The delay time is then determined based on this threshold of 100 pixels. If the threshold is exceeded, an additional 1 ms is added to the initial delay time. This process is repeated until the number of different pixels becomes less than the threshold. Using this technique, the switching delay times between each channel in the manufactured NAS system were determined to be 16 ms, 17 ms and 17 ms.

Next, the cycle time needs to be determined. This is necessary to determine the timing for switching back to Channel 1. Without this determination, Channel 1 may be switched before the measurement screen synchronization is complete, which will result in an error. This process is the same as determining the optimal switching time, as described earlier. This process yields a cycle time of 250 ms.

SYSTEM TEST RESULTS

Figure 8 shows a picture of the assembled NAS system. A different DUT is connected to each channel and its measurement data is displayed on each user’s screen. Following the previously described method for determining high speed switching times, the cycle time was optimized to 250 ms. Therefore, the measurement data on each user’s screen is refreshed four times every second.

It is also important to ensure consistency of measurement results for each channel. To verify this, the measurement results of a given DUT through the NAS system were compared with those measured by connecting it directly to the VNA. The comparisons are shown in Figure 9. The DUT is a cavity filter with a passband between 3.1 and 4.8 GHz. The data measurements from the NAS system agree very well with the data measured directly from the VNA.

CONCLUSION

This article has presented a NAS system that allows four users to access a single VNA simultaneously. The system is designed to replicate the full performance of existing commercial VNAs operating up to 8.5 GHz. Measurements confirm that there is almost no difference in the characteristics between the channels. As an outcome of this technique, the optimal high speed switching time can be automatically set under any measurement conditions. The developed NAS system allows for simultaneous usage by four users, with a refresh time of 250 ms on each user’s screen. This level of refresh is generally not inconvenient for measuring the characteristics of RF components or performing simple tuning tasks. Therefore, the NAS system alleviates the burden of using expensive VNAs and improves productivity.

ACKNOWLEDGMENTS

This research was supported by “Regional Innovation Strategy (RIS)” through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (MOE) (2021RIS-004), and the Soonchunhyang University Research Fund.

References

1. T. G. Ruttan, B. Grossman, A. Ferrero, V. Teppati and J. Martens, “Multiport VNA measurement,” IEEE Microwave Magazine, Vol. 9, No. 3, June 2008, pp. 56–69.
2. S. Azemzadeh, “A Guide to RF Switching Systems,” The RF Microwave Solutions Update, June 2008, pp. 34–44.
3. H. Heuermann, “Multiport S-parameter Calculation from Two-port Network Analyzer Measurements With or Without Switch Matrix,” 67th ARFTG Conference, San Francisco, Calif., U.S., 2006, pp. 219–222.
4. Y. Jang, C. I. Beak, D. M. Kim, D. Ahn and S.-H. Son, “A Sub-6 GHz Switching System for Simultaneous Use of 2-Port Network Analyzer by 4 Users,” International Journal of RF and Microwave Computer-Aided Engineering, Vol. 2023, Article ID 8827119, 2023, 7 pages.
5. A. Rumiantsev and N. Ridler, “VNA Calibration,” IEEE Microwave Magazine, Vol. 9, No. 3, June 2008, pp. 86–99.
6. J. P. Dunsmore, “Handbook of Microwave Component Measurements: With Advanced VNA Techniques,” John Wiley & Sons, 2020.
7. A. C. Bovik, “The Essential Guide to Image Processing,” Academic Press, 2009.