Introduction and Application of Vector Network Analyzer

RF transformers are passive devices that can achieve impedance, voltage, and current transformation, and have functions such as DC isolation, common mode rejection, and single-ended to differential (or unbalanced to balanced) conversion. Therefore, they are widely used in RF circuits such as push-pull amplifiers, double-balanced mixers, and ADC ICs. RF transformers have a certain impedance transformation ratio, and their single-ended impedance is often not 50 Ohm, which makes it difficult to test their performance parameters.

In view of the limitations of the traditional back-to-back test method, this article introduces a test method based on the vector network analyzer R&S ZNB. This method uses the virtual differential test mode and port extension (Offset) function to test the insertion loss, return loss, common mode rejection ratio and other performance indicators of the RF transformer without changing the test device.

Text
RF transformers have the functions of impedance transformation, DC isolation, common mode interference suppression, and balanced and unbalanced conversion, and are widely used in various RF circuits. RF transformers are generally made of two or more mutually insulated copper wires wound on a magnetic core, and power is transmitted from the primary coil to the secondary coil through electromagnetic coupling. Figure 1 shows the equivalent circuit of the RF transformer. Assuming that the number of turns of the primary coil is N1 and the number of turns of the secondary coil is N2, the following relationship is satisfied:

N2 / N1 = n, V2 = n × V1, I1 = n × I2 (Equation 1)

The input and output impedance transformation ratio is: Zout / Zin = (N2 / N1)2 (Formula 2)

Figure 1. RF transformer equivalent circuit

How to test the performance of RF transformer?

Most RF transformers can achieve unbalanced to balanced conversion and can be used as a balun, as shown in Figure 1. The left side is a single-ended form and the right side is a differential form. Typical test parameters include: insertion loss, return loss, CMRR (common mode rejection ratio), amplitude and phase imbalance characteristics, etc.

RF transformers with a single-ended impedance of 50 Ohm and a differential impedance of 100 Ohm can be tested directly in the virtual differential test mode of the vector network analyzer R&S ZNB, because by default, the single-ended impedance and differential impedance of the ZNB in ​​virtual differential mode match the RF transformer to be tested.

But for RF transformers whose single-ended impedance is not 50 Ohm, how can we effectively test their performance?

If the single-ended impedance of the RF transformer is not 50 Ohm, the port matching between the transformer and the vector network needs to be considered. The traditional test method is to directly use two identical RF transformers in a back-to-back arrangement to achieve impedance matching, as shown in Figure 2. Half of the measured loss is the insertion loss of a single transformer. This method can test the insertion loss of the transformer and the return loss of the single-ended port, but the test of the common mode rejection ratio CMRR, amplitude and phase imbalance characteristics is extremely complicated and requires the replacement of the test device.

Figure 2. Traditional back-to-back RF transformer testing

Alternatively, use the impedance transformer shown in Figure 3, use two resistors to build a Mini-Loss Matching PAD, and treat the transformer as a three-port single-ended device. If the differential impedance of the balanced end is 200 Ohm, the corresponding single-ended impedance is 100 Ohm. The values ​​of resistors R1 and R2 must ensure that the input impedance from the transformer output to the vector network is 100 Ohm, and the input impedance from the vector network to the transformer is 50 Ohm. Equation 3 gives the calculation formulas for R1 and R2. Figure 4 shows a schematic diagram of the test device based on a four-port vector network, using the UOSM calibration method. The through calibration between Port1 and Port2 and Port4 also requires connecting an impedance transformation network to achieve matching between ports.

图3. Mini-Loss Matching PAD

(Formula 3)

Figure 4. Test setup using an impedance transformer

After the calibration was completed, an RF transformer was tested with a nominal operating frequency of 600MHz. The measured insertion loss and return loss are shown in Figure 5. In the low frequency band, the test results are consistent with the specifications, but as the frequency increases, the measured results deviate more and more from the specifications. The experiment found that this is because the frequency characteristics of the resistor used in the impedance transformer are poor, and the resistance value changes greatly with the increase of frequency, which limits the application of this method at high frequencies.

Figure 5. Test results using an impedance transformer

The vector network analyzer R&S ZNB supports changing the port reference impedance. Under certain conditions, this allows testing S parameters under non-50 Ohm system impedance. The test process is: first test the S parameters under 50 Ohm system impedance, and then transform the test data accordingly according to the set port reference impedance to obtain the S parameters corresponding to other system impedances. In this way, there is no need to use an external impedance converter, making the test more convenient and flexible.

For RF transformers, the output is in the form of differential pairs. When designing the test evaluation board, the impedance and line spacing of the PCB traces should be arranged according to certain rules to reduce the impact on the test results. However, in practice, this is often difficult to meet. For this reason, after the calibration is completed, it is necessary to perform the port extension function to extend the calibration reference plane to the transformer pin. This is very important, especially for the differential end, because the evaluation board traces are generally impedance controlled according to 50 Ohm, and the single-ended impedance of the differential output end of the RF transformer is often not 50 Ohm. If the port extension function is not performed, accurate performance will not be measured.

Figure 6. Schematic diagram of vector network port extension

Taking the R&S four-port vector network ZNB as an example, the effectiveness of this method is verified by measuring two RF transformers.

1# RF transformer parameters are as follows:

Input single-ended impedance: 50 Ohm

Impedance transformation ratio: 1:4

Frequency range: 0.5MHz~600MHz

In-band insertion loss (Spec.): ≤ 3dB

Test steps:

① First, set the frequency range and perform system error calibration. At this time, the default 50 Ohm port reference impedance can be used.

② Then perform the port extension function, especially for the differential port; if the single-ended impedance of the DUT input side is not 50 Ohm, it is recommended to perform port extension on the single-ended port as well; 

③ Finally, enter the virtual differential test mode and input the differential mode and common mode impedance according to the actual impedance value of the transformer.

Figure 7. R&S ZNB virtual differential test mode

Figure 8. Setting differential mode and common mode impedance

Figures 9 and 10 show the insertion loss, return loss, and amplitude and phase imbalance test results of the RF transformer under test, respectively. The insertion loss meets the specification indicators in the full frequency band, but the amplitude and phase imbalance characteristics are poor at high frequencies, which will affect the ability to suppress common-mode interference signals. For the RF transformer with a center tap on the secondary side as shown in Figure 1, it is generally recommended to ground the center tap to improve the amplitude and phase imbalance characteristics.

Figure 9. 1# Transformer Insertion Loss and Return Loss Test Results

Figure 10. Test results of the amplitude and phase imbalance characteristics of transformer 1#

Figure 11. 2# Transformer Insertion Loss and Return Loss Test Results

Parameters of 2# RF transformer are as follows:
        Input single-ended impedance: 50 Ohm
        Impedance transformation ratio: 1:1
        Frequency range: 0.4MHz~500MHz
        In-band insertion loss (Spec.): ≤ 3dB

According to the test steps described above, after calibration, port extension, and setting the differential mode and common mode impedance to 50 Ohm and 12.5 Ohm respectively, the test results are shown in Figures 11 and 12. The insertion loss meets the specification indicators, and the amplitude and phase imbalance characteristics are relatively good. Figure 13 shows the test results of the common mode rejection ratio CMRR, which is the result obtained using the Trace Math function of the vector network. Now R&S ZNB has supported the direct display of CMRR test results, making the test more convenient.

Figure 12. Test results of the amplitude and phase imbalance characteristics of the 2# transformer

Figure 13. CMRR test results of 2# transformer

The above two test examples show that for the test of RF transformers with single-ended impedance other than 50 Ohm, compared with the traditional back-to-back test method and impedance transformer test method, the virtual differential test mode and port extension functions of the vector network analyzer R&S ZNB will be more convenient. Without changing the test device, the insertion loss, return loss, amplitude and phase imbalance characteristics, and common mode rejection ratio CMRR of the transformer can be directly tested, which greatly simplifies the test of RF transformers.

Conclusion
This paper introduces the RF transformer test method based on Vector Network R&S ZNB. This method not only avoids the drawbacks of the traditional back-to-back test method and impedance transformer test method by adopting virtual differential test mode and port extension functions, but also effectively avoids the impact of the unsatisfactory wiring of the evaluation test board on the test results. It can comprehensively and effectively evaluate the performance of 50 Ohm and non-50 Ohm single-ended impedance RF transformers, making the test of RF transformers more flexible and convenient. Finally, the insertion loss, return loss and other indicators of two RF transformers with different impedance transformation ratios were tested in this paper. After comparison, they were consistent with their specification values, verifying the effectiveness of this method.