Using 50Ω laboratory equipment to test S-parameters of 75Ω circuit
Abstract: RF engineers working in cable, terrestrial, or satellite TV applications often need to measure S-parameters. Using the minimum loss network to convert the traditional 50Ω test port impedance to a 75Ω device provides an economical and convenient method for obtaining reasonable test results.
For most laboratory applications below 1GHz, the minimum loss pads constructed with 1% error 0402 or similar resistors and mounted on the PCB provide a quick and convenient way to test 75Ω circuits using 50Ω laboratory equipment . In most cases, the only factor that needs to be corrected is the MLP insertion loss-5.7dB plus the additional loss of the connector. Usually S-parameter testing does not require the use of complex calculations or even Smith charts.
RF engineers working on cable, terrestrial, or satellite TV applications often need to measure the S-parameters of the circuit. In the past, engineers used a vector network analyzer to verify that the TV tuner input provided the return loss they wanted. They would face the question: How can I use the 50Ω VNA to test the S-parameters of my 75Ω DUT? If there is sufficient funding, the answer is to build equipment specifically designed to test 75Ω circuits (75Ω source impedance and load impedance test ports). Otherwise, the minimum loss pad is used to convert the traditional 50Ω test port impedance to a 75Ω DUT, which provides an economical and convenient solution for obtaining reasonable test results.
When the IC manufacturer gives the return loss (| S11 |) of the cable TV LNA, this test result is referenced to 75Ω. That is, if | S11 | = -30dB (the reflected power is only 1/1000, which is actually quite a perfect match), when driven with a 75Ω source impedance, in fact the device output will allow all power to be transmitted to the LNA.
The same tuner no longer achieves good return loss when driven with 50Ω source impedance. Connect this perfectly matched tuner input directly to the VNA, | S11 | test results are close to -14dB, and the reflected power is now 1/25! Using the same 50Ω VNA, how can we prove that the TV tuner input performance is as we said That's good.
Therefore, a matching circuit is required. It should have a flat frequency response and the lowest insertion loss. The industry standard matching circuit is the "minimal loss pad" (aka "MLP"). The simple resistance network shown in Figure 1. The key performance of this network is to convert the 75Ω DUT load impedance to 50Ω for instrument measurement and the 50Ω instrument source impedance to the 75Ω impedance inside the DUT. This method eliminates the reflection, the response is flat, and the network loss value is easily obtained by subtracting the MLP loss from the DUT measurement value. Most test equipment manufacturers provide "minimal loss pads", and it is easy to build such a network on a laboratory platform when needed.
"Minimum loss" means that the network provides the lowest insertion loss under the same impedance transformation and possible resistance network configuration.
Figure 1. The minimum loss pad matches the 75Ω DUT to the 50Ω test port. Low frequency insertion loss is 5.72dB. The upper limit of the frequency response flatness is determined by the quality factor of the device.
The mathematical method for converting ZLOAD to ZLOAD 'is straightforward and is given in Appendix A. The derived ZLOAD 'expression describes the cascade impedance of (RSOURCE) MLP and DUT viewed from the test port. Transform the equation to get ZLOAD from ZLOAD '. This provides a method to eliminate the effect of MLP. The true ZLOAD value can be obtained from the test data at ZLOAD' Appendix B gives the algebraic operation as a reference, here provides the following results:
Integrity check calculation is reasonable. Suppose we measure the impedance of a 75Ω resistor through MLP. VNA will measure RLOAD '= 50Ω (infinite return loss), we expect math to tell us the result of a load resistance of 75Ω. Setting RLOAD '= 50Ω, we can get R1 = 43.3Ω and R2 = 86.6Ω, so that we get the desired ZLOAD = 75Ω.
It will become more useful to decompose this simple expression into real and imaginary parts and use spreadsheet calculations.
It will become more useful to decompose this simple expression into real and imaginary parts and use spreadsheet calculations. In general, the measurement of 75Ω DUT will be different from the impedance measurement-return loss (dB), reverse isolation, noise figure and input third-order intercept point are often measured. In these cases, it helps to draw the following conclusions about MLP:
(2) Conversely, when RLOAD is different from 75Ω, MLP no longer performs the correct impedance transformation, and the additional VSWR value between the test port and MLP and between MLP and DUT increases, resulting in measurement uncertainty. There are many tools to calculate the mismatch (expressed as VSWR) to measure uncertainty.
(3) MLP can be regarded as a purely resistive network within the specified frequency range, with an insertion loss of 5.7dB, plus the additional loss of connectors and cables.
(4) The path loss measured in S11 and S22 and the complete loss in the S21 or S12 measurement are twice the insertion loss (at least 11.4dB)-this reduces the effective sensitivity and dynamic range of the VNA.
In a practical example, for example, we want to measure the S21 parameters of four LNAs for cable / terrestrial broadcast TV LNAs such as the MAX3558. Place the DUT in the test setup, and place the MLP on both the input and output ports of the device, as shown in Figure 2. Calibrate VNA, excluding MLP. Connect port 1 to the 50Ω port of the MLP, and connect 75Ω to the LNA input. Configure the same for an output port and port 2 of the VNA.
Figure 2. Using a 50Ω device to test four MAX3558 cable / terrestrial TV LNAs
The VNA uses two least loss pads for impedance conversion and measures S21 (forward gain). The measured gain of the VNA is close to -5dB at 500MHz. After simply deducting the 11.5 or 12.0dB insertion loss of the two MLPs and the connector / adapter, the LNA provides 7dB power gain at 75Ω.
S21 (reverse isolation) measurement is not directly obtained. The isolation index of these LNAs is 65dB. After deducting the additional losses of the two MLPs, the VNA itself needs to resolve S12 at 77dB. If you are not careful, the power of the receiving port is too small for the VNA to measure accurately. The solution is to ensure that the receive port power is at least 10dB above the VNA noise floor / internal isolation floor. Without VNA isolation calibration, the noise floor is about -100dBm. We need to set the source impedance power to at least -20dBm, preferably -10dBm or even 0dBm. Port 1 receives sufficient power and makes measurements, adding 12 dB to the measured value to deduct insertion loss. That is, the measured value of -77dB becomes -65dB.
The MAX3558 evaluation board has pads on the PCB that allow engineers to add their own minimum loss pads. 50Ω SMA should be used instead of 75Ω F-type connector.
At frequencies higher than a few hundred MHz, MLPs mounted on PCBs and constructed with 0402 resistors pose measurement accuracy problems. The assumption that parasitic parameters will destroy the network is pure resistance. In this case, a more complicated method is needed to solve the problem. One of the methods is to completely describe the characteristics of MLP, using Smith chart to accurately eliminate the influence of the matching circuit. Another solution is to use an inductance-based transformer for low-loss impedance conversion. RF transformers are usually described by the impedance conversion ratio, rather than the turns ratio, such as "1.5: 1."
For most laboratory applications below 1GHz, the minimum loss pads mounted on a PCB and constructed with a 1% error 0402 packaged resistor provide a quick and convenient way to test 75Ω circuits using 50Ω equipment. In most cases, the only factor that needs to be corrected is the MLP loss -5.7dB plus the connector loss. There is no need for complex calculations and Smith charts for basic S-parameter measurements. When higher accuracy or a wider frequency range is required, and higher accuracy and a wider frequency range are required, product-tested high-quality MLPs are available from most manufacturers.
Appendix A: Derivation of ZLOAD from ZLOAD '
Appendix B: Derivation of ZLOAD from ZLOAD '
Similar articles can be found in the April 2004 issue of RF Design.
Rigid-flexible PCB board, FR4 TG150 + PI material with yellow coverlay and green soldermask.White silkscreen.2oz copper on double-side and the surface treatment is Immersion Gold,Total board thickness is 1.0mm. unit size is 12*37mm. Min.line width/line spacing is 0.2mm/0.2mm. Remote control equipment Circuit Board
We continue to introduce advanced equipment and improve our workers's skills. Now we are able to fully produce 1-12 layers flexible and rigid-flexible boards.
Rigid-Flexible PCB,Flexible Rigid-Flexible PCB,Rounded Rigid-Flex PCB,Rigid-Flex Printed Circuit Boards
Orilind Limited Company , https://www.orilind.com