HP 5988-2923EN free download

File name 5988-2923EN.pdf

Agilent Balanced Measurement Example: Differential Amplifiers Application Note 1373-7 Introduction Agilent Technologies has developed a solution that allows the most accurate method available for measuring differential RF circuits. The approach uses Agilent N444x-series Balanced-Measurement Systems, which provide the ability to test multiport (single-ended) devices to the level of accuracy of traditional two-port systems. The systems can also test differential devices and show the performance under balanced (differential- and common-mode stimulus and response) operating conditions using mixed-mode S-parameters. Mixed-mode S-parameter analysis brings a great deal of insight to differential circuit design. The following figures provide a simple example. The sample circuit A, illustrated in Figure 1, consists of two gain channels that are perfectly symmetrical, and completely isolated from each other. Its multiport single-ended S-parameters are shown in Figure 2. The amplifier is perfectly matched on all ports, and the gain in each channel (S21 and S43) is about 8 dB. Since this is a differential amplifier, it is also interesting Figure 1: Sample circuit A to consider how the circuit behaves with differential and common-mode stimuli. This can be determined from the mixed-mode S-parameters shown in Figure 3. Again, both the input and output ports are matched for both differential and common-mode signals. The differential and common mode gains are also the same, about 8 dB of gain. Figure 2: Single-ended S-parameters of sample circuit A Figure 3: Mixed-mode S-parameters of sample circuit A 2 If the ideal ground is removed, the two channels will no longer be isolated. Figure 4 illustrates the new sample circuit B, schematically. In this example, each port is coupled to each other port when the circuit is viewed as a single-ended device. The impedance seen at each port is now 150 ohms, and the single-ended gain is only about 2 dB, as shown in Figure 5. By looking only at the single-ended S-parameters, it is natural to assume that this circuit is not as good as the previous one. This conclusion would be misleading, as the Figure 4: Sample circuit B, device without channel isolation mixed-mode S-parameters show in Figure 6. Because of the virtual ground that is present with a differential stimulus, the ports are matched, and the gain is Contrary to the indications of the single-ended data, again 8 dB. With a common-mode stimulus, the ports are this is actually a much better device than the previous open circuited and the device has no gain. Therefore, the example because it has a very high common-mode single-ended performance is reflecting a composite of the rejection (ratio of differential-to-common mode gain). differential-mode and the common-mode performance. This underscores the importance of representing the data in the