Authors' Reply.

The authors of the original above-named article [ibid., vol. 60, no. 9, pp. 2791?2798, Sep. 2012] appreciate the comments by M.-J. Park [ibid., vol. 61, 2013]. The authors agree that their proposed balanced-to-balanced power divider/combiner structure can be simplified into the form proposed in the comments of Park when only the circuit performance at the central frequency is considered. The main reason is that the half-wavelength transmission line can be regarded as an ideal phase inverter with special properties, including those described by Park [see (1)?(4)]. However, the two structures have different characteristics at other frequencies, and it should also be pointed out that the curves in Park's Fig. 3 cannot completely describe the frequency responses of the structure since it is not symmetric to its middle plane any more. From a theoretical analysis and simulations presented, it can be concluded that the balanced-to-balanced power dividers originally presented by Park and the authors and this letter are equivalent to each other at the central frequency due to the special properties of ideal phase inverters, and all of them can be converted from a two-way single-ended Wilkinson power divider. The authors' proposed power divider had four resistors and a symmetric configuration, which provides two differential-mode outputs with the same magnitude and the best bandwidth. The power dividers proposed by Park and this letter have fewer resistors, but their two outputs are not symmetric and the operating bands are also narrowed. Therefore, if symmetric performances and relatively large bandwidth are prior in design, the originally proposed four-resistor structure is preferred. When compact size and few resistors are desired, the type C one-resistor structure, given in Fig. 1(d) of this reply, is more suitable. [ABSTRACT FROM AUTHOR]