In RF circuit design, we often need to split one signal into two or more paths, and the component that does this is called a power divider, or more formally, a microwave power divider. Today, we are going to share the simplest type of power divider, the T-type power divider, as shown in the image below.

In this T-type power divider, the signal enters through port 1 and at the T-junction, the signal splits into two paths, exiting through ports 2 and 3. Based on the ratio of signal distribution, power dividers can be classified into equal and unequal types. Of course, if it can split one signal into two, it can also split it into more paths—even resembling a porcupine, if you like.

When implemented with transmission lines, a T-type power divider can be modeled as a junction of three transmission lines, such as the following three examples: E-plane waveguide T-junction, H-plane waveguide T-junction, and microstrip T-junction. At the junction, discontinuities may excite spurious fields or higher-order modes. If these discontinuities cannot be ignored, an equivalent susceptance B can be used to estimate the stored energy.

The transmission line model can be equivalently represented as:

Assume the characteristic impedance of the input transmission line is Z0, and the impedances at the output ends are Z1 and Z2, respectively. To minimize the reflected signal at the input, the equivalent impedance seen from port 1 needs to be matched. That is, the parallel impedance of Z1 and Z2 should equal the characteristic impedance Z0 at port 1.

This condition must be met to avoid reflection at the T-junction:

The power distribution ratio between ports 2 and 3 can be determined by the ratio Z2:Z1. For instance, in an equal-split 1-to-2 power divider, Z2:Z1=1, so Z1=Z2=2*Z0. If the input impedance is 50 Ohms, then Z1=Z2=100 Ohms. In this case, a quarter-wave impedance transformer is typically used to transform the impedance Z1 and Z2 to the desired value, such as 50 Ohms. Note that the two output ports of this divider are not isolated, and the impedance seen from the output ports is unmatched.

For example, a 2:1 power divider with an input impedance of 50 Ohms. If the impedance at port 2 is Z1=150 Ohms and at port 3 is Z2=75 Ohms, the parallel impedance seen from port 1 is 50 Ohms, meaning that port 1 is matched. However, the parallel impedance seen from port 2 is 30 Ohms, and from port 3 is 37.5 Ohms. These two ports are unmatched, with reflection coefficients of:

– Return loss at port 2: RL=3.5dB

– Return loss at port 3: RL=9.5dB

Using a T-type divider for power distribution might be acceptable in many cases, but using it for power combining can result in significant reflection.

Below is an example of a compact-size ultra-wideband power divider for reference:

The design includes a circuit board and a power division circuit placed on the board. The power division circuit has an input signal transmission segment connected to the input port and two output signal transmission segments connected to the output ports. The input and output signal transmission segments are electrically connected through a circuit node, with the signal transmission path lengths of the input and output segments being 1/6λ—1/9λ. Additionally, a matching segment with a signal transmission path length of 1/3λ—1/6λ is connected at the circuit node between the input and output segments to match the line impedance. This design reduces the size of the power divider, making it suitable for a wider range of environments. It also overcomes the issues of parasitic capacitance and inductance effects and impedance mismatch caused by shortened transmission branch lengths, ensuring the performance of the power divider.