Impedance Inverter

The most common physical realization is a quarter-wave transmission line (λ/4). Its defining property is that the input impedance (Z_in) is inversely proportional to the load impedance (Z_L), scaled by the square of the line's characteristic impedance (Z_0):

• Formula: Z_in = Z_0² / Z_L
• Effect: A high-impedance load appears as a low impedance at the input, and vice versa.
• Doherty Role: This inversion is what allows the carrier amplifier to see a high impedance (for voltage saturation) when the peaking amplifier is off, and a low impedance (for current delivery) when the peaking amplifier injects current.

The characteristic impedance (Z_0) of the inverter is a critical design parameter that sets the load modulation ratio. It is not arbitrary; it is chosen based on the optimal impedances of the carrier amplifier.

• Standard Doherty: Z_0 is typically set to the optimal load impedance (R_opt) of the carrier amplifier at peak power.
• Asymmetric Doherty: Z_0 is scaled to accommodate different power ratios between the carrier and peaking amplifiers.
• Practical Constraint: The physical width of a microstrip line for a given Z_0 must be realizable on the chosen substrate, impacting power handling and loss.

A simple quarter-wave transmission line is inherently narrowband. The exact 90-degree electrical length and the required Z_in/Z_L inversion are only perfectly satisfied at a single center frequency.

• Bandwidth Limit: As frequency deviates, the electrical length is no longer exactly 90°, and the impedance transformation deviates from the ideal inverse, degrading the load modulation effect.
• Phase Dispersion: The phase shift through the inverter varies with frequency, causing misalignment between the carrier and peaking paths at the combiner.
• Mitigation: Broadband Doherty designs replace the single λ/4 line with multi-section matching networks or lumped-element equivalents to extend the bandwidth.

At lower frequencies or for MMIC implementations, a lumped-element pi- or T-network can synthesize the impedance inverter function, saving physical space.

• Topology: A low-pass pi-network (shunt C, series L, shunt C) or a high-pass T-network can provide a 90-degree phase shift and impedance inversion.
• Advantage: Significantly more compact than a distributed λ/4 line, especially below 3 GHz.
• Trade-off: Lumped components have finite quality factors (Q), introducing insertion loss that directly degrades the amplifier's overall power-added efficiency (PAE).

The impedance inverter is the physical mechanism that enables the active load-pull effect, the defining feature of a Doherty amplifier. It translates a change in current into a change in impedance.

• Low Power: The peaking amplifier is off (high impedance). The inverter presents a high impedance (2*R_opt) to the carrier, allowing it to reach voltage saturation at half its peak power, maximizing back-off efficiency.
• High Power: The peaking amplifier turns on and injects current into the common load. The inverter transforms this, causing the impedance seen by the carrier to dynamically decrease from 2*R_opt down to R_opt, enabling full current delivery at peak power.

The impedance inverter introduces a nominal 90-degree phase shift at the output of the carrier amplifier. To ensure the carrier and peaking amplifier signals combine in-phase at the output, an identical phase shift must be added to the peaking amplifier's input path.

• Input Splitter: A 90-degree hybrid coupler or a dedicated 50-ohm quarter-wave line is placed before the peaking amplifier's input matching network.
• Misalignment Consequence: Without this compensation, the two amplified signals would arrive at the Doherty combiner out of phase, resulting in destructive interference, severe gain reduction, and catastrophic efficiency collapse.

The most common physical realization of an impedance inverter. A transmission line with an electrical length of 90 degrees (λ/4) at the center frequency. Its defining property is that the characteristic impedance (Z₀) is the geometric mean of the source and load impedances, mathematically expressed as Z₀ = √(Z_in * Z_L). This structure inverts the load impedance: a high load impedance appears as a low impedance at the input, and vice versa, enabling the active load-pull effect.

The output network that integrates the impedance inverter to merge the outputs of the carrier and peaking amplifiers. It performs the critical impedance transformation that allows the carrier amplifier to see a modulated load impedance. At low power, the peaking amplifier is off, and the carrier sees a high impedance for maximum efficiency. At peak power, both amplifiers drive a matched load for maximum output power.

The dynamic mechanism by which the impedance inverter enables load modulation. The peaking amplifier injects a current into the combining node that is proportional to the signal envelope. Through the impedance inverter's transformation, this current injection actively pulls the load impedance seen by the carrier amplifier. This is distinct from passive load-pull measurement systems and is the fundamental operating principle of the Doherty architecture.

The process of dynamically varying the effective load impedance presented to a transistor as a function of the instantaneous signal envelope. In a Doherty amplifier, the impedance inverter translates the peaking amplifier's current into a varying impedance at the carrier's output. This ensures the carrier operates near saturation—and thus at peak efficiency—over a wide range of output power back-off levels, rather than only at the peak.

An advanced Doherty topology where individual matching networks are placed directly after the carrier and peaking transistors, before the impedance inverter and combiner. This architecture decouples the fundamental matching from the impedance inversion, enabling broadband Doherty designs. The impedance inverter can then be optimized solely for bandwidth, while the post-matching networks handle harmonic terminations and optimal fundamental impedance presentation.

The ratio of voltage to current for a wave traveling on a transmission line, a critical design parameter for the impedance inverter. For a quarter-wave transformer to function as an ideal inverter, its Z₀ must be precisely calculated as the geometric mean of the impedances to be matched. In practical Doherty designs, this value is often non-standard (e.g., 35Ω or 70Ω), requiring specialized microstrip or stripline synthesis to achieve the exact impedance transformation ratio.