⚛️ quantum physics

Simple, accurate lumped-element models of distributed resonators for superconducting quantum circuits

This paper introduces a simple, open-source lumped-element modeling framework called `simpleLOMs` that accurately predicts the scattering parameters and Hamiltonian characteristics of distributed superconducting resonators up to strong coupling, thereby reducing the reliance on computationally intensive finite-element electromagnetic simulations.

Original authors: Elizabeth H. Kunz, Eli M. Levenson-Falk

Published 2026-07-23

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Elizabeth H. Kunz, Eli M. Levenson-Falk

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to build a tiny, invisible machine that can solve problems faster than any computer on Earth. This is the world of superconducting quantum circuits, where scientists use super-cold wires to create "qubits," the building blocks of quantum computers. To make these machines work, engineers have to design circuits that behave like perfect musical instruments, vibrating at very specific frequencies. The tricky part is that these circuits often include long, wavy wires called "distributed resonators." Think of these not as simple components like a single spring or a single weight, but as a long, flexible rope that can wiggle in complex ways.

To predict how these wiggly ropes will behave, engineers usually have to run massive, slow computer simulations that act like a virtual wind tunnel, calculating every tiny twist and turn of the electromagnetic waves. It's like trying to predict the weather by simulating every single air molecule in the atmosphere. While accurate, this method is so heavy and slow that it makes designing new quantum machines a nightmare. Scientists have tried to simplify things by treating these long ropes as if they were just simple, single springs (lumped elements), but this often leads to mistakes, especially when the ropes are connected to other heavy parts of the machine. The big question is: Can we find a way to describe these complex, wiggly ropes using simple, easy-to-understand building blocks without losing accuracy?

This paper by Elizabeth Kunz and Eli M. Levenson-Falk says "yes," and they've built a new tool to do it. Instead of trying to simulate the entire complex machine or ignoring the connections at the ends of the wires, the authors developed a clever shortcut. They created a method to turn a long, complicated transmission line (the "wiggly rope") into a simple, single "LC" circuit—a basic combination of an inductor and a capacitor—that acts exactly like the real thing, even when it's heavily loaded with other components.

The authors argue against the old ways of doing this. They show that the standard "analytical" formulas, which assume the wire is just floating in space or cut off at the ends, give inaccurate results when the wire is actually connected to other parts of the circuit. They also show that the popular "Black Box Quantization" method, which tries to replace the whole circuit with a black box, can be too complicated and hard to update when you add new parts. Their new approach, which they call "simple lumped-element models," is different because it looks at the wire and the specific capacitors attached to its ends all at once. By using a computer to simulate how the wire scatters signals (like how a mirror reflects light) and then tweaking a simple LC circuit until it matches that reflection perfectly, they create a model that is both simple and incredibly accurate.

In their simulations, the team tested this method on a coplanar waveguide resonator that was 7000 micrometers long with a characteristic impedance of 45.92 Ω and a bare resonant frequency of 8.58 GHz. They found that their new model could predict the resonant frequency and the "linewidth" (how sharp the note is) with an error of less than 0.5% for the frequency and less than 3% for the linewidth, even when the wire was connected to heavy loads. This is a big improvement over the old analytical methods, which struggled when the connections were strong. The paper also shows that this new model is better at predicting how the wire's frequency shifts when you attach other resonators to it, a crucial detail for building complex quantum networks.

The authors didn't just stop at the math; they packaged their method into an open-source code tool called "simpleLOMs" so other researchers can use it immediately. They demonstrated that this tool works not just for straight wires, but can be adapted for more complex shapes, like wires that twist or have connections in the middle. While the paper relies on simulations rather than physical experiments with a new chip, the results are robust and suggest that this simple approach can replace the heavy, slow simulations that currently slow down quantum circuit design. By turning a complex, distributed problem into a simple, lumped one, the authors have handed engineers a new set of blueprints that are easier to read, faster to use, and just as accurate as the old, complicated ones.

Technical Summary: Simple, Accurate Lumped-Element Models of Distributed Resonators for Superconducting Quantum Circuits

Problem Statement
The design of superconducting quantum circuits relies on accurately mapping physical design parameters to a quantum Hamiltonian. While designers traditionally employ computationally intensive finite-element electromagnetic (EM) simulations, effective circuit-level models are preferred for their ability to provide analytical expressions and clearer physical intuition. However, a significant barrier exists in modeling distributed elements, such as coplanar waveguide (CPW) resonators, which are ubiquitous in device designs.

Existing techniques for quantizing circuits with distributed elements include:

Lumped Oscillator Models (LOM): Treats a selected mode as an LC oscillator using analytical formulas (e.g., assuming open or shorted ends). This approach often yields inaccurate results when the resonator ends are strongly loaded by coupling capacitors or inductors, as the loading distorts the mode's spatial structure.
Energy Participation Ratio (EPR): Calculates the fraction of energy stored in nonlinear elements to generate an effective Hamiltonian. While accurate, this method often bypasses the creation of a lumped-element equivalent circuit, limiting generalizability.
Black Box Quantization (BBQ): Isolates nonlinear elements and models the remaining linear circuit as a "black box" with a specific impedance, often using Foster or Brune synthesis to create a lumped equivalent. Traditional BBQ approaches can be difficult to scale, may reduce intuition by replacing entire networks, and can struggle when new loads are added to existing models without reconstruction.

Methodology
The authors propose a rapid, straightforward procedure to create simple, accurate lumped-element models for distributed resonators within a broader two-port network. The core innovation lies in modeling the distributed element in full awareness of the lumped coupling elements attached to it, rather than treating them as separate or idealized boundary conditions.

The specific procedure involves:

System Definition: Consider a transmission line (e.g., CPW) with characteristic impedance $Z_C$ , length $L$ , and phase velocity $v$ , terminated at both ends by coupling capacitors ( $C_{c1}, C_{c2}$ ) and ground capacitances ( $C_{tog1,2}$ ).
Simulation: The transmission line with its explicit coupling capacitors is simulated to obtain scattering (S) parameters using the scikit-rf package.
Optimization: An equivalent single parallel LC circuit is constructed with the same coupling capacitors. The effective inductance ( $L_{eff}$ ) and capacitance ( $C_{eff}$ ) are varied to fit the S-parameters of the distributed transmission line as closely as possible within a frequency range of $f_0 \pm 5\kappa$ (where $f_0$ is the resonant frequency and $\kappa$ is the linewidth).
Fitting Algorithm: The fit uses Eq. (1) (standard analytical LOM) as an initial seed guess, followed by a least-squares regression on the real parts of $S_{11}$ and $S_{22}$ simultaneously.
Ground Capacitance Handling: Ground capacitances are absorbed into the effective capacitance $C_{eff}$ , as they add in parallel.

Unlike standard BBQ, which might replace the entire circuit including coupling capacitors, this method replaces only the transmission line with the effective LC circuit, leaving the coupling elements unchanged. This ensures the model remains self-consistent when distributed elements are added to a circuit without altering other components.

Key Results
The authors validate their "Optimized" model against traditional Analytical LOM and Foster synthesis approaches:

Scattering Parameter Accuracy: The optimized model reproduces the scattering parameters ( $S_{11}, S_{22}$ ) of the distributed CPW significantly more closely than analytical or Foster methods, particularly under heavy loading conditions (large coupling capacitances).
Resonance Characteristics: Using circle-fit methods on the complex plane, the optimized model demonstrates lower error in predicting the shifted resonant frequency ( $f_0$ ) and linewidth ( $\kappa$ ) across a sweep of coupling capacitances compared to other methods.
Coupling to Off-Resonant Loads: In a test where the distributed resonator is coupled to additional lumped resonators, the optimized model more accurately predicts the frequency shift of the central CPW resonator. The error is defined as $\epsilon = (\Delta f_{LC} - \Delta f_{CPW}) / \Delta f_{CPW}$ . The optimized approach outperforms Foster and analytical models in virtually all cases for the central resonator.
Load Resonator Shifts: The model does not consistently outperform other methods in predicting the frequency shifts of the load resonators themselves, as off-resonant shifts are relatively insensitive to small variations in $C_{eff}$ and $L_{eff}$ . However, the central CPW mode remains highly sensitive to the model choice, where the optimized method excels.
Validation against EM Simulation: The method was cross-validated against ANSYS HFSS driven modal simulations of a straight CPW. The optimized lumped model matched the EM simulation results to within 0.5% for shifted frequency and 3% for linewidth.

Significance and Claims
The paper claims that this technique provides a modular, accurate, and intuitive alternative to full EM simulations and existing lumped-element approximations. Key advantages highlighted include:

Accuracy under Loading: The method maintains accuracy even under heavy loading, a regime where traditional LOM and single-pole Foster synthesis fail.
Modularity: Unlike traditional BBQ, which may require reconstructing the entire circuit model when components change, this approach allows for the replacement of a single distributed element with an effective LC model without altering the values of nearby lumped coupling elements.
Port Independence: The method does not require defining a single reference port; it generates the same effective circuit whether measured via $S_{11}$ or $S_{22}$ .
Accessibility: The authors provide an open-source code package, simpleLOMs, which implements this procedure. The code is modular, allowing any 2-port microwave network to substitute the transmission line, and supports cascading models across networks of multiple distributed resonators.

The authors conclude that simple lumped models, when constructed with this specific awareness of coupling loads, are a viable approach for accurate circuit-level simulation of superconducting devices, reducing the reliance on computationally expensive finite-element simulations while preserving physical intuition.

Technical Summary: Simple, Accurate Lumped-Element Models of Distributed Resonators for Superconducting Quantum Circuits

More like this