TWPA Tune-up with the SHFPPC

February 10, 2025 by Anas El Hamdi

In this blog post, we present a streamlined solution to assist researchers when tuning up and characterizing a Traveling-Wave Parametric Amplifier (TWPA), a critical tool for maximizing qubit readout fidelity.

Our approach leverages the full potential of these devices for advanced quantum applications, achieving high performance with minimal complexity. TWPAs amplify signals through either three-wave mixing or four-wave mixing processes. In a four-wave mixing process, two pump photons from a strong pump tone are converted into a signal and an idler photon through a nonlinear material. This process creates amplification through an energy transfer from the pump to the signal while adding only a minimal amount of noise. 

In order to operate the TWPA at its maximum performance, the properties of the pump tone need to be well controlled. The optimal control parameters are typically determined in a characterization step. Characterizing a TWPA presents challenges such as:

  • Achieving high pump parameter precision.
  • Minimizing tune-up times.
  • Selecting the optimal working point.
  • Configuring the setup and wiring for the probe tone to ensure accurate performance.

Beyond this tune-up step, we aim to operate the TWPA within a large quantum experiment alongside qubits in order to optimize their readout fidelities. This comes with additional challenges, namely:

  • Maintaining long-term power stability.
  • Minimizing frequency drift.
  • Addressing spurious pump tones that can compress amplifiers, impacting fidelity and reliability.

To address the challenges outlined above, our solution combines three key components: the SHFPPC Parametric Pump Controller, the SHFQA+ Quantum Analyzer, and the LabOne Q software platform. Together, these elements provide the following solutions:

  • The SHFPPC is our ready-made solution for characterizing TWPAs, featuring an integrated pump signal source and signal conditioning unit.
  • The SHFPPC stabilizes pump power levels through Automatic Level Control (ALC), ensuring precise pump parameters and reducing tune-up times.
  • It integrates pump tone cancellation techniques to reduce spurious tones and prevent saturation of the amplifiers in the readout line, enhancing overall system reliability.
  • The SHFQA+ aids in precise TWPA tune-up and ensures high-fidelity signal readout, thanks to an application driven feature set including multiplexed qubit readout and power spectrum measurements capabilities.
  • The combination of the SHFPPC and SHFQA+ enables rapid parameter sweeps, approaching the physical speed limit determined by the integration time of the readout signal required to achieve a specific SNR.
  • The LabOne Q software platform seamlessly integrates these components, offering a robust, scalable framework for efficient system control and adaptability to future needs.

This solution ensures precise TWPA characterization, reliable performance, and scalable system management. For more details and implementation guidance, a link to the full TWPA tune-up notebook is provided at the end of this blog post. The data presented here was collected at the ETHZ-PSI Quantum Computing Hub using a TWPA manufactured at MIT Lincoln Laboratory.

TWPA characterization setup

Figure 1: Schematic of the TWPA characterization setup, illustrating the connections between the SHFQA and SHFPPC instruments and the TWPA under test. 

In the setup, the SHFPPC’s Pump Output is routed through a directional coupler to the TWPA. The probe signal coming from the SHFQA+ Signal Output is amplified by both the TWPA and a HEMT amplifier. This signal then passes through the SHFPPC Signal Input and Output for pump tone cancellation before being sent to the SHFQA+ Signal Input. Additionally, the SHFQA's power spectral density (PSD) feature enables the characterization of noise rise and signal-to-noise ratio (SNR). 

In order to be able to accurately measure the response of the TWPA during the tune-up procedure, it is desirable to bypass any qubit chip in the readout line which might otherwise affect the signal. for this, the SHFPPC can generate a probe signal that is routed through the pump line instead.

The TWPA characterization process involves three main steps which we'll explain in the following: optimizing pump parameters, measuring the gain as function of probe frequency, and cancelling the residual pump tone in the readout return line.

Optimize Pump Parameters

The first step is to optimize the pump parameters to maximize the TWPA’s gain or SNR gain for a chosen probe signal. The probe frequency is set to 6.5 GHz, near the typical readout resonator frequency, while its power is kept at -35 dBm to prevent amplifier saturation. This optimization is achieved by sweeping the pump’s frequency and power while measuring the amplified probe signal. The SHFPPC enables application of a probe tone through the pump line, thus bypassing the qubit and ensuring the optimization process does not interfere with the qubit operation. The SHFPPC’s sweeper feature reduces measurement time, providing faster optimization and reliable results. The experiment is performed using the code below, which sets up an experimental workflow to scan pump parameters and executes it.

# Define the experiment workflow by configuring the experiment parameters
exp_workflow = scan_pump_parameters.experiment_workflow(
session=session,# Active session for running the experiment
qpu=qpu,# Quantum processing unit (QPU) to be used in the experiment
parametric_amplifier=twpa_to_measure[0],# Selecting the first parametric amplifier 
temporary_parameters=temporary_parameters,# Temporary parameters for the experiment
pump_frequency=frequencies,# List or range of pump frequencies to be tested
pump_power=powers,# List or range of pump power levels to be tested
options=options,# Additional options for configuring the experiment
)
# Execute the experiment workflow
workflow_results = exp_workflow.run()

Executing the code generates a 2D plot where the color map represents gain values from the TWPA. The optimal working point is identified by the pump power and frequency that yield maximum gain, highlighted by a green dot on the plot. This green dot marks the ideal operating conditions for peak gain. The corresponding pump parameters can be fed back into the tune-up procedure for the subsequent measurements. The optimal working point can also be chosen based on different metrics, such as the maximal SNR improvement.

TWPA gain landscape

Figure 2: Gain landscape. (a) Signal gain measured as a function of pump frequency and power at the SHFPPC output, with a total pump line attenuation of -75 dB, (b) A closer view of the optimal region of interest, corresponding to the area outlined by the green rectangle in (a), and (c) SNR improvement across this region of interest.

Gain Measurement at Different Probe Frequencies

After optimizing the pump settings for a single probe frequency, we measure the frequency response of the TWPA gain across the frequency band of the probe tone, showing how gain varies across different probe tone frequencies. The probe tone frequency is swept from the SHFQA output to generate a full gain curve, leveraging the SHFQA’s sweeper capabilities for optimal performance.

This code defines and executes an experimental workflow to measure the gain curve of a parametric amplifier, configuring it with both probe frequencies and pump powers.

# Define the experiment workflow for measuring the gain curve  
exp_workflow = measure_gain_curve.experiment_workflow(
session=session,# Active session for conducting the experiment  
qpu=qpu,# QPU used in the experiment  
parametric_amplifier=twpa_to_measure[0],# Selecting the first parametric amplifier   
temporary_parameters=temporary_parameters,# Temporary parameters for the experiment
probe_frequency=frequencies,# Frequency range of the probe signal 
pump_power=powers,# List or range of pump power levels   
options=options,# Additional configuration options for the experiment  
)
# Execute the experiment workflow
workflow_results = exp_workflow.run()

This generates a 1D plot of gain versus probe tone frequency, allowing us to assess performance at the qubit readout frequency and evaluate the amplifier's bandwidth, especially when using multiple readout frequencies.

Gain curve diagram

Figure 3: Signal gain as function of probe frequency at the optimal pump power and frequency indicated in Fig. 2

Pump Tone Cancellation

To enhance readout fidelity and reduce compression in the amplifier chain following the SHFPPC, the pump tone is suppressed from the TWPA return signal using an internal cancellation tone, eliminating the need for an external combiner. The phase and amplitude of this cancellation tone are swept, controlled via a variable phase shifter and variable attenuator within the SHFPPC, effectively minimizing the residual pump signal. No probe tone is applied during this measurement, as the focus is solely on reducing the residual pump.

This code defines and executes an experimental workflow to calibrate signal cancellation for a parametric amplifier using cancellation phase and attenuation values.

# Define the experiment workflow for calibrating pump tone cancellation  
exp_workflow = calibrate_cancellation.experiment_workflow(
session=session,# Active session for conducting the experiment  
qpu=qpu,# QPU used in the experiment  
parametric_amplifier=twpa_to_measure[0],# Selecting the first parametric amplifier 
temporary_parameters=temporary_parameters,# Temporary parameters for the experiment  
cancel_phase=cancel_phase,# Phase adjustment parameter for pump tone cancellation  
cancel_attenuation=cancel_attenuation,# Attenuation level for optimizing cancellation  
options=options,# Additional configuration options for the experiment  
)
# Execute the experiment workflow   
workflow_results = exp_workflow.run()

Executing the code generates a 2D plot showcasing the residual power levels of the pump tone after cancellation, mapped against different phase and attenuation settings of the cancellation tone. A green dot marks the optimal point where the residual pump signal is minimized, indicating the best cancellation performance. This method improves readout fidelity by efficiently canceling the pump tone, which reduces the risk of compression and enhances linearity in the amplification chain, allowing the room temperature amplifier to operate at a higher gain before the SHFQA input. This, in turn, prevents signal-to-noise ratio (SNR) degradation.

pump tone cancellation experiment

Figure 4: Pump tone cancellation measurement, the suppression of the residual signal power (in dB) as function of the cancellation circuit attenuation and relative phase shift.

Customizing Experiments with LabOne Q DSL

The experiments above should give you all you need to carry out your standard TWPA characterization. The LabOne Q DSL also provides the flexibility to customize the experiment further by allowing the addition of parameters such as a pulsed probe tone or the sweeping of additional variables. A basic use case of the DSL code below demonstrates how to sweep the pump power and frequency, and it can be easily adapted to other parameters or used for further experiment customization.


# Perform a nested sweep over pump power and pump frequency
# Outer sweep: Vary pump power
with sweep(uid="sweep_pump_power", parameter=pump_power_sweep):
   # Inner sweep: Vary pump frequency      
    with sweep(uid="sweep_pump_frequency", parameter=pump_frequency_sweep):  
        
        # Define the measurement section
        with section(uid="acquire"):  
            acquire(
                signal="acquire",  # Specify the acquisition signal  
                handle="phase_diagram",  # Handle for storing acquired data  
                length=1/acq["psd_bandwidth"], # Acquisition duration  
            )
        
        # Define the resting period between measurements
        with section(uid="rest", length=acq["repetition_delay"]):  
            reserve(signal="measure") # Reserve the measurement signal   
            reserve(signal="acquire") # Reserve the acquisition signal

In this example, the pump power and frequency are swept, and the acquisition process is handled within the defined sections.

Conclusion

After characterizing the TWPA, we use the updated TWPA parameters during operation with qubits, enhancing readout fidelity and ensuring a stable and reliable solution. Seamlessly integrating into our comprehensive qubit readout system, this TWPA solution is a valuable tool for both TWPA technology research and broader quantum applications. It provides all the necessary components for efficient qubit control and readout, supported by open-source software that enables fast and streamlined workflows. Whether you're focused on TWPA technology or leveraging TWPAs in quantum research, our complete solution is designed to meet your needs. For a detailed guide on TWPA tune-up, check out the full TWPA tune-up notebook in our GitHub repository. Curious to learn more? Get in touch.

Contact us