About the Juice and Education category

Hi! :waving_hand: Use this category to discuss Juice and its use in education.

Topics can include notebooks, examples of how Juice is used in courses, tips and best practices, and any questions you may have.

Discussions about extensions, tools, or frameworks built on top of Juice that are used for teaching, training, or course delivery are also welcome.

When working on a multiqubit device:

“quantum_device.draw()
plt.show()”

We can initialize the qubit by resonator spectroscopy, similar to tutorial 2, hence qubit is there. But Rabi frequency and further can not be found (terrible fit to the data (see attached)). It seems that q0, q1, q2 can not be driven by Rabi, or do we make a mistake here? It also seems there is no response on the flux, hence fit failed?

Hi Tjeerd, Thanks for your post!

I have a couple questions before narrowing down the problem.

  • Have you run the calibration of the reference magnitude (or in other words, the power) for the readout line already? It is essential to set the readout power set such that it is in dispersive regime where we can perform qubit-state dependent dispersive readout.
  • Secondly, have you performed a qubit spectroscopy measurement to calibrate the drive frequency of the qubit? The data shows a drift across the sweep indicating that the drive is not in resonance with the qubit transition frequency.

Hi,

Here what we could run so far up to not being able to obtain a parabolic fit anymore for the response on the flux:

3qubit_star.zip (718.2 KB)

Hi Tjeerd, thanks for sharing your notebook.

I noticed one missing piece of action after the measurement of the resonator power scan. After the measurement has finished running, the power that is extracted from the fit for the dispersive regime needs to be set for the qubit. Only then the readout operation can be performed in the dispersive regime. This can be done by running the following lines of code after running the resonator power scan measurement:

q0.spec_ro.reference_magnitude.dBm(analysis.quantities_of_interest['critical_reference_power'])
q0.measure.reference_magnitude.dBm(analysis.quantities_of_interest['critical_reference_power'])

In the next experiment where you attempt to measure the flux dependence of the resonator frequency, what happens is that you apply a current on the SQUID loop of the qubit that the resonator is coupled to. The qubit transition frequency gets shifted based on the flux current in the SQUID loop. The resonator frequency in-turn also gets shifted as it is coupled with the qubit.

And secondly, on your question about the parabolic fit. The flux quantum (I_\phi) is the period of the arc, and hence can’t be obtained from the parabolic fit as it has only a vertex position, a vertex height and a curvature. Perforning the scan over a small range of flux bias currents is useful for accurately estimating the current and resonator frequency at the sweetspot (vertex) using the parabolic fit. If you wish to characterize the flux period, then the best way would be to

  1. Widen the current sweep to cover at least one full period, ideally more.
  2. Pass flux_arc_model="sine".

Hi,

Added the both lines and broadened the linspace on the currents as suggested. Getting this parabolic curve, which now seems reasonable, but still the message “unable to determine with parabolic fit” and indeed solved the problem, now getting clear Rabi.

Thx

1 Like