KI spirals and a TWPA line

A structure sheet for kinetic-inductance (KI) devices, not a measured chip. A travelling-wave parametric amplifier (TWPA) style line runs from one bond pad to the other: a finger-loaded section, a meandered section and a second finger-loaded section, all 10/6 µm CPW. The fingers (2 µm wide, reaching into both grounds) add shunt capacitance per unit length, as a KI TWPA’s artificial line does. Above it sit six spiral inductors (double-Fermat and Fermat spirals), the compact high-inductance elements KI circuits use.

No KI or TWPA chip script exists in Quantcore’s library. The cells here are ported building blocks from the legacy files spiral.lym, draw.py and cpwe.py (internal; the finger line is from G. Long, Glasgow 2021), laid out as a sheet on the PDK die with the device ID KI TWPA. Build it to see the spiral and finger-line components side by side and how a snapped chain of CPW cells forms one continuous line.

The whole ki_twpa_structures sheet rebuilt on the Quantcore PDK

At a glance

Notebook

examples/designs/legacy/ki_twpa_structures.ipynb (design JSON ki_twpa_structures.json)

Source

Legacy building blocks spiral.lym, draw.py, cpwe.py in Quantcore’s library (internal); no KI / TWPA chip script exists

Chip

7.5 × 7.5 mm Quantcore die, sapphire, CPW 10 / 6 µm

On the chip

2 CpweBondPad · 2 CpweFingerLine (150 fingers each) · 1 CpweMeanderLine · 6 SpiralLine

Builds today

yes, rebuilt from its JSON on 2026-11-19 in 2.2 s (the notebook itself was not re-run)

Pictures

Rebuilt ki_twpa_structures

The sheet: the TWPA-style line at the bottom, six spiral inductors in their square pockets above.

Zoom on the centre of a spiral

The centre of spiral S3, a double-Fermat inductor: the metal line (white, 3 µm wide) winds out of the middle (an 80 µm view).

Zoom on the finger-loaded line

The finger-loaded line: 2 µm fingers reach into the grounds on both sides of the centre line (an 80 µm view).

Zoom on the left bond pad

The left bond pad, an exponential taper into the 10 / 6 µm line.

LOM estimates

The notebook has no saved LOM output. The chip has no qubit and no resonator whose frequency the LOM estimates (it is a structure sheet).

Build spec

The rest of this page is the detailed spec the notebook was built from: every component, the changes made while rebuilding, and the notes.

Request: “import the kinetic-inductance and TWPA design chips to render, especially the spirals”. Notebook: examples/designs/legacy/ki_twpa_structures.ipynb (+ .json). Quantcore internal.

What the library holds

No KI or TWPA chip script exists in components/ (or the sibling project folders): no chip_* script, notebook or GDS uses spirals, KI lines or a TWPA. The building blocks:

legacy source

draws

Quantcore cell

components/pymacros/spiral.lym (library “MyLib”, PCell Spiral)

Archimedean spiral path inner_r → outer_r, pitch width + spacing per turn, n points per turn

SpiralPath

components/python/draw.py::spiral(x, y, w, start_ang, end_ang, spiral_r, spiral_type, npoints, layer)

spiral by angle: 0 Archimedean, 1 Fermat, 2 double Fermat in/out (the double-Fermat branch has a typo, y_bls / y1s)

SpiralLine (archimedean / fermat / double_fermat; golden XOR in tests/test_spirals.py)

components/python/cpwe.py::finger_line (G. Long, Glasgow 2021)

CPW with interdigital fingers into both grounds: the shunt-C-loaded artificial line of a KI TWPA

CpweFingerLine

cpwe.py::meander_line, b_pad, tapper, term_open

meandered CPW, bond pad, taper, open end

CpweMeanderLine, CpweBondPad, CpweTaper, CpweOpenEnd

ustrip.py, pymacros/Cong/bondpad_ustrip.lym

microstrip versions

not ported (microstrip is ignored)

The chip (Designer frame, 7.5 × 7.5 mm PDK die, ID KI TWPA)

instance

cell

params

at

PL

CpweBondPad

pad 500 / gap 300, end 10, pad 50, taper 500 (exponential, weight 25), CPW 10/6

(250, 2600)

F1, F2

CpweFingerLine

150 fingers, 2 µm wide, 10 µm long, 2 µm apart, CPW 10/6

snapped: PL.a → F1, M1.b → F2

M1

CpweMeanderLine

r 50, straight 600, bias 0.5, 18 periods, CPW 10/6

snapped to F1.b

PR

CpweBondPad(mirror=True)

as PL

snapped to F2.b

S1–S4

SpiralLine (double Fermat, pocket_margin=20; S5 Fermat, S6 double Fermat 8 turns, added)

legacy sizes kept where they map (spiral_r, width in µm, angles in rad); exact parameters and the 1100 µm pitch from x = 1000 µm: see the notebook layout cell

(1000 + 1100 k, see notebook)

PS1–PS4

metal:Rectangle(subtract=True)

square pocket 2 r_out + 40 µm

same centres

LOM

The line: one conductor PL → PR (50 Ω ports), finger sections with their shunt C (CpweFingerLine model). Spirals: shapes without ports; the notebook tabulates N turns, wire length, geometric L (modified Wheeler, circular) and kinetic L_k = L_□ ℓ / w for example sheet inductances (0.1 / 10 / 50 pH/□; the KI film is not in the PDK, so these are assumed values). No sim groups.

Open

  • The actual KI / TWPA chip sources (if they exist elsewhere): add them to components/ to rebuild them as chips.

  • Done: the draw.spiral Fermat / double-Fermat spirals are ported as SpiralLine, which the notebook uses.

  • The KI film (material, thickness, sheet kinetic inductance) for the PDK.