SPITE 0.1.0

Scattering Parameter Inspection Tool for Engineers

"The secret ingredient is SPITE."

1. Introduction

SPITE is a lightweight single-file Python library for RF/microwave network analysis, built so undergraduate students can learn microwave engineering and run simulations outside of class or internships without needing expensive, license-gated software.

SPITE is built around the Network class, which represents RF networks using frequency data, scattering parameters, reference impedance, and schematic information. Python operators are used for 1-port and 2-port network composition: @ for cascading and ** for repeating networks

2. Features

• The entire library is in one file
• Conversion between S, Z, Y, and ABCD parameters in both directions
• Schematics for 1-port and 2-port networks
• Cascade 2-port and 1-port networks with @
• Repeat a 2-port network with **
• Linear, dB, and Smith chart plots for S, Z, and Y parameters
• Time-domain gating
• Touchstone file import and export
• Very important functions such as esqueleto() and merendola()

3. Limitations

Features that haven't been added yet

• Frequency dependent / complex reference impedance
• General N-port topology / circuit connection
• Components such as waveguides, microstrips, etc.
• Noise factor as a network parameter

Try using scikit-rf for more advanced RF engineering.

4. Installation

SPITE depends on NumPy and Matplotlib. Nothing else.

pip install spite

5. Examples

Quarter-wave matching network

import numpy as np
import spite as sp

f = np.linspace(0.1, 4e9, 100)
matching_network = sp.net2p_tline_series(f=f, f0=1e9, EL_deg=90, Zc=np.sqrt(50 * 75))                     @ sp.net1p_R(f=f, R=75)

matching_network.plot_schematic()
matching_network.plot_dB()

A quarter-wave transformer of Zc = sqrt(Z0 * R_L) matches a 75Ω load to a 50Ω system exactly at f0 — you can see it in the resulting S11 dip.

Lumped-element low-pass filter

import numpy as np
import spite as sp

f = np.linspace(0.1, 4e9, 100)
filt = sp.net2p_L_series(f, 3.9785e-9)        @ sp.net2p_C_shunt(f, 3.184e-12)        @ sp.net2p_L_series(f, 3.9785e-9)

S11_at_2GHz = filt.sample(2e9, param="S", port_indices=(1, 1))
print(20 * np.log10(np.abs(S11_at_2GHz)))

filt.plot_dB(floor=-30, port_indices=[(1, 1), (2, 1)])
filt.plot_smith(port_indices=[(1, 1), (2, 1)], chart_type="Y")

This exact filter was cross-checked against an AWR Microwave Office simulation of the same design. SPITE's predicted response matched to within 0.01 dB at the 2 GHz cutoff.

RLGC ladder

import numpy as np
import spite as sp

f = np.linspace(0.1, 4e9, 100)
R, L, G, C = 1.5, 5e-9, 3e-4, 2e-12

ladder = sp.net2p_RLGC(f, R, L, G, C) ** 3
ladder.plot_schematic()
ladder.plot_dB(port_indices=(2, 1))
ladder.plot_smith()

6. License and Citation

MIT License. Use it however you want.

But if you want to be nice, please put my name somewhere. I don't care about the format. SPITE is above citation formats because it is cool.