Zero documentation¶
Note
Zero is still under construction, with the program structure and behaviour not stable. The program, and this documentation, may be altered in ways that break existing scripts at any time without notice.
Zero is a linear circuit simulation library and command line tool. It is able to perform small signal ac analysis on collections of components such as resistors, capacitors, inductors and op-amps to predict responses and noise.
Why Zero?¶
Given that tools such as LTspice and Qucs exist, why use this tool?
The answer is: it depends. For circuits where dc analysis is needed, or where you must model non-linear or time-variant effects, then the above tools are very useful; however, whilst component manufacturers often provide SPICE models to represent their parts, these often do not correctly model noise and output impedance. This is especially true for op-amps, at least historically. One of the key advantages of LISO, upon which Zero is loosely based, was that measured op-amp parameters were provided as standard in its library, available to be used in simulations to provide accurate gain and noise results. This feature has become incredibly useful to LISO’s users. LISO furthermore provided an optimiser to be able to use to tune circuit component values, something which is much trickier to do with SPICE.
Zero implements around half of what LISO is capable of doing, but extends it in a few ways to provide greater customisability and ease of post-processing. You can for example easily add new noise sources to components and simulate how they propagate through circuits without having to edit any core code, or implement your own types of analysis using the base Zero circuit objects. The results are also provided as a so-called solution which contains the simulation data as well as means to plot, export and further process it.
The other half of LISO is the so-called root mode, which includes a powerful optimiser. This is not part of Zero, but given that Zero exists within the Python ecosystem is is possible to use it some other optimisation packages such as scipy.optimize. LISO’s optimiser may one day be implemented in Zero.
What Zero does¶
Zero can perform small signal analyses on circuits containing linear components. It is inherently AC, and as such can compute frequency responses between nodes or components and noise spectral densities at nodes. Inputs and outputs can be specified in terms of voltage or current.
For more information, see the available AC analyses.
What Zero does not do¶
Zero’s scope is fairly limited to the problem of simple op-amp circuits in the frequency domain using linear time invariant (LTI) analysis. This means that the parameters of the components within the circuit cannot change over time, so for example the charging of a capacitor or the switching of a transistor cannot be simulated. This rules out certain simulations, such as those of switch-mode power supply circuits and power-on characteristics, and also effects such as distorsion, saturation and intermodulation that would appear in real circuits. Effects at dc such as op-amp input offset voltages and currents are also not modelled. Instead, the circuit is assumed to be at its operating point, and the circuit is linearised around zero, such that if the current through a component is reversed, the voltage drop across that component is also reversed. A small signal analysis is then performed to simulate the effect of the circuit’s output given small variations in its input. This is perfect for computing transfer functions and ac noise, but not for non-linear, time-varying effects. You should bear these points in mind before choosing to use Zero for more complete analyses.
For more information, see AC analyses.
LISO¶
Zero is loosely based on LISO by Gerhard Heinzel. It (mostly) understands LISO circuit mode input files, meaning that it can be used in place of LISO to simulate circuit signals. It also understands LISO output files, allowing results previously computed with LISO to be plotted, and for LISO results to be directly compared to those of this program.