Circuit Simulator: DC, AC and Transient Analysis

The circuit simulator solves a SPICE-subset netlist with modified nodal analysis: DC operating point, DC sweep, small-signal AC response with a Bode plot, or transient run, plus a schematic canvas that converts both ways between drawing and netlist. The page opens on a resistive divider that already reads V(out) = 7.5 V.

Updated 2026-10-01Reviewed by CalcX Editorial Team6 sources5 min read

What the circuit simulator does

The component set covers resistors, capacitors, inductors, independent and controlled sources, diodes, BJTs, MOSFETs and an ideal op-amp; the readout gives node voltages and branch current and power per element. Filter and control design belongs to the Engineering Lab, signal spectra to the Science Lab.

How to use it

Choose a preset or paste a netlist.Eighteen presets cover dividers, steps, diodes, op-amp stages and a bridge; netlists accept R/C/L/V/I/E/G/H/F/D/Q/M/X elements with .op / .dc / .ac / .tran / .ic / .model directives and * comments. Suffixes: k, meg, m, u, n, p, f (M = milli; 1MEG = one megohm).
Pick the analysis tab.Operating Point, DC Sweep, AC Response, Transient or Schematic; results update as you type.
Fill the tab's fields.DC Sweep sets source and start/stop/step; AC Response sets dec/oct/lin, points and f start/stop with optional output and reference probes; Transient sets t stop, max step, integrator, LTE step control, UIC and Energy.
Read the result.Operating Point prints a node/branch table; DC Sweep and Transient plot the listed probes; AC Response draws magnitude and phase and names the −3 dB corner; Transient adds step counts.
Draw it if you prefer.In Schematic, place parts, wire with two clicks and name nets, then press Canvas → netlist; netlist → canvas rebuilds the drawing. Results can be published as a card, added to a report, or sent onward.

Worked readouts

Every number below is read from the tool.

Set up	Readout
Resistive divider preset: 10 V, 1 kΩ / 3 kΩ, Operating Point	V(in) 10V, V(out) 7.5V, I(V1) -0.0025A, P(V1) -0.025W, P(R2) 0.0187W
Inverting amplifier preset: 1 V, Rin 1 kΩ, Rf 10 kΩ, Operating Point	V(out) -10V, I(Rf) 0.001A, P(Rf) 0.01W, P(X1) -0.01W
RC low-pass: 1 kΩ, 159.1549 nF, AC Response, dec 40, 10 Hz–1 MHz	−3 dB corner ≈ 1.0001kHz, phase there -45°
RC step: 1 kΩ, 1 µF, 0→5 V pulse, max step 20 µs, 5 ms	steps 250, rejected 0, h ∈ [2.00e-5, 2.00e-5] s, Newton iterations 251
Lossless LC with `.ic VC(C1)=1`, Trapezoidal vs Backward Euler	energy drift -0.0000% versus -86.4529%

The divider solves in one iteration with zero residual; both resistors carry 0.0025A. The inverting stage shows the ideal op-amp's virtual short: node m reads 0V, the input current is 1 V / 1 kΩ = 0.001A, and X1 reports -0.01W. In the transient row τ = RC = 1 ms; only the trapezoidal LC run keeps its stored energy.

Limits

Teaching scale, explicit failure.: Up to 120 nodes, 400 elements, 400 unknowns and 60,000 transient steps; past that it returns E_TOO_LARGE instead of degrading quietly. Devices are first-order models (1N4148 / 1N4007 / 1N5817 / LED diodes, simplified BJTs, square-law MOSFETs) with no series resistance, breakdown, temperature sweep, noise, distortion, Monte Carlo, switches or transformers. A failed Newton solve returns E_NO_CONVERGE with the residual it reached; the last iterate is never passed off as an answer.
Some circuits have no DC solution.: Pure capacitive feedback, a capacitive divider, or any node reachable only through capacitors has no DC operating point. It points to uic with .ic initial conditions instead of inventing a value.
What it doesn't do.: It does not design filters, tune controllers, derive transfer functions, run FFTs or convert power units; those jobs belong to the Engineering Lab, the Science Lab and the power converter. It is also not production SPICE: check first-order stages with it, but do not sign off a silicon design.

Why the source current is negative

The current column follows the SPICE branch convention: a branch current flows into the element's + terminal. A source that delivers power therefore reads negative — the divider shows I(V1) = -0.0025A with P(V1) = -0.025W, and the inverting op-amp shows I(X1) = 0.001A with P(X1) = -0.01W. The power column keeps the passive convention: positive absorbs, negative supplies. Resistors read positive, and the signed columns balance. The negative is the direction, not a solver defect — do not flip it before comparing with a textbook.

Where it is useful

Learning the four analyses

Load a preset, change one component, watch the readout move. The divider teaches KCL arithmetic, the RC step ties τ = RC to the curve, the LC pair shows numerical damping between two integrators, and the diode sweep shows where a pn junction stops acting as a resistor.

Checking a stage before lab

The series RLC preset reads f₀ = 5.0329kHz, ζ = 0.3162, Q = 1.581 and 35.09% overshoot, and the transient plot shows how closely the run tracks that reference. A Sallen–Key low-pass shows its corner and roll-off; the rectifier shows the smoothing capacitor's effect on ripple.

Moving data onward

Operating-point tables and AC or transient waveforms can be sent onward as tables and series. Send waveforms to the Science Lab for spectra, coefficients to the Engineering Lab for transfer functions and controllers, and power values to the power converter.

References

C.-W. Ho, A. E. Ruehli and P. A. Brennan, The Modified Nodal Approach to Network Analysis, IEEE Transactions on Circuits and Systems — branch currents added to nodal analysis. ieeexplore.ieee.org （访问日期：2026-10-01）
Wikipedia, SPICE — netlist syntax and engineering suffixes, including M = milli and MEG = mega. en.wikipedia.org （访问日期：2026-10-01）
Wikipedia, Operational amplifier — ideal op-amp assumptions and the virtual short. en.wikipedia.org （访问日期：2026-10-01）
Wikipedia, Bode plot — magnitude in decibels, phase, and the −3 dB corner. en.wikipedia.org （访问日期：2026-10-01）
Wikipedia, Trapezoidal rule (differential equations) — A-stability of the trapezoidal method. en.wikipedia.org （访问日期：2026-10-01）
Wikipedia, Backward Euler method — L-stability and numerical damping of oscillatory systems. en.wikipedia.org （访问日期：2026-10-01）