Ztarget— mΩ
first crossing—
worst |Z| peak—
Φ Z-PDN — target-impedance visualizer.
Build your decoupling rack, set the rail and the transient step, and watch |Zpdn(f)| cross or not cross the target line. The same closed-form math the PDN target-impedance guide teaches — runs entirely in your browser, no install, no signup.
Caps in the rack
| name | SRF | |Z|@SRF | qty |
|---|
What this is — and what it isn’t.
What it solves
Each cap is a series-RLC: Z = √(R² + (ωL − 1/ωC)²). The rack in parallel is summed as complex admittances, |Zpdn| = 1/|ΣYi|. The plane is treated as a fourth lumped RLC with parallel-plate C, low ESL, low ESR.
Target line: Ztarget = V·ripple/ΔI — the same formula the PDN article teaches. We sweep 1 kHz to 1 GHz across 200 log-spaced points.
What it ignores
- Mounting inductance from cap → via → plane. Real ESL is the datasheet value plus the layout mounting. Add ~0.2–0.5 nH per via.
- Frequency-dependent ESR (real caps trend up at high f). Constant-R is a good first approximation.
- Plane resonance peaks (cavity modes). We treat the plane as a single capacitance.
- VRM output impedance — we assume the regulator is the “low frequency floor” and the rack handles everything else.
When to use it
Use Φ Z-PDN to rapidly compare two rack designs — “does removing the 1 µF row break my margin?” — in a few seconds rather than a SPICE iteration.
For the real sign-off, run TRM (which couples PDN with thermal IR-drop) or a SPICE bench-correlated model. Φ Z-PDN is the funnel; TRM is the authority.
Sign off the rail, not just the rack.
TRM couples PDN and thermal so the IR-drop and the temperature meet on the same network. 14-day trial.