A parallel LC circuit (an inductor and capacitor wired in parallel) forms a resonant "tank" that exhibits maximum impedance at its resonant frequency. Unlike series LC circuits that pass their resonant frequency, parallel tanks reject it when placed in the signal path, or pass it when used as a shunt to ground. This topology is the backbone of RF oscillators, IF (intermediate frequency) filters, and impedance matching networks.
This guide skips the abstract theory and goes straight to the bench: how to calculate real values, how component drift destroys your Q factor, and exactly what to buy for your next build.
The Parallel LC Topology: Nodes, Currents, and the Tank Concept
In a parallel LC topology, the inductor (L) and capacitor (C) share the exact same voltage. They connect at two common junctions: Node A (the top rail/signal input) and Node B (the bottom rail/ground).
When an AC signal hits Node A, current divides between the two branches. Below resonance, the inductor's low reactance dominates, pulling current through L. Above resonance, the capacitor's dropping reactance dominates, pulling current through C. Exactly at the resonant frequency ($f_r$), the reactive currents through L and C are equal in magnitude but 180 degrees out of phase. They cancel each other out at the input nodes, causing the circuit to draw near-zero current from the source. The energy simply sloshes back and forth between the magnetic field of the inductor and the electric field of the capacitor—hence the term "tank circuit."
Why Parallel Over Series? Choose a parallel LC when you need a band-stop (notch) filter in a series signal path, or a band-pass filter when shunted to ground. Choose a series LC when you need a band-pass filter in a series signal path. Parallel tanks are heavily favored in RF oscillator feedback networks because their high impedance at resonance maximizes voltage gain.
Design Walkthrough: Sizing Real Components for a 1 MHz Tank
Let's design a parallel tank tuned to exactly 1.000 MHz. The governing formula is:
$$f_r = \frac{1}{2\pi\sqrt{LC}}$$
Step 1: Pick the Capacitor First. Capacitors are available in much tighter tolerances (1% or 2%) and lower temperature coefficients than inductors. We will select a 250 pF silver mica capacitor (e.g., Cornell Dubilier CD15CD250JO3, roughly $1.50 on Mouser). Silver mica offers excellent stability and low ESR (Equivalent Series Resistance).
Step 2: Calculate the Required Inductance. Rearranging the formula to solve for L:
$$L = \frac{1}{(2\pi f_r)^2 C}$$
$$L = \frac{1}{(2\pi \times 1,000,000)^2 \times 250 \times 10^{-12}}$$
$$L \approx 101.32 \text{ \mu H}$$
Step 3: Select a Real Inductor. Inductors rarely come in 101.32 \mu H. We will select a standard 100 \mu H RF inductor. A solid choice is the Coilcraft RFS-1317-101KL (100 \mu H, 5% tolerance, high Q). It costs about $2.00.
Step 4: Verify Actual Resonance. Plugging 100 \mu H and 250 pF back into the original formula yields an actual $f_r$ of 1.006 MHz. For a standard 1 MHz IF or oscillator stage, this 0.6% deviation is well within acceptable limits and can be trimmed with a small parallel variable capacitor if absolute precision is required.
Behavior Matrix: How Component Drift Alters Resonance
Real-world components drift with temperature and age. Here is how parameter shifts affect your parallel LC circuit's behavior. For high-Q RF designs, always specify C0G/NP0 dielectrics for capacitors to avoid the massive capacitance shifts seen in X7R or Y5V ceramics under bias or temperature changes.
| Parameter Changed | Direction of Change | Effect on Resonant Freq ($f_r$) | Effect on Q Factor / Bandwidth |
|---|---|---|---|
| Capacitance (C) | Increases (e.g., thermal drift) | Decreases ($f_r$ shifts lower) | Unchanged (assuming ESR is stable) |
| Capacitance (C) | Decreases | Increases ($f_r$ shifts higher) | Unchanged |
| Inductance (L) | Increases (e.g., core saturation) | Decreases ($f_r$ shifts lower) | Drops severely (core losses increase) |
| Inductor DCR (Parasitic) | Increases (e.g., winding heat) | Negligible shift | Q drops, bandwidth widens |
| Capacitor ESR (Parasitic) | Increases | Negligible shift | Q drops, peak impedance lowers |
Breadboard Testing: Step-by-Step Verification
Do not trust the datasheet values blindly. Parasitic breadboard capacitance (typically 2-5 pF per node) will pull your high-frequency resonance down. Here is how to measure the actual tank response on the bench.
- Isolate the Source: Connect your function generator (e.g., Siglent SDG1032X) to Node A through a 1 k\Omega series resistor. This converts your low-impedance voltage source into a high-impedance current source, preventing the generator's 50\Omega output impedance from loading down the tank and artificially destroying your Q factor.
- Wire the Tank: Place your L and C in parallel between Node A and Node B (ground). Keep leads as short as physically possible.
- Probe Node A: Connect your oscilloscope probe (e.g., Rigol DS1054Z) to Node A. Crucial: Use a 10x probe to reduce the probe's parasitic capacitance from ~100 pF down to ~10 pF.
- Sweep the Frequency: Set the function generator to output a 1 Vpp sine wave. Use the scope's built-in sweep or manually step the frequency from 500 kHz to 1.5 MHz.
- Find the Peak: The voltage at Node A will peak sharply at resonance. Record the exact frequency where $V_{peak}$ occurs. This is your true $f_r$.
- Calculate Q: Find the two frequencies where the voltage drops to 0.707 of the peak voltage ($f_1$ and $f_2$). Calculate bandwidth ($BW = f_2 - f_1$). Your measured Q is $f_r / BW$.
Failure Modes: What Breaks at the Extremes?
Understanding how a parallel LC fails compared to a series LC is critical for troubleshooting dead RF boards. A short or open in a parallel tank yields drastically different symptoms than in a series configuration.
| Fault Condition | Result in Parallel LC Circuit | Contrast: Result in Series LC Circuit |
|---|---|---|
| Capacitor Shorts | Dead short across Node A and B. Blows upstream fuse or pulls signal to ground. Tank action ceases. | Circuit becomes purely inductive. Passes high frequencies, blocks DC. No resonance. |
| Capacitor Opens | Circuit becomes purely inductive. Low impedance at high frequencies. Resonance destroyed. | Dead open. Signal path is completely broken. Zero current flows. |
| Inductor Shorts | Dead short across Node A and B (DC and AC). Signal is grounded. | Circuit becomes purely capacitive. Passes high frequencies, blocks DC. No resonance. |
| Inductor Opens | Circuit becomes purely capacitive. High impedance at low frequencies. Resonance destroyed. | Dead open. Signal path is completely broken. Zero current flows. |
Note on Inductor Shorts: In a parallel tank, an inductor shorting internally (winding insulation failure) is catastrophic because it directly shorts the AC signal node to ground. In RF power amplifiers, this often destroys the driving transistor.
Decision Tree: Selecting Your Inductor and Capacitor
Stop guessing component types. Follow this decision path to select the exact physical components for your parallel LC tank based on your target frequency and Q requirements.
- Is your target frequency > 10 MHz?
- Yes: You need minimal parasitic capacitance. Choose an air-core inductor (or very low permeability powdered iron) and a silver mica capacitor.
- Is your target frequency between 100 kHz and 10 MHz?
- Yes: Choose a ferrite-core RF inductor (shielded to prevent cross-talk) and a C0G/NP0 multilayer ceramic capacitor.
- Is your target frequency < 100 kHz (e.g., audio crossover or low-frequency switching)?
- Yes: Inductance values must be large. Choose a high-permeability ferrite or laminated iron core inductor and a polypropylene film capacitor for low dielectric absorption.
The Default Concrete Pick: If you are building a standard 455 kHz AM radio IF filter (the most common hobbyist parallel LC application), do not wind your own inductor. Buy the Toko 10RB-144 (a canned, shielded 455 kHz inductor with an integrated tuning slug, approx. $3.50). Pair it with a 1 nF C0G/NP0 ceramic capacitor (e.g., KEMET C315C102J1G5TA, approx. $0.15). This combination guarantees a high-Q, easily trimmable tank without the parasitic nightmares of hand-wound toroids on a breadboard.
For deeper reading on how inductor core materials affect Q factor at resonance, review the Coilcraft Q Factor application notes. For a comprehensive mathematical breakdown of tank circuit circulating currents, consult the All About Circuits AC theory chapter on parallel resonance.
