The Capacitance in Series Equation: Topology and Node Behavior
When you wire capacitors end-to-end, you are building a series topology. Unlike resistors in series (which simply add up), the capacitance in series equation calculates the equivalent capacitance ($C_{eq}$) using the inverse sum of the individual values. For a basic two-capacitor string, the math is straightforward:
C_eq = (C1 × C2) / (C1 + C2)
For three or more capacitors, the generalized equation is 1/C_eq = 1/C1 + 1/C2 + 1/C3 + .... This inverse relationship means that adding capacitors in series always decreases the total capacitance. The physical reason is that you are effectively increasing the distance between the outermost plates while keeping the plate area constant, and capacitance is inversely proportional to plate separation.
Node Labels and Voltage Division
To analyze this on the bench, let's define our nodes. Assume a DC voltage source ($V_{in}$) applied across the string:
- Node A: The positive input terminal (connected to the top lead of C1).
- Node B: The junction point between C1 and C2.
- Node C: The ground/reference terminal (connected to the bottom lead of C2).
Because charge ($Q = C imes V$) must be identical on all series elements in a steady state, the voltage divides inversely proportional to the capacitance. The voltage across C1 is calculated as:
V_C1 = V_in × [C2 / (C1 + C2)]
Behavior Matrix: What Changes When an Element Shifts?
| Parameter Changed | Effect on Total Capacitance (C_eq) | Effect on Voltage Division (Node B) |
|---|---|---|
| Increase C1 value | C_eq increases (approaches value of C2) | Voltage across C1 drops; Node B voltage rises closer to V_in |
| Decrease C1 value | C_eq decreases (approaches new value of C1) | Voltage across C1 rises; Node B voltage drops closer to GND |
| Short Circuit C1 | C_eq becomes exactly C2 | Node B is pulled to V_in; C2 takes 100% of applied voltage |
| Open Circuit C1 | C_eq drops to 0 (acts as DC block) | Circuit broken; Node B floats or bleeds to GND via leakage |
Why Series Over Parallel? (And What Breaks at the Extremes)
If putting capacitors in series reduces your total capacitance, why do we do it? The primary reason is voltage stacking. When you need a capacitance value that must withstand a voltage higher than what is commercially available (or cost-effective), you wire lower-voltage units in series. A secondary use is in AC capacitive dropper circuits, where the series impedance limits current without dissipating heat like a resistor would.
However, series topologies introduce severe failure modes that parallel banks do not. According to standard circuit theory referenced by Electronics Tutorials, understanding these extremes is critical for safe design.
The Short-Circuit Extreme
If C1 fails as a short circuit (common in electrolytics subjected to overvoltage), Node B is instantly tied to Node A ($V_{in}$). C2 now sees 100% of the supply voltage. If your supply is 400V and C2 is rated for 250V, C2 will rapidly overcurrent, vent, or explode. This cascading failure is why series strings require overvoltage protection or robust balancing networks.
The Open-Circuit Extreme
If C1 fails open, the DC path is broken. $C_{eq}$ becomes zero. In a DC smoothing application, your output ripple will spike to catastrophic levels. In an AC coupling application, the signal is simply lost. Unlike a short, an open failure usually doesn't cause a fire, but it silently kills circuit functionality.
Design Walkthrough: Building a 600V DC Snubber Bank
Let's build a practical circuit. We are designing an IGBT inverter snubber and need approximately 0.5µF of capacitance rated for a 600V DC bus. Finding a single 0.5µF 600V high-frequency film capacitor is expensive and physically massive. Instead, we will use the capacitance in series equation to build it from standard parts.
Component Selection
- Capacitors: Two 1µF, 350V DC Polypropylene Film Capacitors (e.g., KEMET R76 series or WIMA MKP10).
Math check: (1 × 1) / (1 + 1) = 0.5µF.
Voltage check: 350V + 350V = 700V total stack rating, safely derated for a 600V bus. - Balancing Resistors: Two 220kΩ, 1W Metal Film Resistors. Film capacitors have incredibly low leakage current, but minor manufacturing variances mean they won't share DC voltage perfectly. The resistors force Node B to sit at exactly half the bus voltage.
Breadboard and Bench Testing Steps
Before soldering this into a high-voltage PCB, prove the topology on the bench. As noted in All About Circuits' DC textbook, verifying voltage division at low voltage prevents high-voltage surprises.
- Wire the String: Place C1 and C2 in series on a high-voltage rated breadboard or terminal block. Node A is the free lead of C1, Node B is the junction, Node C is the free lead of C2.
- Add the Balancing Network: Connect the first 220kΩ resistor in parallel with C1 (Node A to Node B). Connect the second 220kΩ resistor in parallel with C2 (Node B to Node C).
- Low-Voltage Verification: Set your bench power supply to 50V DC. Connect it across Node A and Node C. Using a multimeter, measure Node B to Node C. You should read exactly 25V (±0.5V). If you read 40V, one of your resistors is open or miswired.
- Ramp and Monitor: Slowly ramp the supply to 200V. Verify Node B sits at 100V. Check the resistors for heat; at 200V, each resistor drops 100V, dissipating roughly 45mW (well within the 1W rating).
- Measure Parasitics: If you have an LCR meter, measure the total string from Node A to Node C at 100kHz. You should read ~0.5µF. Note that the Equivalent Series Resistance (ESR) of the string is the sum of both capacitors' ESR plus the physical wire resistance. Keep leads short to minimize parasitic inductance (ESL), which defeats the purpose of a high-frequency snubber.
Frequently Asked Questions
How do you calculate total capacitance in series with 3 or more capacitors?
Use the reciprocal sum formula: 1/C_total = 1/C1 + 1/C2 + 1/C3. For example, if you place three 30µF capacitors in series, the calculation is 1/30 + 1/30 + 1/30 = 3/30. Inverting that gives 30/3, resulting in a total capacitance of 10µF. A quick shortcut for identical capacitors in series is to simply divide the capacitance of one unit by the total number of units ($C/n$).
Does the capacitance in series equation change for AC vs DC circuits?
The fundamental equation for calculating the equivalent capacitance ($C_{eq}$) remains exactly the same for both AC and DC. However, the behavior of the circuit changes drastically. In DC, once the capacitors charge to their divided voltages, current stops flowing (acting as an open circuit). In AC, the series string presents a continuous capacitive reactance ($X_c = 1 / (2\pi f C_{eq})$), allowing alternating current to flow continuously while blocking any net DC offset.
Why is my measured series capacitance lower than the equation predicts?
If your LCR meter reads lower than the math suggests, you are likely measuring the parasitic effects of your test setup. Breadboard contacts, long alligator clips, and the internal inductance of the capacitor leads add series resistance and inductance. At high measurement frequencies (like 100kHz or 1MHz), these parasitics skew the impedance reading. To get an accurate reading, use a proper 4-terminal Kelvin fixture, keep leads under 2 inches, and zero out the meter's lead capacitance before testing.
Can I put polarized electrolytic capacitors in series for AC?
Not directly. Polarized electrolytics will suffer dielectric breakdown if subjected to reverse voltage. If you must create a non-polarized electrolytic equivalent for an AC application (like a passive speaker crossover), you have to wire two identical polarized electrolytics in "back-to-back" series (negative-to-negative or positive-to-positive). This creates a bipolar capacitor with half the original capacitance. However, for high-fidelity audio or precision AC filtering, it is vastly superior to use non-polarized film capacitors instead of hacking electrolytics.
