The Direct Answer: Which Supercapacitor for Which Job?
Selecting the right supercapacitor requires matching the internal chemistry and construction to your specific discharge profile. A coin cell will fail catastrophically if asked to deliver the pulse current required by a GSM modem, while a massive cylindrical EDLC is a waste of board space for simple RTC (Real-Time Clock) backup.| Type | Construction | Tolerance | Tempco / Operating Range | Typical Use Case |
|---|---|---|---|---|
| Coin Cell EDLC | Activated carbon, organic electrolyte, stacked | -20% / +80% | -25°C to +70°C (High ESR at cold) | RTC backup, SRAM memory retention, low-current energy harvesting |
| Radial/Cylindrical EDLC | Wound carbon foil, aqueous or organic electrolyte | -10% / +30% | -40°C to +85°C (Low ESR) | Motor start assist, audio amplifier buffer, solar tracker pulse loads |
| Pseudocapacitor | Metal oxide (RuO2 / MnO2) faradaic reactions | ±20% | -20°C to +65°C | High-density micro-energy, wearable medical sensors |
| Hybrid (Li-Cap) | Li-ion intercalation anode, activated carbon cathode | ±10% | -40°C to +70°C | EV regenerative braking, UPS bridge power, smart metering |
Decoding the Can: How to Read Supercapacitor Markings
Unlike ceramic capacitors that rely on the EIA 3-digit pF code (where104 means 100nF), supercapacitors use entirely different marking conventions. Misinterpreting these codes is a common reason hobbyists order parts that are off by a factor of 10,000.
1. Direct Printing (Cylindrical and Large Coin Cells)
Most radial and large coin supercaps print the nominal values directly on the heat-shrink sleeve or metal can. You will see 2.7V 10F 85°C. The critical marking here is the polarity stripe. The stripe on a supercapacitor denotes the negative terminal (just like standard aluminum electrolytics). Applying reverse voltage to an EDLC causes rapid electrolyte decomposition and gas generation.
2. The 3-Digit Code Trap (SMD and Micro Coin Cells)
On tiny 5.5V SMD or micro-coin supercaps, manufacturers often use a 3-digit code. However, the multiplier is usually in microfarads (µF), not picofarads (pF).
If you see 104 on a tiny supercap, it means $10 \times 10^4$ µF = 100,000 µF = 0.1F.
If you mistakenly apply the ceramic pF rule, you would calculate 100,000 pF (0.1 µF) and wonder why your circuit dies in milliseconds. Always verify the multiplier unit on the specific manufacturer's datasheet (e.g., Panasonic, Elna, or Eaton).
3. Voltage and Date Codes
Look for the maximum DC voltage rating (usually 2.5V, 2.7V, 5.4V, or 5.5V). Operating a 2.7V supercap at 2.8V cuts its operational lifespan by more than half. Date codes are typically 4 digits (e.g., 2341 for week 41 of 2023). Supercaps degrade on the shelf if stored at high temperatures; avoid using stock that has been sitting in a hot warehouse for over two years.
Real-World Scenario: The 5V Backup Power Miscalculation
Let us walk through a classic bench failure involving a 3.3V microcontroller RTC backup circuit. According to Eaton's supercapacitor technical guides, ignoring leakage current is the number one cause of backup power failure.
The Numbers (Theoretical):
The DS3231 draws $I_{load} = 1\mu A$ at 3.3V. The RTC stops functioning at $V_{dropout} = 1.8V$.
$\Delta V = 3.3V - 1.8V = 1.5V$.
Using the standard capacitor discharge formula $t = \frac{C \times \Delta V}{I}$:
$t = \frac{0.47F \times 1.5V}{1\mu A} = 705,000$ seconds $\approx$ 195 hours (8.1 days).
The Outcome:
The RTC lost time after just 14 hours on the bench.
What Went Wrong:
The theoretical math ignored two massive real-world factors:
- Leakage Current: The datasheet specified a leakage current of $1.5\mu A$ at 24 hours and 25°C. The total current draw was actually $1\mu A$ (load) + $1.5\mu A$ (leakage) = $2.5\mu A$. This alone cuts the time to 78 hours.
- Dielectric Absorption and Temperature: The bench was sitting at 35°C. Supercapacitor leakage current roughly doubles for every 10°C increase above 25°C. At 35°C, leakage spiked to ~$3\mu A$. Combined with the initial dielectric absorption (which acts like a parasitic load soaking up charge in the first few hours), the effective drain exceeded $4\mu A$, collapsing the 195-hour theoretical runtime down to roughly 14 hours.
The Fix: We replaced the 0.47F coin cell with a 1.5F radial EDLC and added a Schottky diode (low forward voltage drop) to isolate the supercap from the main rail's parasitic paths, extending the verified backup time to 9 days.
Visual Symptoms and Failure Modes
Supercapacitors rarely fail with the violent explosions seen in reverse-polarity aluminum electrolytics, but they do exhibit distinct visual and electrical symptoms when abused. For a deeper look into EDLC degradation mechanisms, refer to the Cornell Dubilier (CDE) application notes.| Failure Mode | Visual Symptom | Electrical Symptom | Root Cause |
|---|---|---|---|
| Reverse Polarity | Can bulges, safety vent pops, sweet chemical smell (PC/ACN electrolyte) | Dead short (near 0Ω ESR), 0F capacitance | Electrolyte decomposition generating gas; internal separator breakdown |
| Overvoltage (>110% Vmax) | Heat-shrink sleeve shrinks, discolors, or becomes brittle | Capacitance drops >30%, ESR spikes >200% | Accelerated oxidation of the carbon electrode; gas generation increases internal pressure |
| Excessive Ripple Current | Sleeve cracking, terminal epoxy softening or weeping | Thermal runaway, sudden open-circuit failure | $I^2R$ heating exceeds the thermal dissipation limit of the can; internal connections melt |
| High Temp + Max Voltage | No external visual changes (silent failure) | Gradual capacity loss over weeks/months | Electrolyte evaporation through the seal; standard Arrhenius degradation rate |
Safe Substitution: When You Don't Have the Exact Part
When your BOM calls for a specific supercapacitor and it is out of stock, you can substitute, but you must follow strict electrical rules. As noted in comprehensive guides on All About Circuits, treating supercaps like standard capacitors during substitution leads to immediate circuit failure.
Rule 1: Voltage Ratings Are Absolute Maximums
You can always substitute a higher voltage supercapacitor for a lower one (e.g., using a 5.5V part in a 2.7V circuit). You cannot substitute a lower voltage part. If your circuit peaks at 3.0V and you sub in a 2.7V supercap, it will degrade rapidly and eventually vent. Unlike MLCC ceramics, supercapacitors do not tolerate overvoltage gracefully.
Rule 2: Capacitance Can Increase, But Watch the Leakage
Substituting a 10F part for a 5F part is generally safe and will increase your backup time. However, larger capacitance physically means more surface area, which inherently increases leakage current. If your circuit is ultra-low-power (e.g., a nano-amp energy harvesting node), a massive 100F supercap might leak more current than your harvester can generate, resulting in a net-zero charge state.
Rule 3: Series Stacking Requires Active or Passive Balancing
If you need 5.4V and only have 2.7V 10F supercaps, you must put two in series.
The Math: Two 10F caps in series yields 5.4V at 5F (capacitance halves in series).
The Danger: Because no two supercaps have identical leakage currents, the voltage will not divide evenly. The cap with the lower leakage will absorb more than 2.7V, overvolting and failing.
The Fix: You must add balancing resistors across each cap. Calculate the resistor value using the worst-case leakage current. If $I_{leak(max)} = 5\mu A$ at 2.7V, $R = \frac{2.7V}{5\mu A} = 540k\Omega$. Use standard 470kΩ 1% resistors across each capacitor to bleed off excess voltage and ensure equal distribution.
