When an original equipment manufacturer (OEM) capacitor part is discontinued or allocated, the replacement engineer’s job is not simply to match capacitance and voltage rating. In high-ambient-temperature applications—enclosures above 70°C, near motors, or in outdoor telecom cabinets—the thermal derating curve becomes the deciding factor between a reliable ten-year service life and a field failure within eighteen months. This article outlines a practical method for comparing derating behavior across brands, focusing on the parameters that must match exactly versus those that can reasonably deviate.
Why the Original Part Is Often Unavailable
Obsolescence in aluminum electrolytic and film capacitor lines is frequently driven by raw material changes or by the manufacturer consolidating its portfolio toward higher-voltage families. Distributors see this most often with 400 V and 450 V electrolytic capacitors used in industrial power supplies, where the original part carried a specific ripple current rating at 105°C. Because the replacement must operate in the same thermal environment, you cannot simply select a part with a higher voltage rating and assume the derating behavior is equivalent. The original manufacturer’s datasheet may have specified a ripple current at 105°C that the replacement cannot sustain, even though both parts share the same can size.
The core issue is that derating is not a single multiplier. It is a curve that depends on ambient temperature, internal hot-spot temperature, ripple current frequency, and the thermal resistance of the capacitor construction. When you cross-reference a capacitor, you are effectively comparing two thermal models, and the common practice of “10°C rule” substitutions—halving life for every 10°C increase—can lead to grossly inaccurate predictions if the replacement has a different ESR profile.
Parameters That Must Match Exactly vs. Permitted Deviation
For a high-ambient replacement, separate the specifications into hard constraints and tradeable parameters.
Hard Constraints (Must Match)
- Rated voltage and surge voltage: A 450 V rated part must not be replaced by a 400 V part, even if the applied voltage is only 350 V. The surge rating and the voltage derating curve for high ripple are not transferable.
- Capacitance tolerance: For filtering applications, a ±20% tolerance is often acceptable, but for timing or snubber circuits, the tolerance band must match the original exactly.
- Maximum operating temperature: If the original was rated for 105°C, a replacement rated for only 85°C is unsuitable regardless of any derating calculation. The internal chemistry differs.
- Ripple current at the highest rated temperature: This single number defines the thermal limit. If the replacement’s ripple current rating at 105°C is lower than the original, the application will push the part beyond its design point.
Tradeable Parameters (May Deviate with Analysis)
- ESR at 100 kHz: A lower ESR is generally acceptable, but a much lower ESR can shift the thermal hot spot toward the terminations, changing the failure mode. A ±30% deviation is workable if the ripple current is adjusted accordingly.
- Impedance at low temperature: For applications that see cold starts, a higher low-temperature impedance may still work if the inrush current is limited elsewhere.
- Can size and mounting: A larger can improves thermal resistance, so a physically larger replacement can sometimes meet the ripple requirement even if the original cannot.
Cross-Reference Table for High-Ambient Derating Comparison
The table below shows a generic comparison framework. Values are illustrative but typical of the ranges seen in industrial-grade electrolytic capacitors. Use this as a template for your own cross-referencing spreadsheets.
| Parameter | Original Part | Candidate Replacement | Decision Rule |
|---|---|---|---|
| Rated Voltage (V) | 450 | 450 | Must match exactly |
| Surge Voltage (V) | 500 | 500 | Must match or exceed |
| Capacitance (µF) | 470 | 470 | Within tolerance band |
| Max Temp (°C) | 105 | 105 | Must match |
| Ripple Current at 105°C (A) | 2.1 | 1.8 | May be lower if ambient is below 85°C |
| Ripple Current at 85°C (A) | 3.4 | 3.1 | Acceptable if ambient stays below 85°C |
| ESR at 100 kHz (mΩ) | 45 | 38 | ±30% deviation acceptable |
| Load Life at 105°C (hours) | 10,000 | 12,000 | Higher is better |
In this example, the candidate replacement has a lower ripple current rating at 105°C. Whether this is acceptable depends entirely on the actual operating ambient. If the equipment is rated for continuous operation at 80°C, the relevant ripple current is the one specified at 85°C, not the 105°C value. The table forces you to compare at the application’s real temperature, not at the datasheet’s extreme.
Verification Steps Before You Commit
Once you have narrowed the candidate list, run a structured verification sequence. This is especially important when you are sourcing from a secondary distributor or a broker, where the part may be a pull from another production line.
- Confirm the thermal impedance (Rth): Request the thermal resistance value from the manufacturer or calculate it from the ripple current and ESR. The formula is Rth = (T_hot_spot – T_ambient) / (I_ripple² × ESR). Compare this value against the original part.
- Measure ESR at the operating frequency: ESR is frequency-dependent. The 100 kHz value in the datasheet may not reflect the 120 Hz ripple present in a full-wave rectifier. Measure the ESR at the actual switching or line frequency using an LCR meter set to the application frequency.
- Check the ripple current derating multiplier: Most reputable manufacturers provide a frequency correction factor. If the factor at 10 kHz differs by more than 10% from the original, the internal foil construction is different, and the part may not handle the same harmonic content.
- Run a 24-hour thermal soak: For high-ambient applications above 70°C, mount the replacement on a test board with the same thermal environment, apply the rated ripple current, and measure the case temperature rise. Compare it to the original part’s measured rise. The replacement must stay within 5°C of the original’s case temperature.
- Verify voltage derating during surge events: If the application has intermittent surge conditions (e.g., motor braking or line transients), apply the surge voltage at the maximum ambient temperature. The capacitor must show no leakage current increase beyond the datasheet limit after 1,000 surge cycles.
A further consideration is the ripple current interaction with the ripple voltage. In a replacement scenario, the capacitance value remains the same, but if the ESR is lower, the ripple voltage will decrease. This is usually beneficial, but it can mask a resonance issue with the downstream converter if the ESR change shifts the zero of the output impedance. Verify the control loop stability if the ESR deviation is on the lower edge of the tolerance window.
The final approval should rest on the temperature rise measurement, not on the datasheet’s life rating. The life rating is a statistical estimate under controlled conditions. The thermal soak tells you whether the replacement actually fits your thermal budget. If the case temperature rise at your worst-case ambient is within 5°C of the original, the replacement is electrically and thermally equivalent for practical purposes. If it runs hotter, no amount of calculated derating will save the part from premature wear-out.