ULH 150 750 J FL=500 Measured: Full Data for Ripple Current vs. ESR Temperature Rise Curve

Internal temperature rise in capacitors caused by ripple current is an invisible killer most easily underestimated by power supply engineers during component selection. Taking the ULH 150 750 series as an example, under the condition of a lead length of FL=500mm, how can the coupling effect between ripple current and ESR be quantitatively evaluated? Based on measured data, this article fully presents a multi-dimensional analysis of the temperature rise curve, providing practical selection criteria for high-frequency switching power supply designs.

Technical Background and Test Conditions of the ULH 150 750 Series

As a representative specification of high-capacitance aluminum electrolytic capacitors, the ULH 150 750 J series, with a nominal capacitance of 150 μF and a voltage rating of 750V, is widely used in photovoltaic inverters, industrial frequency converters, and electric vehicle OBC (on-board charger) applications. The FL=500mm lead specification directly determines the baseline level of equivalent series inductance (ESL), which in turn affects the distribution characteristics of high-frequency ripple currents.

Product Structure and FL=500 Lead Specification Definition

This series features an axial leaded structure, where FL=500 denotes an exposed lead length of 500mm after forming. This length represents typical medium-to-long distance routing in system-level wiring, with a lead resistance of approximately 12-18 mΩ and an inductance of up to 80-120 nH. For switching frequencies above 100 kHz, the inductive reactance caused by ESL cannot be ignored and must be incorporated into the frequency-domain analysis framework of the ripple current.

Test Platform Setup and Environmental Control Standards

The measurement setup utilizes a programmable electronic load paired with a wideband LCR meter, and temperature data is collected using K-type thermocouples embedded at multiple points inside the capacitor core. The ambient temperature is controlled at 25±1°C, with forced air cooling turned off to simulate the worst-case natural convection cooling conditions. The ripple current source covers the 100 kHz to 1 MHz frequency band, with amplitudes loaded in steps from 20% to 120% of the rated value.

Measured Ripple Current Data: Multi-Frequency, Multi-Load Sweep

The root-mean-square (RMS) value of the ripple current (Irms) is not a simple sinusoidal wave at a single frequency, but rather the superposition of multiple harmonic components within a switching power supply. In the measurements, the thermal contributions of the fundamental wave and major harmonic components were extracted separately to accurately evaluate the actual thermal load of the capacitor.

Current Carrying Capacity under Typical 100 kHz/200 kHz Operating Conditions

Under the conditions of a 100 kHz fundamental frequency and a 105°C upper case temperature limit, the rated ripple current of ULH 150 750 J FL=500 is approximately 2.8 Arms. When the frequency is increased to 200 kHz, the effective conductive cross-sectional area of the leads decreases due to the skin effect, causing the equivalent current-carrying capacity to drop by about 12% to 15%. Measured data shows that after loading 2.5 Arms at 200 kHz for 30 minutes, the hotspot temperature of the capacitor core reached 98°C, approaching the safety boundary.

Non-Linear Relationship Mapping of Ripple Current versus Frequency

As the frequency rises, the ripple current carrying capacity does not decay linearly. In the frequency band above 500 kHz, ESL and the capacitor body form a series resonance, leading to current amplification at specific frequency points. A weak resonance peak was observed near 1 MHz during testing, indicating that a 15% to 20% current margin should be reserved in the design to avoid localized overheating.

ESR Frequency Spectrum Characteristics: Attenuation Curve from Low to High Frequencies

As the core variable in ripple current thermal loss, the frequency-domain behavior of ESR directly determines the transient thermal power of Ploss = Irms² × ESR. The temperature sensitivity of the ESR of the ULH 150 750 J constitutes a critical link in the positive feedback mechanism of temperature rise.

Measured Temperature Sensitivity Coefficient of Equivalent Series Resistance

At a 25°C baseline, the typical ESR of this specification is 180 mΩ; when the core temperature rises to 105°C, the ESR decays to approximately 65 mΩ. This negative temperature coefficient might seem advantageous, but it actually conceals a risk—during the low-temperature startup phase, the ESR is exceptionally high, resulting in thermal loss up to 7.6 times that of the high-temperature state under the same ripple current conditions, which easily leads to "startup overshoot temperature rise."

Fitting Model of ESR Variation with Frequency

The bivariate ESR(f, T) model constructed from measured data shows that at 100 kHz, ESR is dominated by electrolyte ion conduction loss, whereas above 1 MHz, it transitions into a superimposed control region dominated by oxide film dielectric loss and the lead skin effect. The long FL=500 leads introduce an additional 8-12 mΩ of high-frequency ESR, which must be accounted for separately as a lead impedance term in the model.

Core Temperature Rise Curve Data: Ripple Current × ESR Coupling Effect

Comparing the dual curves of steady-state and transient temperature rise reveals fundamental differences in thermal management design. Under steady-state conditions, the heat capacity effect is fully released, and the temperature rise is strictly positively correlated with the thermal resistance θca. During transient loading, the thermal capacity of the core absorbs part of the thermal shock, resulting in a significant time lag in the apparent temperature rise.

Dual-Curve Comparison Between Steady-State and Transient Temperature Rise

Measurement under 2.8 Arms / 100 kHz step-loading: The transient temperature rise rate in the first minute can reach up to 12°C/min, slows down after 5 minutes, and reaches a steady-state plateau at 30 minutes. This thermal inertia characteristic implies that peak currents during short-term overload (such as motor startup) can moderately exceed the rated value, but the duration must be strictly controlled within 20% of the thermal time constant τ (measured at approximately 180 seconds).

Quantitative Analysis of Thermal Resistance θca and Dissipation Paths

The FL=500 leads function as both the primary heat dissipation path and an additional heat source. The measured temperature gradient from the lead base to the solder joint reaches 15-20°C, indicating that the lead body accounts for approximately 30% of the total thermal resistance. Optimization strategies include shortening the leads to under 300mm to reduce θca by about 18%, or utilizing copper busbars for parallel current division.

Engineering Application Guide: Selection Strategies Based on Measured Data

Translating measured data into actionable engineering criteria requires establishing an integrated calculation framework linking ripple current, ESR frequency spectrum, and the thermal environment, rather than looking at datasheet nominal values in isolation.

Derating Curves and Safety Margin Design Recommendations

A three-level derating strategy is recommended: frequency derating (applying a 0.7-0.85 factor at high frequencies), temperature derating (leaving a 10°C margin below the upper case temperature limit), and lifetime derating (designing for target lifetime at 50% of the nominal value). For the FL=500 specification, a comprehensive derating factor of 0.55-0.65 is suggested, meaning actual operating ripple currents should not exceed 1.5-1.8 Arms (referenced at 100 kHz).

Parallel Current Sharing and Thermal Management Optimization Solutions

When multiple capacitors are connected in parallel, the variance in ESR (typically ±20%) leads to uneven current sharing. In measurements, the current deviation between two parallel branches can reach up to 35%, with hotspots concentrating on the lower ESR unit. Current-sharing mitigation measures include symmetrical trace design, individual series current-sharing resistors (in the 5-10 mΩ range), or selecting ESR-matched/binned products.

Key Summary

• The negative temperature coefficient characteristic of the ESR of ULH 150 750 J FL=500 causes low-temperature startup thermal loss to reach up to 7.6 times that of the high-temperature state, requiring focused protection against startup overshoot risks
• The long FL=500 leads increase high-frequency ESR by an additional 8-12 mΩ, deteriorating the thermal resistance θca to above 25°C/W in frequency bands above 500 kHz; shortening leads or using parallel busbars is recommended for high-frequency applications
• The ripple current carrying capacity decays non-linearly with frequency, with an ESL resonance peak existing near 1 MHz; a design margin of 15%-20% should be reserved
• A comprehensive three-level derating factor of 0.55-0.65 is recommended, keeping actual operating ripple currents within 1.5-1.8 Arms to ensure long-term reliability
• In multi-parallel scenarios, attention must be paid to uneven current distribution caused by ESR variation; symmetrical layout and current-sharing resistors can effectively mitigate hotspot concentration

Frequently Asked Questions