Component Derating Guide for PCB Assembly

Derating is the practice of using a component below its rated maximum so that it has margin against the conditions the product will actually see. The rated values in a datasheet describe a part at a reference condition, and the reference condition is rarely the one the board experiences. Derating rules translate the rating into a working limit, and the margin they create is what turns a design that works in the laboratory into one that still works after five years in the field.

Why Ratings Are Not Limits

A datasheet rating is the value at which the manufacturer guarantees the part will function, usually at a specified temperature and with a specified mounting. Operating at the rating is therefore not forbidden and is not a good idea, because the guarantee applies at the reference condition and the part is being used at the edge of its capability.

The margin between the reference condition and the application is where the difference lies. A resistor rated at a hundred and fifty degrees on a large pad may run hotter on a small one, a capacitor rated at its nominal voltage loses life rapidly as the voltage approaches the limit and a semiconductor rated at a hundred and fifty degrees junction has a much shorter life at that temperature than at a hundred.

The stress margin that derating creates is therefore not conservatism for its own sake. It is a way of accounting for the difference between the datasheet condition and the application, for the tolerances of the components themselves, for the variation of the supply and for the ageing of the material over the life of the product.

Voltage Derating

Voltage derating is most critical for capacitors, where the applied voltage relative to the rating has a strong effect on the failure rate and on the life. Ceramic capacitors also change their capacitance with the applied voltage, and a part used close to its rating may deliver a much smaller capacitance than its nominal value, which affects the circuit as well as the reliability.

Electrolytic capacitors have a different behaviour and are usually derated to a lower percentage of their rated voltage, partly because their life is governed by temperature and partly because the oxide layer behaves better below a certain field strength. Tantalum capacitors are derated for a different reason, since they can fail short when subjected to surges near their rating.

For semiconductors the relevant voltage is usually the blocking voltage and it should include the transients the circuit can produce rather than only the steady state. A switching node that rings above the supply rail during a transition will exceed a rating that was chosen from the steady state value, and the resulting failure appears as a random event until the waveform is examined.

Power and Current Derating

Power derating applies mainly to resistors and to semiconductors. A resistor rated at a watt is not intended to dissipate a watt at the maximum ambient, and the derating curve in the datasheet shows how the allowable power falls as the temperature rises. Using the rated power in a hot enclosure is one of the most common and easily avoided reliability problems.

Current derating applies to connectors, tracks, inductors and cables, and it usually includes a temperature rise allowance. A track rated for a current at a specified temperature rise will run hotter when it is in a bundle with other tracks or when the copper is thinner than the calculation assumed. Rechecking the current capacity against the actual construction is worth the few minutes it takes.

Power derating in semiconductors covers the junction temperature rather than the package dissipation alone, and it depends on the thermal path. A device that is comfortably derated on a board with a large copper pad may be close to its limit on a board with a small one, and the derating calculation should follow the actual thermal design rather than the nominal one.

Temperature and Other Stressors

Temperature is the stressor that dominates most reliability calculations, and it enters twice: as the ambient that sets the operating point and as the cycling that causes fatigue. Derating reduces both effects, because a part running cooler is further from the failure mechanisms that depend on temperature and experiences a smaller swing during a cycle.

Mechanical stress matters for ceramic capacitors in particular, where board flex and thermal expansion can crack the body, and for components mounted on a board that will be subject to vibration. The derating rules for these cases are about placement and orientation as much as about electrical values, and they should be written into the layout rules rather than left to the designer’s judgement.

Humidity and contamination are the third group, affecting insulation resistance, migration and the corrosion of fine features. Where the product is exposed to a humid environment, the voltage derating should be more generous because the surface resistance falls as the material absorbs water.

Derating Standards and Rules

Most organisations use a derating standard rather than a set of improvised rules, and the common ones specify a percentage of the rated value for each component class and each stress. The percentages are not arbitrary; they come from failure data and they are a starting point rather than a precise engineering result.

Component derating should be applied as a design check rather than as an afterthought. A derating table filled in after the design is complete will find violations, and at that point the options are limited to changing components or accepting the risk. A table used during the selection of components produces a design that passes without rework.

The rules also need a route for exceptions. Some circuits are required to operate close to a rating for the function to work at all, and the exception should be recorded with its justification and its mitigation rather than quietly ignored. The record is what makes the design review meaningful.

Verifying Derating in Production

Derating is verified by measuring the actual stresses rather than by trusting the calculation. The temperature of key components at the worst case ambient, the voltage across a capacitor in the actual circuit and the current in a track are all measurable, and each of them should be checked on the first article. The measurement resolves the argument about whether the calculation was conservative.

The measurements should be taken at the extremes rather than at a convenient operating point. Maximum load, maximum ambient and minimum supply voltage together produce the worst case for most stresses, and a measurement made at a comfortable operating point proves nothing about the design.

The results should be recorded with the design and retained, because the same board built later may use a different component or a different supplier. Comparing the new measurement against the original record is a quick way to confirm that the substitution did not compromise the margin.

Practical Rules

Apply a derating standard during component selection rather than after the design, and record any exception with its justification. Measure the actual stresses on the first article at the worst case conditions.

Recheck the derating whenever a component is substituted, and keep the results with the design and build records. Derating is the cheapest reliability measure available, because it costs nothing beyond attention at the point of selection, and the field quality of a product usually reflects how well it was applied.

FAQ

Why not use components at their rated values? Because the rating applies at a reference condition that the application rarely matches, and because life falls rapidly as a stress approaches the limit.

Which component needs the most careful voltage derating? Capacitors, particularly ceramics, where applied voltage also changes the capacitance, and tantalums, which can fail short near their rating.

When should derating be checked? During component selection, again at the design review, and whenever a component is substituted.