Why this measurement is difficult to get right
Power supply testing lives at the extremes of the measuring range, and both extremes are normally reported from the same instrument in the same report. Full-load efficiency is measured at the top of the range, while no-load and standby consumption are measured at a fraction of a watt. The specification is therefore decided by the end of the range that is hardest to measure, which is the bottom.
This is why low-end performance dominates the choice. An instrument that is excellent at 20 A but cannot resolve a fraction of a milliamp will not produce a usable standby figure, however good its headline accuracy. A wide-range design that keeps its resolution across the whole span handles the entire test sequence without being reconfigured between steps, which also removes a common source of operator error.
The second factor is the shape of the current, not its size. A switching supply draws a highly peaked input current, and an instrument that cannot tolerate a high crest factor will clip or distort the reading. On the output side the question is whether the instrument accepts DC at all, because many units specified for AC-only work cannot produce the output-side figure that the efficiency calculation needs.
What has to be measured
- Input power at no load and in standby, typically far below the rated figure.
- Input and output power at rated load, from which efficiency is calculated.
- Power factor and the behaviour of the input current waveform.
- Temperature rise and output regulation under sustained load.
- The load condition and duration of each reading, because a short reading and a soaked reading of the same unit are different results.
Which instrument measures what
| Measurement | What the reading has to establish | Instruments commonly specified |
| Wide-range input power with fine low-end resolution | That one instrument can capture both the standby figure and the rated figure without a range change that would invalidate the comparison | Digital power meters and AC/DC electrical parameter testers |
| Low-power and standby measurement | A figure at the bottom of the range that is stable enough to be compared against a regulatory threshold | Low-current power meters with high resolution near zero |
| Efficiency calculation and comparison | The ratio between two readings taken at different points, so both contribute their uncertainty to the result | Power meters with communication output and recording |
| Thermal behaviour under load | Whether the unit stays inside its declared limits once it has soaked at rated load rather than on a cold start | Temperature measurement and data logging |
Efficiency is a ratio of two measured values, so the uncertainty of both ends reaches the answer.
Narrowing the choice for a bench supply
- Find the regulatory or contractual threshold the unit has to meet, because it sets the required accuracy rather than the instrument catalogue.
- Note the lowest figure that has to be reported. Standby consumption is usually the number under scrutiny and it is usually three orders of magnitude below rated load.
- Confirm the instrument resolves that low figure while still handling rated load, ideally in one range.
- Ask for the crest factor the instrument tolerates, since the input current of a switching supply is highly peaked.
- Establish whether the output side has to be measured, which decides whether DC measurement is required or whether an AC-only unit will do.
- Decide whether efficiency must be calculated inside the instrument or from exported readings, which changes the interface requirement.
- Check whether the test has to run unattended, because a soak test that runs overnight needs logging rather than a display.
- Confirm the calibration document the customer expects, since efficiency results are frequently contractual.
Where this application ends and the next begins
| Adjacent application | What separates it from power supply testing |
| Electrical parameter testing | That work judges a device at one declared operating point, while this one spans the whole range because standby and full load appear in the same report. |
| Energy efficiency testing | Here efficiency is an instantaneous ratio between input and output, whereas that work accumulates consumption across a representative cycle and reports energy. |
| Motor & drive testing | The decisive quantity there is bandwidth against a changing switching frequency, while here it is resolution at the bottom of a fixed range. |
| Industrial test bench & calibration | A bench verifies that the supply itself is telling the truth, which is a check on the instrument rather than a test of the product. |
| Product quality inspection | Inspection passes or fails a physical item, while this work produces performance figures that a regulation or a customer will read. |
What to send with the enquiry
- Type of unit - adapter, charger, open-frame supply or module.
- The threshold it has to meet, and where that threshold comes from.
- Rated input and output, including the lowest figure that must be reported.
- Whether the output is AC or DC, and whether it must be measured.
- Expected crest factor or the nature of the input current waveform.
- Whether readings must be exported, and to which interface.
- Whether the test runs attended or unattended, and for how long.
- Supply arrangement at the test position - single or three phase.
- Quantity required and the date it has to be in use.
- Whether a calibration certificate has to accompany the instrument.
