Overunity measurement error, not exotic physics, explains nearly every reported case of a device producing more power than it consumes. Measuring real power in a switching circuit is genuinely difficult, and the standard failure modes all push the answer in the same direction: input understated, output overstated, apparent efficiency above one hundred percent.
Real power is an average of a product, not a product of averages
Power at any instant is voltage multiplied by current at that same instant, and real power is the average of that product over a cycle. Many people instead multiply an average or root-mean-square voltage by an average current, which gives the correct answer only for sinusoidal waveforms in phase. With pulsed or distorted signals, the two calculations can differ enormously. If voltage peaks while current is near zero and current peaks while voltage is low, the true average product is far smaller than the product of the separate averages, or the reverse. Devices built around switching transients produce exactly this kind of waveform, which is why they are the natural habitat for false surplus readings. Output measurements fail in mirror-image ways. Reading a peak voltage across a load and treating it as continuous, or ignoring the duty cycle of a pulsed output, inflates the computed result by a large factor.
Instrument bandwidth and sampling are limiting factors
A meter or oscilloscope only reports what it can see. A handheld multimeter designed for power frequency has a bandwidth of a few kilohertz at best and will simply not register the content of a pulse with a microsecond rise time, producing an input reading that omits most of the energy. Digital oscilloscopes alias if the sample rate is too low relative to the signal, producing waveforms that look plausible and are wrong. Probe bandwidth, current probe frequency response, and the roll-off of the meter’s own input stage all matter. Measuring a fast switching circuit correctly requires instruments rated well above the highest significant frequency present, verified against a known source rather than assumed.
Phase and timing errors are subtle and decisive
Voltage and current are usually measured through different paths: a voltage probe with its own delay and a current probe or shunt with another. A small timing difference between the two channels shifts the apparent phase, and because real power depends on the cosine of that phase angle, a shift of a few degrees changes the computed power substantially when the power factor is already low. In a circuit with large circulating reactive current and small real power, a modest phase error can flip the sign of the answer entirely. Any serious measurement must include a de-skew step using a known resistive load, and the result should be sanity-checked against a case where the answer is already known. Shunt resistors bring their own trap, since a shunt with appreciable inductance develops a voltage proportional to the rate of current change rather than the current, which badly distorts fast pulses.
Ground loops and common-mode error contaminate the reading
Connecting two oscilloscope probes with grounded leads to different points in a circuit creates a loop and can short part of the circuit through the instrument ground. The displayed waveform then describes a circuit that the probes themselves have altered. Common-mode voltage on a differential measurement can exceed the probe’s rejection and inject error. Long ground leads act as inductors and ring. These effects are well known in power electronics work, and they produce waveform features that look like genuine phenomena. A basic check is to short both probe tips to the same point and confirm the display shows nothing; a surprising number of anomalous results fail that test.
The measurement that settles the question
Rather than arguing about waveforms, measure total energy over a long period. Run the device from a fully charged battery of known capacity and see how long it operates, or from a supply through a certified energy meter and record kilowatt-hours. On the output side, dissipate the energy into a resistive load and measure the temperature rise of a known mass of water, which is calorimetry and cares nothing about waveform shape. Energy over hours is a scalar quantity that no phase error can distort. When builders of these devices are asked for that measurement rather than an oscilloscope screenshot, the reported surplus reliably disappears.
What to check before you act
- Compute real power as the average of the instantaneous product.
- Verify instrument bandwidth exceeds the fastest edges in the circuit.
- De-skew voltage and current channels against a resistive load.
- Confirm results with long-duration energy measurement or calorimetry.
Common questions
Can a good multimeter measure switching circuit power?
Generally no. Bandwidth and averaging assumptions make ordinary meters unsuitable for pulsed waveforms. Use a wideband power analyzer or an energy measurement over time.
Why do reactive circuits confuse these measurements?
Large circulating current with little real power means the answer depends on a small cosine term, so even a minor phase error changes the computed result dramatically.
The practical takeaway
Whenever a circuit appears to produce surplus power, suspect the measurement first. Check bandwidth, de-skew the channels, and then settle it with energy over hours or with calorimetry, because those methods are immune to the errors that generate the illusion.
References and further reading
- Department of Energy: measuring appliance energy consumption accurately
- ENERGY STAR: standardized test procedures behind efficiency claims
- Department of Energy: energy efficiency measurement and savings resources
Numerical scenarios are illustrative unless identified otherwise. Follow the exact product instructions; component ratings and local installation requirements can differ.
