Almost all AC-powered electronics products today use switch-mode power supplies (SMPS), most with the potential to generate significant power-line disturbances. Since equipment developers want power supplies that will not cause any undue interference, more engineers now are performing some type of power-consumption analysis.

At the same time, more electronics equipment users are concerned about how to safeguard system performance in spite of power-line disturbances, whether caused by SMPS, motors, or the utility company. Consequently, more users want to conduct power analysis.

But equipment designers and end users have different power-analysis objectives: one to perform load-related tests, the other for source-related tests. While some power analyzers perform both functions, most serve either one purpose or the other.

Source-Related Tests

Since you usually cannot predict the type and timing of the next disturbance, the equipment suited for your purpose must record a range of anomalies over some period of time. A typical single-phase analyzer of this type, such as the International Power Technologies Sherlock Power Analyzer, measures and time-stamps voltage aberrations, brownouts, blackouts, dropouts, harmonic distortion, normal- and common-mode voltage spikes, and noise as well as line frequency and phase-shift changes.

Normal/abnormal limits can be pre-set for the three-phase PowerTroncics disturbance and voltage recorders, which profile voltages and record up to 6,000 transient events. Subsequent to acquisition, data is transferred to a PC and profiles can be graphed for 11 days per channel.

The Valhalla 3030A single- and three-phase harmonic energy analyzer-datalogger performs both source- and load-related tests. Logging periods are selectable from one second to 99 minutes and the unit can be left unattended for up to two months.

An analysis of the data collected over a well-chosen time period usually will reveal what action should be taken to eliminate the line disturbances or what to do to protect your equipment. For the latter case, this set of commonly encountered situations and appropriate solutions is outlined by Tim Bailey, marketing manager at International Power Technologies:

Brownouts, overvoltages, and abnormal average line voltages—A voltage regulator, an uninterruptable power supply (UPS) with voltage regulation, or a dedicated electrical wiring circuit may be required. The collected data shows the voltage boost or buck range required to correct the problem.

Blackouts and dropouts—A UPS is necessary. The outage periods recorded and the application determine the required backup time to be supported by the UPS battery.

Average neutral voltage and common-mode noise—If the noise is severe, an isolation transformer may be required. If less severe, a good surge suppressor may be adequate. Also, better grounding may correct the problem.

Normal-mode voltage spikes and common-mode voltage spikes—A surge suppressor or a UPS with effective surge suppression is needed.

Normal-mode noise—A surge suppressor or a UPS with excellent EMI/RFI line noise filtering is required.

Line frequency changes and AC line phase shift—An on-line, double-conversion UPS is needed to regenerate a stable sine wave.

Load-Related Tests

Load analyses are performed to qualify the design of a product or to determine countermeasures needed to eliminate disturbances caused by an offending power-consuming device. Disturbances fall into two classes: steady-state and transient.

The most readily correctable steady-state aberration is the current vs voltage phase shift commonly encountered with industrial loads. For instance, the phase lag produced by AC motors, which represent a combined inductive and resistive load, can be corrected by adding a compensating capacitor.

Most power analyzers provide a direct readout of the magnitude of the leading or lagging phase angle, making it very easy to select a suitable capacitor value for zero phase shift. Uncorrected phase shifts or a power factor of <1 result in a discrepancy between the true power consumed in watts (W) and the product of voltage times current (VA). Utility companies are entitled to assess penalty fees if power factors are substantially <1.

With constant motor loads, the phase shift does not change. However, during start-up or sudden load changes, not only does the phase change but additional transients also may occur.

Measuring transients and start-up phenomena requires special facilities. “The Voltech PM3000A Power Analyzer uses a unique operating mode which collects cycle-by-cycle rms data throughout a 30-s to 40-s acquisition period,” explained Tom Mahr, vice president and general manager of Voltech. The PM3000A, the Valhalla D6000, and the Yokogawa WT1000 also accept motor-torque and speed inputs to automatically calculate motor efficiency.

Steady-state performance assessments of nonlinear loads, such as lighting ballasts, pulse-width modulated motor controllers, and electronic power supplies, require an additional set of analysis features. Current consumed by any nonlinear load is nonsinusoidal. But regardless of its shape, it may be represented and treated as a fundamental plus a number of harmonic sine waves.

The reasons for concern about this type of current consumption are twofold. Current drawn at the fundamental frequency produces no current flow in the neutral wire of the power-distribution system. However, odd-order harmonics reinforce each other on the neutral line and, when of significant magnitude, cause overheating.

Secondly, any power-distribution system has a small but finite impedance. The harmonic currents flowing through this impedance produce harmonic voltages on the power grid, which may cause interference on any sensitive equipment fed from the same line.

Two types of test instruments can assess the degree of nonsinusoidal current consumption of equipment. The first is an inexpensive true-rms-reading power meter with peak measurement capability. Comparing the measured crest factor with the pure-sine wave crest factor of 1.414 yields an indication of the severity of harmonics.

Alternatively, the true-rms-reading power meter may be used in conjunction with an average-reading current clamp. In this case, a ratio of average/rms equal to one indicates the absence of harmonics and values of 0.5 would show high harmonic content.

The second type of instrument is a harmonics analyzer, which can be used to determine the exact magnitude and, in some cases, the phase relationships of each harmonic. Most full-featured power analyzers contain extensive harmonic analysis capabilities. Their bandwidth and computational capabilities determine the highest order harmonic they can measure.** **

**Regulatory Requirements**

** **With the proliferation of nonlinear loads, especially electronic power supplies, more emphasis is placed on limiting allowable harmonic magnitudes and, if found excessive, applying power-factor correction. The European Community (EC) has defined strict test procedures for measuring harmonic content, based on the original International Electrotechnical Commission (IEC) 555 Part 2 requirements but modified and formalized as IEC 1000-3-2 or EN61000-3-2.

The IEC 555 Part 3 specification for allowable flicker also has been updated, and the present test requirements are defined in IEC 1000-3-3 or EN61000-3-3. Flicker pertains to temporary voltage disturbances caused by a load during defined short-term and long-term periods.

Equipment to be sold in EC markets must pass the IEC 1000-3-2 and the IEC 1000-3-3 tests. The Voltech PM3000-A, the Valhalla D6000, the Yokogawa WT2000, the Marconi/Zimmer LMG 310, and the California Instruments iL Series provide the manual and automated measurement capabilities needed to perform both sets of tests.

A Range of Choices

The comparison chart accompanying this article lists instruments for line and load analysis. They are only a sampling of the power analyzers offered by the companies. Call for more detailed specifications or information on other models.

SIDEBAR

Definitions of Selected Measurement Capabilities

The Power Analyzers Comparison Chart that accompanies this article shows some specific measurement functions performed by the instruments. These include Power Factor, Phase Angle, Crest Factor, Real Power, Reactive Power and Apparent Power. Here are the equations that define these functions.

P (in watts)

Power Factor = ———————————–

E (in V_{rms}) × I (in A_{rms})

For pure sine-wave voltage sources and linear loads, this equation can be rewritten as:

E × I × cos f

Power Factor = ——————— = cos f

E × I

where f = the phase angle between E and I.

For nonlinear conditions, values for E, I, or P may be obtained in the time-domain by integrating incremental values over a period of one cycle. More often, they are obtained by frequency-domain conversion (using FFT) and calculating the root mean square of a summation of the fundamental component and all harmonics.

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Peak Value Amplitude

Crest Factor = ————————————-

True rms Value Amplitude

For sine-waves, the crest factor (cf) = 1.414; for triangular waves, cf = 1.73; and for square waves, cf = 1.

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Real Power is the component of apparent power that represents true work. It is defined as:

Real Power = E (in Vrms) × I (in Arms) × power factor

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Reactive Power = E (in V_{rms}) × I (in A_{rms}) × sin f

Apparent Power = E (in V_{rms}) × I (in A_{rms})

**Copyright 1997 Nelson Publishing Inc.**

June 1997