Back to Basics: What does Power Factor Mean and Why Must We Correct it?

February 4, 2013

Today’s commercial, industrial, retail and even domestic premises are increasingly populated by electronic devices such as PCs, monitors, servers and photocopiers which are usually powered by switched mode power supplies (SMPS). If not properly designed, these can present non-linear loads which impose harmonic currents and possibly voltages onto the mains power network.
Harmonics can damage cabling and equipment within this network, as well as other equipment connected to it. Problems include overheating and fire risk, high voltages and circulating currents, equipment malfunctions and component failures, and other possible consequences.
A non-linear load is liable to generate these harmonics if it has a poor power factor. Other loads can present poor power factors without creating harmonics. This post looks at these issues, the circumstances that can lead to damaging harmonic generation, and practical approaches to reducing it.

The two causes of poor power factor

At the simplest level, we could say that an electrical or electronic device’s power factor is the ratio of the power that it draws from the mains supply and the power that it actually consumes. An ‘ideal’ device has a power factor of 1.0 and consumes all the power that it draws. It would present a load that is linear and entirely resistive: that is, one that remains constant irrespective of input voltage, and has no significant inductance or capacitance. Fig. 1. shows the input waveforms that such a device would exhibit. Firstly, the current waveform is in phase with the voltage, and secondly both waveforms are sinusoidal.

Input voltage and current waveforms for a device with PF = 1.0

Fig 1: Input voltage and current waveforms for a device with PF = 1.0

In practice, some devices do have unity power factors, but many others do not. A device has a poor power factor for one of two reasons; either it draws current out of phase with the supply voltage, or it draws current in a non-sinusoidal waveform. The out of phase case, known as ‘displacement’ power factor, is typically associated with electric motors inside industrial equipment, while the non-sinusoidal case, known as ‘distortion’ power factor, is typically seen with electronic devices such as PCs, copiers and battery chargers driven by switched-mode power supplies (SMPSs). We shall look briefly at the displacement power factor before moving on to the distortion case, which is of more immediate concern to electronic power system designers. However it is important to be aware of both cases. For example, some engineering courses discuss the power factor issue only in terms of motors, which causes confusion when their students later encounter poor power factor as exhibited by an SMPS.

Electric motors and displacement power factor problems

Electric motors create powerful magnetic fields which produce a voltage, or back emf, in opposition to the applied voltage.  This causes the supply current to lag the applied voltage. The resulting out of phase current component cannot deliver usable power, yet it adds to the facility’s required supply capacity and electricity costs. Fitting capacitors across motors reduces the phase lag and improves their power factor.

SMPSs and distortion power factor problems

Non-sinusoidal current waveform drawn by SMPS with poor power factor

Fig 2: Non-sinusoidal current waveform drawn by SMPS with poor power factor

Whereas displacement power factor loads do not cause harmonics and their associated problems, distortion power factor loads such as SMPSs will do so unless their power factor is improved.

An SMPS’s AC front end typically comprises a bridge rectifier followed by a large filter capacitor. This circuit only draws current from the mains when the line voltage exceeds that across the capacitor. This causes current to flow discontinuously, resulting in the non-sinusoidal current waveform shown in Fig. 2.

It is possible to use Fourier transforms, a mathematical process, to analyse this waveform and break it down into a set of sinusoidal components. These comprise the fundamental frequency – 50 Hz in Europe, 60 Hz in America – and a set of predominantly odd multiples of the fundamental, known as harmonics. The third harmonic is 150 (or 180) Hz, the fifth, 250 (300) Hz and so on. Fig. 3 shows a typical harmonic spectrum for an electronic SMPS load. The fundamental component is usefully consumed by the SMPS, while the harmonics are reactive and create the problems described above. The ratio of the fundamental amplitude to the sum of all the harmonic amplitudes gives the device’s power factor.

Typical harmonic spectrum for an electronic SMPS load

Fig 3: Typical harmonic spectrum for an electronic SMPS load

International standard

An international standard exists to describe and set acceptable limits for a product’s mains harmonics generation. Within the EU, its reference is IEC 61000-3-2, covering equipment power levels from 75 W to 600 W. The standard assigns equipment into four Classes – A, B, C and D. Class D covers personal computers, personal computer monitors and television receivers.

Established and innovative PFC solutions

Although passive power factor solutions exist, the general industry view is that active designs offer the best power factor improvements. These are typically based on boost converter technology, as in the example shown in Fig. 4.

Active power factor correction circuit using voltage boost

Fig. 4: Active power factor correction circuit using voltage boost

In this active power factor correction circuit the incoming line voltage passes through a bridge rectifier, which produces a full wave rectified output (Figure 5–A). Since the peak value of the line is less than the bus voltage, no current will flow into the holdup capacitor unless the line voltage is boosted above that present on the holdup capacitor. This allows the control circuit to adjust the boost voltage (5–B-A) to maintain a sinusoidal input current.

To do so, the control circuit uses the input voltage waveform as a template. The control circuit measures the input current, compares it to the input voltage waveform, and adjusts the boost voltage to produce an input current waveform of the same shape (5–I). At the same time, the control circuit monitors the bus voltage and adjusts the boost voltage to maintain a coarsely regulated DC output (5–B). Since the primary function of the control circuit is to supply a sinusoidal input current, the DC bus voltage is allowed to vary slightly.

Voltage and current waveforms for active boost circuit of Fig 5

Fig 5: Voltage and current waveforms for active boost circuit of Fig 5

The use of an active power factor correcting circuit results in few discontinuities in the input current and consequently low distortion and harmonic content imposed on the input current drawn from the line. However Vicor has recently introduced a modular VI Brick® AC Front End, based on their new dynamic converter architecture, called Adaptive Cell.
The AC Front End offers a number of improvements for systems designers. In particular, it provides 85 V to 264 Vac universal input, high efficiency and high power density, especially considering that it is a complete solution including isolated and regulated DC output as well as rectification and power factor correction. The device reduces propagation of AC line harmonics, improving overall power quality at system and facility level. Total Harmonic Distortion exceeds EN61000-2-3 requirements, while high switching frequency and resonant transitions simplify external filtering and EMI standards compliance.

Other ‘Back to Basics’ Posts

Back to Basics: Understanding and Mitigating the Growing Problem of Distribution Losses

Back to Basics: Handling High Input Transients

Back to Basics: Meeting EMI for AC-DC Systems

Tags: ,

2 Responses to:
Back to Basics: What does Power Factor Mean and Why Must We Correct it?

  1. Jonathan Allen on February 6, 2013 at 11:04 AM

    Thank you for the article. The more usual definition of PF is the ratio of watts to volt-amps. The former is measured with an electrodynamometer (or electronic equivalent) while the latter is product of RMS voltmeter and ammeter readings. <1 displacement (phase) PF occurs when the load looks either inductive (motors) or capacitive to the generator.

    • Webmaster on February 8, 2013 at 11:42 AM

      Hi Jonathan – thank you for reading through our blog post, and your feedback.

      It’s true that PF is often defined as the ratio of watts to volt-amps. However our objective was to provide a simple definition of power factor as a concept. In this ‘Back to Basics’ context, we’re extending our reach to a less-technical audience, and wanted to focus on what is actually going on.

      Inductive and capacitive displacement PFs – We do mention that a load has to be free of both inductance and capacitance to achieve the ideal 1.0 power factor. Although it’s true that capacitive loads do exist and we could have exlored this further, we wanted to focus on the issues most likely to be encountered – inductive linear loads arising from electric motors, and nonlinear loads from SMPSs.

Find out more about our Cool-Power Buck Regulators subscribe to vicor newsletter Contact Us

Get Connected