Testing – Input Voltage Line Regulation and Power Factor

This content was originally featured on Amdmb.com and has been converted to PC Perspective’s website. Some color changes and flaws may appear.

During the previous load tests we set the AC input voltage to 115 VAC.  This is an optimum value for most of the power supplies under test.  To find out how well each power supply can handle under and over voltage conditions on the AC mains, I lowered the input voltage to 100 VAC and then raised it to 130 VAC with a Variac (variable autotransformer).  Once again we are interested in seeing how well each PSU can maintain the various output voltages as the input voltage fluctuates +/- 13% (+/- 15 VAC).

Ten ATX Power Supplies Compared - Cases and Cooling 107

Ten ATX Power Supplies Compared - Cases and Cooling 108

Ten ATX Power Supplies Compared - Cases and Cooling 109

All of the PSUs tested exhibited excellent line regulation with varying input voltage.  None of the power supplies tested showed more than a few one hundredths of a volt change as the line input went from 100 VAC through 115 VAC all the way up to 130 VAC.  Very good!

Testing – Power Factor (PF)

Power factor (PF) is one of those mysterious properties of AC that even most electrical engineers have a hard time explaining.  A thorough technical discussion goes beyond the scope of this review (not to mention this author’s understanding). 

Ten ATX Power Supplies Compared - Cases and Cooling 110

Power factor is defined as the ratio of true power (measured in watts) to apparent power (measured in Volt Amps).  It measures how effectively AC power is being used by a device.  The difference between true power and apparent power is expressed as the power factor and results from the way true power and apparent power are measured.  Ideally we would like to have true power and apparent power equal to one another, which would result in a PF of 1.00 or 100% effective power utilization. 

True power (also referred to as working power) defines power that produces work, heat and light.  We see true power in DC circuits and in AC circuits that are powering purely resistive loads like a resistance heater or light bulb.

Apparent power is found only in AC circuits. Along with true power it also takes into account the extra power needed to create the alternating magnetic fields inside inductors (transformers and motors) and to charge capacitors.  Devices that incorporate inductance and capacitance in an AC circuit are referred to as reactive loads.  Reactive power does not produce any work.  Nearly all AC devices include some form of reactive load, which causes the PF to be less than 1 and the apparent power to always be greater than the true power.

Apparent power can easily be found by multiplying the AC voltage times the AC current using the RMS (Root Mean Square) values.  Apparent Power = Volts (RMS) x Amps (RMS) = VA

The basic formula for true power includes the power factor (the power factor compensates for the extra reactive power).  True Power = Volts (RMS) x Amps (RMS) x PF = Watts

When AC is applied to a purely resistive load, the current rises and falls in almost perfect harmony with the voltage.  Plotting a graph of the AC voltage and current will result in two classic sine waves that are in alignment with each other.  But power supplies are not resistive loads.  They include inductors and capacitors among other things, which result in a complex reactive load.  A reactive load causes the alternating current to become out of phase with the alternating voltage.  A load with predominantly inductive reactance will cause the current sine wave to lag behind the voltage sine wave by a certain amount (phase angle).  When the two sine waves are in perfect alignment (purely resistive load) the Power Factor is 1.00.  The more the current lags behind the voltage, the smaller the Power Factor value becomes.  Having the current out of phase with the voltage can also induce harmonic distortions back onto the power lines.

For example, when AC is used to power a light bulb (resistive load), electricity flows thru the filament and is converted into heat and light.  Under these conditions the true power and apparent power are essentially the same so the PF ~ 1.  All of the incoming AC power is being effectively used.  If that same AC has to first go thru a transformer or power supply (reactive load) before reaching the light bulb filament, extra current is required to create the magnetic fields inside the inductors and keep the PSU’s capacitors charged. 

The added reactive power causes the apparent power to now be greater than the true power, which in turn decreases the PF to something less than 1.00.  This extra reactive power does no real work so is factored out (power factor) when the true power of the circuit is measured.

Switching mode power supplies can have a detrimental affect on the AC mains because of the reactive load they induce and by the harmonic distortions they generate.  This type of power supply can draw highly distorted current from the AC mains, which may adversely affect other equipment on the same circuit.

Reactance and waveform distortions work together causing the PSU to draw more power than is actually converted into DC power and heat.  With a PF of 0.66, a power supply will use 50% more power (100 watts / 150 VA = 0.66 PF) than it converts into DC power and heat. 

Ten ATX Power Supplies Compared - Cases and Cooling 111 Ten ATX Power Supplies Compared - Cases and Cooling 112

Some PC switching mode PSUs contain Power Factor Correction (PFC) components and circuits.  Passive PFC typically adds a capacitor onto the AC input while Active PFC incorporates more sophisticated circuitry.  Active PFC is usually only found in higher wattage and more expensive units and is required in many European Union countries.  A power supply that does not have any PFC will normally exhibit a PF of <0.70 and will generate significant harmonics, which can distort the AC source waveform.  Power supplies with active PFC will have power factors >0.95 with minimal harmonics. 

What this really means to the end user is that a PSU with PFC will pull less current from the AC mains to generate the same amount of DC power as a similar non-PFC unit.  For commercial users who are billed based on VA usage and PF, this may save you money.  However, most residential users are billed per kilowatt-hour, which ignores the load’s reactive power component.  Because of this a PSU with active PFC won’t save the typical residential user any money on their electric bills.  However if you have a room full of computers operating on a single circuit, equipping them with PFC power supplies will draw significantly less current (therefore allowing more computers 🙂

Ten ATX Power Supplies Compared - Cases and Cooling 113

OK, that was a lot of theory – let’s see how all of this applies to our ten ATX power supplies.  I measured the AC Power Factor of all the PSUs with a WattsUp? Pro power analyzer.  The following table lists the results.

Ten ATX Power Supplies Compared - Cases and Cooling 114

The only two units in this roundup that incorporate active power factor correction are the PC Power & Cooling Turbo-Cool 510 ATX-PFC and the Zalman ZM400A-APF.  Both of these PSUs produced an almost perfect power factor of 0.99.  All others had a power factor of ~0.70.

The AC current draw (apparent power) was noticeably less on the two PFC units as compared to all the others.  While at the same time, the AC watts (true power) pulled by all the units was approximately the same. 

The PSUs without PFC drew an average of 4.2~4.5 Amps on the AC mains while operating under load.  By comparison, the Turbo-Cool 510-PFC pulled 3.12 Amps and the Zalman Noiselss 400-APF pulled only 3.03 Amps while supporting the same DC load.  Once again it is the power factor that primarily causes the difference between the non-PFC and PFC input current values. 

Since the input voltage is the same for both PFC and non-PFC power supplies, we can see the relationship between apparent power and true power by just looking at current and the PF.

Ten ATX Power Supplies Compared - Cases and Cooling 115

One final note on power factor: a power supply that incorporates active PFC circuitry might seem to be more efficient than one that does not have PFC. It’s true a PFC power supply will pull less current than one without PFC but because of the way power supply efficiency is calculated, it is technically not more efficient than a comparable unit without PFC. This is because true power (watts) is used in the efficiency calculation, not apparent power (VA).  A power supply with active PFC is more effective at converting electrical power but is not necessarily more efficient.

A power supply with active PFC is more environmentally friendly (doesn’t pollute the transmission grid) and will use less current, but it will not save you money unless you are a commercial user whose bill is based on PF and usage.  As we will see in the next section, the two power supplies with active PFC were not any more efficient than the non-PFC units even though they pulled substantially less current while operating under the same load.

« PreviousNext »