Measuring the PC power consumption
It is not always easy to determine the consumed power, especially when it a electronic load on an alternating voltage concerns. A simple measurement of the voltage and current with a multimeter and multiplying it isn't sufficient. This will be demonstrated on the basis of a power measurement on a PC.
At the mains terminal of a desktop PC the voltage (green trace) and the current (blue trace) are measured. It stands out that the mains voltage isn't so sinusoidal as may expected. The degree of the distortion is dependently on the kind of district (residential, industrial) one is located. Second, the current drawn is far from sinusoidal. The current flows only at the moments where the voltage is maximum, and has therefore a distinctive impulse shape. These kind of impulse shaped currents are very common for electronic equipment. Nowadays more and more devices will draw a current that is sinusoidal, but especially low budget will still draw highly distorted currents. One may never assume that an unknown device draws an clean sinusoidal current. The phase shift between the voltage and current isn't exactly 0º, if there is any phase relation in such a case. Measuring the cosine-phi is therefore meaningless.
If the voltage and current is measured with individual multimeters than the values of the voltage and current would be 228 VRMS and 466 mARMS. Multiplying these two values would result in a power reading of 106.25 W. Such a measurement with individual meters doesn't take into account that the current mainly flows during the voltage maximums, and beside these moments the current is almost zero.
On the other hand a real Watt meter will show a reliable value. Just as the measurement with the digital oscilloscope in this example, a wattmeter will multiply the instantaneous values and shows the average value of this. The result of the power measurement in this example is 69,9 W.
A power measurement with multimeters is totally unreliable. One can not speak of a small deviation: 106,25 W while the real power consumption is 69,9 W, a measure fault of no less than 52 %! There is no difference if there is measured with a True-RMS or a cheaper multimeter. Powers can't always be measured with multimeters.