The limits permitted for Class D equipment, which is the special waveshape of current associated with a full-wave rectifier and capacitor input filter, is given in Table 3 of IEC-1000. Table 3 provides 2 limits for each harmonic, an absolute maximum and a power-weighted maximum based on the real input watts required by the device under consideration. For example, Table 3 limits for the 3d harmonic are 2.3 Amps rms max and 3.4 mA/watt. Thus a 100W product is limited to a 3d harmonic of 0.34 Amps.
A device not required to meet Class D limits is permitted to meet the limits given in Table A. Table A provides the same maximum values for each harmonic as Table D, but does not have the power-limited feature. Thus a 100W product acceptable per Table 1 is permitted a 3d harmonic of 2.3 A. This is in reality the 'old' IEC-555. The driving force behind changing the current waveshape is to avoid the special waveshape critria and permit larger harmonic values. I know it can be done---but it is costly and takes a lot of space. I forced a 200W power system into IEC-555 compliance with a 10 mH inductor about the size of a baseball. Please let me add a word about power factor. This can be very confusing since we all tend to think about the cosine of the phase/angle. One must think the more basic definition which is the ratio of effective power to apparent power. There is zero energy transferred to a load in the harmonics! The THD of the 'special waveshape' is approximately 100%. This explains the power factor of uncorrected switch-mode power supplies of about 0.5 in spite of the fact there is virtually no phase angle between the voltage and the current. The rms current in a power-factor corrected design will be about 1/2 that of the 'special waveform' design and this is one of the driving forces behind IEC-1000. Even though there is no energy delivered by the harmonics, the losses incurred in the delivery system are very real indeed.
