Small switch-mode PSUs - Poor PF or something else?

Below 60 watts there is no requirement to consider power factor in the requirements for CE marking. So For LEDs a series capacitor to drop the volts and then a diode bridge with the LEDs as the load.yes it  is a poor PF but at least it is more or less a sine wave and will probably be balanced out by the fridge motor to a degree.

For small switchers USB chargers etc direct rectification feeds a capacitor to provide a DC bus for the oscillator that drives the ferrite transformer. Direct rectification means no current flows except at the top 10-20 degrees of each half cycle. Adding PFC to turn short pulses into a longer more sinusoidal splurge is hard, and needs is more components, so in turn more unreliability and more heat to be got rid of so why bother if it is not mandated ?

Mike.

That certainly makes sense.  There cannot be much inside the LED string one since it is featherweight and that includes an optional 6 hours on 18 hours off option via the push-button switch!

In comparison, PF wise, how do the now unavailable linear PSUs compare? I am thinking of those for routers etc - the linear PSU that my DrayTek ADSL router - circa 2006 - had, weighed a "ton".

Clive

Quite variable - most iron core designs had a 50hz transformer and then a rectifier on the secondary, If the rectifier fed directly into a smoothing cap the same 'conduct at the crest of the waves' applies but the series L & R of the transformer conspire to limit the peak current and to spread the conduction angle, especially the rather thinly optimized transformers found in 'wall wart' adapters,   so the conduction angle is longer 30-40 degs perhaps, and the diode switch off is less snappy and RFI prone.
so PF better than some,  but not great.

Mike

Most small switched mode power supplies have a poor power factor. And unlike "classic" poor power factor there is no simple way to improve this.

A lagging power factor is easily dealt with by adding suitable capacitors in parallel  with the load.

A classic leading or capacitive power factor is rare and is generally tolerated or even welcomed as it helps to correct the generally lagging power factor present on most non domestic installations.

A SMPSU draws current in phase with the mains voltage, it is NOT significantly displaced  with regard to the supply voltage, but current is only drawn when the mains voltage is close to the peak voltage. At other points in the mains cycle no current is drawn.

If such loads are significant then high peak currents at the peak of each sine wave tend to depress the peak voltage and result in the mains voltage becoming a flat topped semi-sinewave. 

A SMPSU draws current in phase with the mains voltage, it is NOT significantly displaced  with regard to the supply voltage, but current is only drawn when the mains voltage is close to the peak voltage. At other points in the mains cycle no current is drawn.

Agree totally with that. And there is an additional problem of the hot neutral with 3 phase supplies- while 3 equal resistive loads, cancel in the neutral, and so do 3 identical capacitors or 3 inductors, the waveforms from 3 identical SMPS, if you have one per phase do not overlap in time and cannot cancel. The neutral sees a 3 times as many pulses of current per cycle than any one phase, and if that is the only sort of load you have then the neutral needs to be sized for 3 times the current of any one line.

These funny waveforms are sometimes called 'triplen' harmonic but really all this means is that a spectrum of continuous  waves and their harmonics is a most awkward way to think about it, and to consider the pulses in the time domain is easier.

  This is one of the reasons that for high power stuff some effort has made to correct the power factor but it involves tricking the electronics to provide a varying load that has a draw current that is more or less proportional to the voltage of the incoming AC, as that is what a resistor does.

In effect there is rectification without smoothing,  a step up circuit whose control follows the input  voltage and then from the DC that generates, a large reservoir cap at perhaps 300-400V a more or less normal step down SMPS.

Mike