Author Topic: Advice for better understanding PFC with regard to EN/IEC61000-3-2/12  (Read 518 times)

0 Members and 1 Guest are viewing this topic.

Offline lordnoxx

  • Contributor
  • Posts: 22
  • Country: de
Hi everybody,

the international and European standards EN/IEC61000-3-2 and EN/IEC61000-3-12 state limits for harmonics in AC mains currents. It is well known that these are best to achieve with some sort of PFC. But let's assume an active PFC here. Searching the web for active PFC always shows that not only the THD of the current is optimized but also the displacement power factor i.e. cos-phi is forced to be 1.

From an efficiency point of few it is obvious why cos-phi has to be controlled to be 1 (no reactive power) but did I get it right that the mentioned standards do not call/require a cos-phi of 1. As mentioned the standards only set limits on current harmonics. So in theory I could operate an active PFC with cos-phi=0 (only reactive power, 90° phase shift) at 75Arms and close to perfect THD (0%) and still pass the standard? ???

A simple ohmic resistor placed on the mains would have perfect behavior with regards to the current drawn. At least that is what can be read on many places when searching the web: No Phase shift (only active power), no harmonics. But that is no longer true when already the mains voltage waveform is distorted. In most areas of the world in nowadays the mains voltage sine wave is no longer a perfect sinusoidal shape. Instead it is flat-topped and thus introduces harmonics itself.

If an active PFC follows the standard an draws a perfect sine shaped current from the mains, does it then contribute even more to the flatness of the voltage waveform? I mean with an PFC there would be more current drawn near the flat region of the voltage waveform when compared to a simple resistor which has its current followed the flat-top shape of the voltage. So shouldn't a PFC behave more like a simple resistor and let the current follow the voltage shape?

In fact there are two approaches PFC controllers implement:

1. Using the mains voltage to get a reference time base and feeding that into a PLL. The PLL then outputs a phase-locked sinusoidal reference waveform for the current controller --> perfect sinusoidal mains current with no harmonics

2. Taking the mains voltage (with some scaling and other calculations performed on it) as a direct reference for the current controller --> non-perfect sinusoidal mains current with harmonics. Because the flat-top shape is also present in the reference. But the PFC behaves as a perfect resistive load "seen" from the mains.

So what is the way to go to meet the standard's requirements? Is it the perfect no-THD sinusoidal current, or is it the PFC that behaves like a simple and variable resistor with some harmonics? And why does the standard not talk about cos-phi in particular?

Thanks and cheers
Find out what you cannot do and then go an do it!
The following users thanked this post: Alti

Offline Faringdon

  • Super Contributor
  • ***
  • Posts: 1161
  • Country: gb
If an active PFC follows the standard an draws a perfect sine shaped current from the mains, does it then contribute even more to the flatness of the voltage waveform?
I know what you mean, yes, there is more reactive power if an active PFC draws perfect sine current when the mains voltage is flat-topped.
....But if all PSU's were active PFC'd, then you woudlnt get flat topping (not nowhere near so much anyway).......because AYK, flat topping is mostly caused by the multiple use of "txfmr/rect/smoothing cap" power supplies being used....which have poor PF.
I think the EMC test places have to monitor flat-topping whilst they are testing your product, and if its too much, then they have to invalidate the test......otherwise, 1 or 2 would get you a pass as the  test limits are not that tight......PFC approvals is  brought in to reduce flat-topping so woudlnt let flat-topping "rain on its parade" so-to-speak.

Yes i think you can draw high current at cos phi  = 0 and have no THD and it passes.....but there is no equipment that generally would be available for someone  to do that i believe.

AYK, industrial customers are allowed cos phi = 0 but have to pay for all reactive current drawn....Or more accurately, they pay for their total "Apparent power" drawn....Apparent power being the complex combo of active and reactive power.

AYK, the cheapest way to do domestic PFC for residential type products is to use a boost PFC, and use the divided down mains itself as the "sinusoidal reference"...when you do this you pretty much end up with most of your input current (the fundamental) being in phase with the mains voltage, so the points you raise dont really come into it that much, though academically speaking,  your points are spot on.

The PLL method you speak of is convenient to do if you are doing PFC with a micro.....but in truth.....who wants to design a PFC for a mass market product that needs a software engineer to be involved in all testing and maintenance of the design, when you can just use a cheap offtheshelf Boost PFC controller...which uses the mains as a reference....and get away cheaply like that , perfectly satisfactorily.
« Last Edit: June 22, 2022, 03:40:14 pm by Faringdon »

Offline jonpaul

  • Super Contributor
  • ***
  • Posts: 1747
  • Country: fr

One effect of current harmonics is to overload the neutral conducteur of a three phase power installation.

A boost PFC synthesizes the mains current wave to be a sinewave or to mirror the mains voltage waveform, thus reducing the harmonics and bringing power factor nearly to 1.0.

But, The OP post has some misconceptions and assumptions. I advise you to study the topic a bit....

The motivation and design for compliance with the IEC 61000 3-2 is well covered in books, papers, app notes and seminars.

Bon courage

« Last Edit: June 22, 2022, 08:39:05 pm by jonpaul »
Jean-Paul (EE 1968, the Internet Dinosaur)

Offline f4eru

  • Frequent Contributor
  • **
  • Posts: 974
  • Country: 00
    • Chargehanger
The question is valid and very interesting.
What happens when cosphi is very low, but there are no harmonics, like with a capacitive dropper scheme....

Worst case : A big commercial building with a lot of lights, each one capadropper based.

Offline Marco

  • Super Contributor
  • ***
  • Posts: 6022
  • Country: nl
EU seems to have realized there were some perverse incentives there, they switched to Displacement Factor per harmonic to judge power quality instead.

Though LED lighting below 25W is exempt, so a whole flat full of capacitive dropper lighting is still a head ache for power companies.
« Last Edit: June 23, 2022, 04:43:04 pm by Marco »

Offline Faringdon

  • Super Contributor
  • ***
  • Posts: 1161
  • Country: gb
Yes thats right, and even then it leaves us wondering, because its only the fundamental (50Hz) that contributes to active power the displacement factor of the harmonics is actually irrelevant, but yes, it was brought in as told.

To be honest, i think cap droppers are not so common now, specially above a few watts......ive took apart 5w bulbs and they contain pretty well PFC'd cheap offline Buck controllers......what is the company name "staybright"?....."brightLED"?.......I  forget now, but they make  these buck controllers cheap....running in BCM, and giving good PF, because the input cap after bridge is low.
...actaully i think its "brightpower"?
« Last Edit: June 23, 2022, 09:27:18 pm by Faringdon »

Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo