In this discussion of Power Factor Correction in AC/DC Power Supplies I did not hear much (any?) talk about Harmonic Distortion. As I understand it PFC front ends will of course maintain a unity power factor but the real important issue is that they pull current in a sinusoidal fashion proportional to the applied voltage.
In contrast the average bridge front end on an AC/DC converter pulls current only at peak when the applied voltage is greater than that stored on the front end capacitor. This causes current to be consumed in a harmonically distorted form. It is not uncommon for a PC power supply, for example, to have 130% THD current consumption. If you break the harmonics down into components you will see odd harmonics starting very large with the 3rd and diminishing in amplitude with a sinx/x function so that by the 27th harmonic or so the contribution is insignificant. All of this is then pulled against the system impedance causes voltage distortions and resonance problems. Think about your office environment 20 years ago versus today and all of the non linear loads that have been added to the system. PFC front ends are not perfect but they do extend the harmonic content out past the 27th harmonic where the contribution is not a problem. In commercial buildings with Y secondary the 3rd and 9th harmonics (the tripplens or zero sequence harmonics) do not cancel but add in the neutral causing overheating conditions and distort the voltage waveform for other loads. This is why manufactures of power distribution gear have been putting in 200% neutrals. As you all know the use of AC/DC converters is only going to increase. My reading of these PFC requiring standards is an attempt to protect a fragile power grid. There is more to this story but I think I have made my point. Regards Chris Wells Sr Des Eng. Cutler Hammer Power Distribution Components [email protected] ------------- Original Text From: C=US/A=INTERNET/DDA=ID/GrassC(a)louisville.stortek.com, on 7/11/97 2:11 PM: Nice summary Rich Charles Grasso EMC Engineer StorageTek Tel:(303)673-2908 Fax(303)661-7115 >---------- >From: Rich Nute[SMTP:[email protected]] >Sent: Thursday, July 10, 1997 5:10 PM >To: [email protected] >Cc: [email protected] >Subject: Re: EN 61000-3-2/3 In the Great Scheme of Things > > > >Hello from San Diego: > > >I appreciate Eric's point of view regarding the harmonics >standard, EN 61000-3-2. However, I do have some rebuttal >remarks. > >(Understand that the products that I am responsible for >are less than 75 watts and therefore are Class A, not Class >D, under the standard. As such, the products do not need >PFC circuits. So, I am largely unaffected by the outcome >of the CENELEC BT effective date deliberations.) > >Personally, I disagree with the need for this standard. >I equate it to saying that you shall not place an inductive >load on the power line because it will cause a phase shift >which will affect other subscribers. But, power suppliers >do not say this, instead quietly installing capacitors at >strategic points on their own power lines to correct the >phase shift caused by customers' inductive loads. > >Power suppliers could likewise correct for non-linear loads >by installing "zig-zag" transformers at strategic points on >their own power lines. I'm sure these are more expensive >than capacitors, hence they don't want to do this. > >The problem could be minimized by constructing distribution >systems with much smaller LV loads and using more, yet >smaller, distribution transformers. In Europe, the LV >distribution can be as high as 600 homes, while in North >America the LV distribution is typically 8 homes! > >So, I don't like the power suppliers saying, on the one >hand, "We will correct for whatever inductive load you >connect to the system," while, on the other hand, "You >will correct for whatever non-linear load you connect to >the system." > >June 1, 1998, has been the "effective" date since the >standard was announced nearly 4 years ago. The date has >NEVER been changed, although there have been many attempts >to change (delay) it. > >As of today, effort is continuing to change the effective >date to January 1, 2001. I have NO knowledge as to who >is driving the proposal to change the date, nor do I know >the process, nor do I have any knowledge that suggests >a probability for adoption. > >The antics of those who have proposed a delayed effective >date together with the antics of the CENELEC BT have indeed >confused the situation. Nevertheless, there has been NO >official word at any time as to any change to the June 1, >1998 effective date. > >Here's an anecdote. In 1994, we started a new product >with lifetime expected to extend beyond June 1, 1998. We >developed two power supplies, one with PFC and one without >PFC. Our plan is to use the non-PFC until early 1998, >and then switch to the PFC power supply. The two power >supplies use essentially the same board, the non-PFC >having fewer parts. We continue to use the non-PFC after >June 1, 1998, for non-CE markets. > >In all of my postings to this group, I have been very >careful to say that you should not rely on my reports of >the efforts to delay the effective date. In my opinion, >a manufacturer must deal with the "official" effective >dates first, and have a backup plan to deal with a date >change. > >The current status is that June 1, 1998, is the effective >date for the standard. > >The risk of non-compliance is no or delayed sales in the >EC. > >The risk of early compliance is higher prices. > >The risk of a back-up plan is higher development costs. > >I don't like any of the choices! > > >Best regards, >Rich > > > >
