Power factor correction
in [Design]
|
From: Hammy on 29 Dec 2009 13:38 On Tue, 29 Dec 2009 13:07:50 -0500, Hammy <spam(a)spam.com> wrote: >On Tue, 29 Dec 2009 07:17:31 -0600, "Tim Williams" ><tmoranwms(a)charter.net> wrote: > >>PFC is a flyback thing. But flybacks suck over 100W, where forward looks >>better. But forward sucks over wide duty cycle ranges. What do they do, >>just bigger flybacks? Nothing at all? >> >>Tim >They use CCM Boost topology for PFC. > >See NCP1654 for example. > >http://www.onsemi.com/PowerSolutions/product.do?id=NCP1654 Coilcraft has the inductor for the Onsemi APFC. http://www.coilcraft.com/prod_pfc.cfm
From: Fred Bartoli on 29 Dec 2009 14:20 Tim Williams a �crit : > PFC is a flyback thing. But flybacks suck over 100W, where forward looks > better. But forward sucks over wide duty cycle ranges. What do they do, > just bigger flybacks? Nothing at all? > > Tim > Flybacks work great in PFC but have pulsed input current which isn't so great. The usual PFC is a boost, but you don't get isolation and have to manage inrush current. I once did a 250W isolated sepic PFC with zero input ripple. Worked great with none of the above pb. I think I've made a couple of posts here. 't was circa 2001-2002. -- Thanks, Fred.
From: Tim Williams on 29 Dec 2009 23:54 "Jan Panteltje" <pNaonStpealmtje(a)yahoo.com> wrote in message news:hhd149$b7b$1(a)news.albasani.net... > For much higher power 3 phase rectifiers are used. > That leaves only a small ripple. > I am not sure if you need PFC in such a case. Maybe, maybe not. You get 60 degree conduction, which isn't quite 90, so maybe it would benefit some. Where cap input filters are used, conduction can still be narrower and peakier. What I'm most interested in is single phase, since there's a small band between 2kVA-20kVA where single phase is available. My welder runs from a 240V, 50A circuit, for instance. Just imagine how many home machinists have their shops wired with, say, a 240V 100A circuit, and have all their machines running from a rotary phase converter because nobody mades single phase motors over 1-2 HP. Three phase simply isn't an option for residential in the U.S.. So I guess flyback (well, boost) is still the way to go? Just more phases? Three phase I suppose would be a good start for this power level? I don't like the idea of CCM, even at 200kHz those inductors will be bigger than the rest of the circuit after it! Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
From: MooseFET on 30 Dec 2009 09:50 On Dec 29, 8:54 pm, "Tim Williams" <tmoran...(a)charter.net> wrote: > "Jan Panteltje" <pNaonStpealm...(a)yahoo.com> wrote in message > > news:hhd149$b7b$1(a)news.albasani.net... > > > For much higher power 3 phase rectifiers are used. > > That leaves only a small ripple. > > I am not sure if you need PFC in such a case. > > Maybe, maybe not. You get 60 degree conduction, which isn't quite 90, so > maybe it would benefit some. Where cap input filters are used, conduction > can still be narrower and peakier. > > What I'm most interested in is single phase, since there's a small band > between 2kVA-20kVA where single phase is available. My welder runs from a > 240V, 50A circuit, for instance. Just imagine how many home machinists have > their shops wired with, say, a 240V 100A circuit, and have all their > machines running from a rotary phase converter because nobody mades single > phase motors over 1-2 HP. Three phase simply isn't an option for > residential in the U.S.. > > So I guess flyback (well, boost) is still the way to go? Just more phases? > Three phase I suppose would be a good start for this power level? I don't > like the idea of CCM, even at 200kHz those inductors will be bigger than the > rest of the circuit after it! You can make the current go continuous at the peaks and discontinuous at lower points. It makes the control a bit harder but makes the inductor a lot smaller. You can also make the frequency rise as you go to the lower voltages. The frequency and inductance can be optimized for near the peak voltage. Since this is where most of the power transfers, the efficiency is dominated by it. You can then raise the frequency as you go to lower voltages to semi-optimize at each voltage point. > > Tim > > -- > Deep Friar: a very philosophical monk. > Website:http://webpages.charter.net/dawill/tmoranwms
From: Tim Williams on 30 Dec 2009 10:58
"MooseFET" <kensmith(a)rahul.net> wrote in message news:3ae8196e-a5ee-418b-a543-f9496a6fa614(a)d20g2000yqh.googlegroups.com... > You can also make the frequency rise as you go to the lower voltages. > The frequency and inductance can be optimized for near the peak > voltage. Since this is where most of the power transfers, the > efficiency is dominated by it. You can then raise the frequency as > you go to lower voltages to semi-optimize at each voltage point. I was looking at the FAN9611 the other day. BCM is nice because it uses "all" the inductor's capacity. Now, the datasheet says it's useful for 100-1000W, but do you think it'd be good for another decade beyond that? What scares me is the frequency clamp stops at "only" 520-750kHz. I don't think I want to deal with that much dV/dt and dI/dt just yet... It doesn't look like Fairchild is making more than biphase chips... TI has the UCC28070 which looks fairly equivalent (with constant frequency instead). Offhand, I'm not finding anything with more than two phases. Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms |