Power factor correction
in [Design]
|
From: MooseFET on 30 Dec 2009 13:36 On Dec 30, 7:58 am, "Tim Williams" <tmoran...(a)charter.net> wrote: > "MooseFET" <kensm...(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... You can never trust the makers "good for from here to there" on controller chips. The limitation is largely the result of how much current the outputs can produce. They assume "reasonable" values for the gate charge per drain current rating of the MOSFET and work out how much current that comes out to. As new MOSFETs are introduced, the ranges become sort of nonsense. > 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. The amount of advantage decreases with increasing number of phases. You can always make N copies of the 2 phase circuit to get N times the power. Making them share isn't going to be too hard to do.
From: Jon Elson on 30 Dec 2009 18:01 Tim Williams wrote: > 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.. For home shop use, a VFD can't be beat. You get variable speed, soft start and dynamic braking. Running a 5 Hp + 3-phase machine motor off one of those "phase converters" will sure dim your lights on every start. An RPC is better, but makes noise and sucks power all the time it is running. Neither gives you braking. if you have a big lathe with no mechanical clutch/brake setup, the long coast time can be an actual safety concern. My 15" Sheldon lathe would coast 20 seconds by itself, but stops in 3/4 second by the VFD. > > 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! The typical larger PFC scheme uses a boost converter that is modulated by the voltage waveform, so the current draw matches the instantaneous voltage, scaled by average current demand. This stage precedes the main converter, which would usually be a push-pull design, and you can use the topology and drive scheme of your choice in that converter. Jon
From: Alan W on 30 Dec 2009 22:27 On 12/29/2009 5:17 AM, Tim Williams 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 > I've done a 2kW CCM PFC using a boost converter - the inductor was pretty massive (several pounds at least), but then again it was running at 50 kHz. (It was DSP-based, so the control loop couldn't run much faster.) It was for a 2HP VFD, and there wasn't any size or weight constraint, otherwise I would have figured out how to use a higher frequency with a smaller inductor. I seem to remember Bi Technologies made some power modules designed for PFC converters in the 2-3kW range, and had some app notes for them. - Alan
From: Tim Williams on 31 Dec 2009 00:31 "Jon Elson" <jmelson(a)wustl.edu> wrote in message news:JridnVJ_VK2HRqbWnZ2dnUVZ_oCdnZ2d(a)giganews.com... > For home shop use, a VFD can't be beat. Unless you're cheap, which most of rec.crafts.metalworking IS. ;-) When you get a lathe for $200 and a case of beer, VFDs don't look so cheap. They're definitely nice though. As for electrical characteristics, VFDs are a fine example: what do they do for PFC? Nothing? The VFDs installed in the school's power lab are 10 or 15HP, three phase input, and according to the Fluke PQAs, current is spikey as hell, no PFC, big gulps of capacitor charging. Doesn't seem very nice to me, and when you're losing half your VAs in harmonic current, it's just not practical from a single phase supply. > The typical larger PFC scheme uses a boost converter that is modulated by > the voltage waveform, so the current draw matches the instantaneous > voltage, scaled by average current demand. This stage precedes the main > converter, which would usually be a push-pull design, and you can use the > topology and drive scheme of your choice in that converter. Yeah, but that doesn't work well at let's say 400V and 30A. The inductor is huge, and it's an awful lot of ripple on the capacitors. There's gotta be a better way (multiphase helps). Tim -- Deep Friar: a very philosophical monk. Website: http://webpages.charter.net/dawill/tmoranwms
From: JosephKK on 3 Jan 2010 14:41
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 What? Even i have done flyback at about 15 W and have heard of them used in the 10s of milliwatt range. I think i have seen PFC that was done with SEPIC in the 20 watt range. |