DIY Power Factor Correction - 1.5kW at 230VAC

[rentner]

I was looking for a power factor correction IC on mouser.com and was finding quite a lot of them. The problem with the circuits for them is, that you cannot use a simple inductor for that - the inductor has a primary that is coupled back into the IC to manage the AC sweep of the line Voltage, as far as I know.

I mean, I could wind my own transformers, but that would be quite complicated, I'd say. I don't know the ratio of the windings and also I don't know, how inductive the transformer would be...

But for a 1.5kW Supply I want to use PFC, because I gess, that can't be a bad Idea. I could theoretically use an Inductor as a passive PFC, but it would weight many kilos, so that does not make any sense. I'm more into active components but still don't know, how to get around with the inductors. Can somebody help me with that? I don't mention a specific part number, because this will count to all descrete PFC ICs.


Thanks for help.

If somebody is interested in the rest of the supply: It will be a regulated supply with 4 Outputs, done by 2 SG3525, which works perfectly for my application. Tested and success. <3 (Edit: Yes, that worked out well without an oscilloscope! )

[rentner]

Does nobody have an Answer? I don't want to distort the mains too much.

Please help me somehow.

[Harvs]

Believe it or not, winding your own inductors/transformers isn't that hard.  There seems to be a bit of "black magic" surrounding them for people who haven't tried it before.

Yes you'll need to get hold of some decent material to teach you the basics.  Personally I like the A. Pressman book "Switching Power Supply Design", which has a chapter or three on magnetics, but it's far from the only one out there.

I've also been looking at building a PFC lately, but not sure if I'll bother yet.  I think the inductor is probably going to be the easiest bit of the design to get right, the PCB layout I see as being much more difficult.

One of the best bits about being confident with winding your own magnetics is it opens up a whole new realm of circuit configurations for one off devices.

[rentner]

Winding a transformer isn't hard for me, but looking at the PFC IC's I net a transformer of unknown Windings Ratio. That's the problem. How do I wind a Transformer, where I don't even know, what the Ratio must be? I cannot find that in the Datasheets. That's, why I cannot do that.


Already wound a transformer, working fine at 20khz - 1 winding per volt. Worked very nice. (Of corse will decrease the windings and increase the frequency to 50khz, to increase efficency of the transformer - My setup with high power Push-Pull Stage will be able of propably 300khz with enought cooling of the mosfets - Output is able to push-pull over 100 Amps at 12V continuosely!)



Well, the Transformers for the power factor correction are still a problem, since I realy don't know anything about the required parameters like Windings Primary:Secondary. Can you help me with that? o.O

[rentner]

Auxiliary supply? I don't need that at all.

All I need is a way for perfect power factor correction at very high powers.

The problem is, that inductors are coupled. I don't know eather the windings ratio or the inductance ratio, so it is impossible for me, to find out, what the coupled inductors (transformers) have to be. Giving the fact, that there are many thousands of inductors out there. I cannot wind an inductor of ungiven values.

[codeboy2k]

The active PFC is simply a boost converter, with the boost voltage chosen to be above the peak rectified AC input voltage.

I've drawn a simple picture below ... I think you are talking about the inductor value for T1, which is coupled to a sense/reset winding.

You are saying you don't know the values of these windings.



The value for the inductor's main winding is chosen like any other boost converter inductor.  The PFC controller will control the gate of the switch Q1 to charge the bus capacitor C_bus up to the bus voltage; it will do so while maintaining  inductor current to closely follow the voltage cycle of the rectified input. It will gate the voltage onto the DC bus at a high frequency, like 20khz to 50khz, charging the capacitor in small bursts timed with the voltage waveform.  This high frequency ripple can be filtered out of the AC line with standard passive AC filtering before the bridge rectifier.  I didn't draw the AC line filter here, but I'm sure you've seen them before, and I've even drawn one on the forum in the past, over here.

When the PFC controller's gate is on, the induced voltage in the reset winding is sensed by the PFC controller to determine where in the cycle it is, i.e. the peaks and valleys of the cycle.  When the PFC controller switches the gate off, the inductor current collapses in the primary winding and reverses in the reset winding, which is fed back into the PFC controller and used to reset the core.

The datasheet for your chosen PFC controller should show a minimum and maximum voltage allowable for that pin's input.  On one datasheet I looked at, this was specified as min -25V, max +25V.  So you would size the turns ratio of T1 to ensure that it is no more than 25V referenced to ground, based on what you calculate your primary turns should be, which would be the number of turns required on your chosen core to get the calculated inductor value (for your bus voltage and current)

Does that help?

[NiHaoMike]

PFC is a scam for home use. In that case, the costs and efficiency penalty of using PFC usually outweigh its benefits.
http://nlcpr.com/Deceptions1.php

That said, it's best to think of a PFC inductor as an inductor that just happens to have a sense winding, since that's what it is. On an old HP I took apart, the sense winding was just a single turn.

[mkissin]

The ON Semiconductor NCP1653 controllers are very simple to use, and have an incredibly simple feedback mechanism.

They also provide an excel spreadsheet which basically does the entire PFC design for you, including boost and feedback components.

http://www.onsemi.com/PowerSolutions/product.do?id=NCP1653

Needless to say, be careful. You're going to end up with a huge capacitor with about 400V DC on it. It won't be forgiving.

[ejeffrey]

PFC is a scam for home use. In that case, the costs and efficiency penalty of using PFC usually outweigh its benefits.
http://nlcpr.com/Deceptions1.php

The article is correct, but your interpretation of it is wrong.  The article is talking about external PFC devices for correcting large inductive loads.  It very clearly states that what does make sense is for appliances to be designed to have near unity power factor, which is what the poster is asking about.

[TerraHertz]

Seen this?
http://www.ti.com/product/uc3855b
TI/Unitrode data index page for the UC3855B

http://www.ti.com/litv/pdf/slus328b
Data sheet Rev B, Oct 2005.

Notes----
http://www.ti.com/lit/an/slua146a/slua146a.pdf
Application note

http://www.ti.com/litv/pdf/slua263
AC Requirements for Power Factor Correction Circuits

http://www.ti.com/litv/pdf/slua253
Accurate PWM Duty Cycle Clamp

http://www.ti.com/litv/pdf/slua196a
DN-66 UC3854A/B and UC3855A/B Provide Power Limiting with Sinusoidal Input (Rev. A)

The UC3855B is about $5 each on ebay.