Valley filler (passive power factor corrector)

[eem2am]

valley fill circuit

Hello,

A valley filler (passive power factor corrector) is here...figure 1, page 1...

https://www.coilws.com/Publications/ImprVF.pdf

it is possible to use a Valley Filler circuit to smooth the DC after a rectifier. It increases the conduction angle of the diodes and thus improves the power factor.

Here is the waveform (bottom trace) of a valley fillers current....

...Do you know what the power factor of this valley filler would be?

Thankyou for listening.

 

[oddbudman]

valley fill pfc

<quote>it is possible to use a Valley Filler circuit to smooth the DC after a rectifier</quote>

Not really, valley fill if used to smooth the dc would use large capacitors and you would find power factor start to drop. Generally you would find higher ripple on a valley fill circuit vs a simple bridge-diode rectifier. (Capacitor current needs to follow the rectified line voltage to get a high power factor).

If you want smooth DC after the rectifier use a active power factor correction ic such as the L6561.

oddbudman

 

[eem2am]

valley fill passive pfc

thankyou oddbudman, though i beg you to reconsider...

valley filler has better power factor correction than normal high value cap after rectifier.

the thing is, i am woindering what the power factor would be.........increasing the value of the valley filler caps will not make the power factor worse.

the valley filler caps have less AC ripple current than a normal cap after a rectifier....this is because the conduction angle is longer for valley filler caps.

i am grateful for thoughts on the power factor of a valley filler

 

[FvM]

passive power factor correction

t is possible to use a Valley Filler circuit to smooth the DC after a rectifier.

As also your sketched waveforms and literature shows, DC is not smooth, it has about 60%-65% voltage ripple.

i am wondering what the power factor would be

The literature link has FFT respectively THD measurements, that can be easily converted to a power factor. The real value is somewhat higher beacuse too few harmonics are used in calculation, but it's good for an estimation. The best circuit has a power factor of about 0.97, mainly caused by displacement apparent power (cos φ).

P.S.: Your idealized sketched waveform misses the current peak that i still present with basic valley filler circuit and causes considerable higher order harmonics. To my opinion, nonlinear circuit simulation with any of the available SPICE variants is a good tool to evaluate the properties of this circuit. It allows precise determination of waveforms and extraction of parameters as power factor, THD, voltage ripple, whatever.

 

[eem2am]

valley-fill circuit

FvM, thankyou for replying,

interestingly i did simulate a valley filler using a SPICE program and it showed the valley fill caps discharging just after the mains peak voltage which is not physically possible. I wrote to their support site but got no answer.

Anyway, i am believing that many companies are wrongly stating power factor as displacement factor (cos phi) alone. As you know, power factor is the product of displacement and distortion factors......with a pure sine, distortion factor is 1 and so power factor ends up being the same number as the displacement factor.

...if you please see this app note ...

http://www.irf.com/technical-info/appnotes/an-1074.pdf

on page 2 they are stating the power factor of a valley filler as baing 0.964.

To me, this is ridiculous, a valley filler could not possibly give power factor of 0.964, and i am wondering what the power factor really is with a valley filler....

i will look into your suggestion of calculating it myself from the harmonics.

 

[FvM]

passive power factor

To me, this is ridiculous, a valley filler could not possibly give power factor of 0.964

It does, exactly. The waveform already suggests a power factor value near to 1.

Also without recalculating the IRF AN details, I'm convinced, that the power factor results are correct. They match the results from your previous literature link, that basically promotes the same circuit voltage doubler and additional series resistor. In addition, the IRF AN systematically varies the series resistor and shows what must be done to meet certain power quality standards. You'll notice, that the current waveform of the first circuit is nearly indentical to the best circuit in the literature and the power factor similar to my caculated value of 0.97.

i am believing that many companies are wrongly stating power factor as displacement factor (cos phi) alone.

I don't. In case of the present AN, although the displacement (phasee angle of fundamental) isn't given, the given power factor results are only correct if both terms are considered.

As I mentioned in another discussion, you'll have difficulties to find an instrument that is reading a pure displacement power factor. It's very easy to determine P, Irms and Vrms in digital signal processing and calculate the true power factor. Displacement would require an additional measurement or calculations.

i did simulate a valley filler using a SPICE program and it showed the valley fill caps discharging just after the mains peak voltage which is not physically possible

Although you can achieve a lot of interesting artefacts with inappropriate SPICE simulation setup. It's a real simple simulation circuit and should work from the scratch, normally.

Finally, to take up a previous statement from the discussion. An active PFC has still some important advantanges. It can hardly be avoided, if a low ripple output voltage is needed. But that's another discussion.

 

[vicbsee]

Power Factors of 0.90 and higher using Valley Fill are achievable, however only with a resistive load. As soon as the load becomes reactive, the PF begins to drop drastically.

I wanted to use VF for a SMPS that I am working on right now, however, the large ripple was limiting the Pout of the supply.

Great idea in theory, but practically, not so great. If I had the real estate needed for an active PFC circuit, I would use that instead.

Victor