Why bother with a BCM Boost PFC?

[grizedale]

Hello,

The Boundary Condution Mode Boost PFC is good because you avoid harsh reverse recovery problem in the boost diode.

...but SiC Schottky diodes of 500V rating are cheap now.

So why not always do Continuous mode boost PFC?

...After all, with CCM Boost PFC the inductor is cheaper as it doesnt need an extra winidng to indicate when the inductor has discharged.

 

[RCinFLA]

You need to flush the coil on downward slope of AC input cycle to get sinusoidal current profile. Also during load variations you cannot get caught with too much energy in the coil for CCM.

 

[grizedale]

But I dont think sudden load removal is a problem for CCM Boost PFC.

PFC controllers detect overvoltage and immediately stop switching,

-then, whatever energy thats in the PFC inductor simply discharges into the output electrolytic...
-that electrolytic is always going to be bug enough to absorb the energy from the PFC inductor without going miles overvoltage.

 

[mtwieg]

The Boundary Condution Mode Boost PFC is good because you avoid harsh reverse recovery problem in the boost diode.

Diode reverse recovery isn't the only source of turn on losses though.

...but SiC Schottky diodes of 500V rating are cheap now.

They're certainly not as cheap as silicon equivalents. They probably never will be, at least for several more years. As long as that's true a lot of people won't use them.

So why not always do Continuous mode boost PFC?

...After all, with CCM Boost PFC the inductor is cheaper as it doesnt need an extra winidng to indicate when the inductor has discharged.

An extra sense winding should be negligible to the cost of the inductor. And there's no reason you can't use a sense resistor in a BCM supply.

BCM/TCM converters also have the advantage that their conducted interference has an inherently spread spectrum, which can make it easier to meet interference specifications, despite the rms ripple actually being higher.

 

[doraemon2011]

BCM/TCM converters also have the advantage that their conducted interference has an inherently spread spectrum, which can make it easier to meet interference specifications, despite the rms ripple actually being higher.

Would you like give us a further explanation about "inherently spread spectrum",
thank you very much!

 

[mtwieg]

TCM converters operate with variable frequency, and with a PFC the frequency changes greatly over the course of one half line cycle. So that mean the spectral density of the conducted current will be spread over a large bandwidth, rather than being concentrated at one frequency (plus harmonics), as is the case with a constant frequency converter.

 

[grizedale]

though strangely constant frequency PFC stages and SMPS's are said to be easier to provide EMC filtration for.

 

[mtwieg]

though strangely constant frequency PFC stages and SMPS's are said to be easier to provide EMC filtration for.

Possibly because you can use narrowband EMI filters with a notch filter characteristic, giving very high attenuation at the fundamental frequency. Though that won't work if the user wants to sync it to run at a different frequency.

Overall the best solution for (conducted, at least) EMC is probably a CCM converter with some frequency dithering built in, so that you get low rms ripple and also a spread spectrum. Nowadays many chips offer frequency dithering.

 

[doraemon2011]

TCM converters operate with variable frequency, and with a PFC the frequency changes greatly over the course of one half line cycle. So that mean the spectral density of the conducted current will be spread over a large bandwidth, rather than being concentrated at one frequency (plus harmonics), as is the case with a constant frequency converter.

My English is not good enough, maybe there are some places I misinterpreted.
The whole time, I was thinking that constant frequency converter can got a EMI filter easiler than variable frequency
converter, because of the fixed switching frequency follows some fixed interference frequency.