Buck-Boost PFC Peak Current in Critical Conduction Mode

[Endymion]

I am considering to use Buck-Boost topology for Power Factor Correction but I'm having trouble finding much literature on it. My appliance is an LED fixture so it will always have a (fairly) constant load (150mA @ 30V). Since the input voltage range is going to be between 100-240 VAC and the output is 30V, using a Boost Converter with PFC is a bit unfeasible. I would prefer if the solution is single-stage due to space and cost. I want to avoud using a Flyback topology because at 240V input it will place a huge voltage stress on the power switch.

From looking at the waveform (page 10 and 11) of a Boost topology working in Critical Conduction Mode, the peak current is:
Ipk=(2 x √2 x Pin)/Vrms
My question is, does this result hold for Buck-Boost as well? I feel that it should because it is based on the waveforms and not on the topology. Another way to look at this would be that during ton, the Buck-Boost and the Boost topology's inductor both have Vin applied across them.

I am actually not considering the above linked part. I am considering Richtek RT7302. Unfortunately, for this IC, most of the literature is for Flyback topology and not for Buck-Boost - including their Excel Design Tool.

power-factor-correction

 

[biswaIITH]

I am considering to use Buck-Boost topology for Power Factor Correction but I'm having trouble finding much literature on it. My appliance is an LED fixture so it will always have a (fairly) constant load (150mA @ 30V). Since the input voltage range is going to be between 100-240 VAC and the output is 30V, using a Boost Converter with PFC is a bit unfeasible. I would prefer if the solution is single-stage due to space and cost. I want to avoud using a Flyback topology because at 240V input it will place a huge voltage stress on the power switch.

From looking at the waveform (page 10 and 11) of a Boost topology working in Critical Conduction Mode, the peak current is:
Ipk=(2 x √2 x Pin)/Vrms
My question is, does this result hold for Buck-Boost as well? I feel that it should because it is based on the waveforms and not on the topology. Another way to look at this would be that during ton, the Buck-Boost and the Boost topology's inductor both have Vin applied across them.

I am actually not considering the above linked part. I am considering Richtek RT7302. Unfortunately, for this IC, most of the literature is for Flyback topology and not for Buck-Boost - including their Excel Design Tool.

power-factor-correction

How r u operating the PFC converter???In DCM hopefully..u have to include the ON period there..The peak (if u r operating in DCM) is

Ipeak=(Vmsinwt*D1Ts)/L...
Where D1Ts is the on period...

The analysis with the Flyback is almost similar to buck-boost

 

[Endymion]

I'm operating in Critical Conduction Mode.

 

[biswaIITH]

I'm operating in Critical Conduction Mode.

Yes the above expression applies to Critical conduction mode also...

 

[Endymion]

Yes, it's the standard inductor expression. I understand that.

However, the question isn't what will be the inductor current, it's what should be the inductor current so that the average of that current appears in-phase with the voltage waveform for PFC.

 

[FvM]

The topology designated "buck-boost" in the RT7302 datasheet isn't but an non-isolated flyback. It saves some transformer copper but the primary waveforms are essentially the same as with regular isolated flyback. It reduces the "huge voltage stress on the power switch" only in so far that additional overvoltage caused by the leakage inductance can be avoided.

Designing a non-isolated flyback for 30V output however forces a very low primary duty cycle and respective high peak current. Although you can probably chose a lower transistor voltage class, switching losses will be most likely higher.

 

[biswaIITH]

Would you be kind enough to point me to where you found this expression? Just want to derive it to understand it.

U can derive it...When the mosfet is on,the voltage across the input inductor is mod of VmSinwt....Now u can apply the universal V=L*(di/dt) across the inductor...So long as the switch will be on,the current through the inductor will increase...So integrate current (from 0 to D1Ts for the first switching cycle or t to D1Ts+t for any switching cycle) and integrate current from 0 to Ipeak(as input current to buck-boost will be discontinuous).....u will get the expression using simple math

The input current to the buck-boost converter will be triangular(over a switching cycle Ts)in nature,u can find the avg of the input current as u wud find the avg of a triangle...
The expression will be
Iinavg= (D1^2*VmSinwt)/(2*L*Fs)
U can see the Input avg current is proportional to the input avg voltage....i.e. the input of the buck-boost is behaving like a resistor

 

[Endymion]

The topology designated "buck-boost" in the RT7302 datasheet isn't but an non-isolated flyback. It saves some transformer copper but the primary waveforms are essentially the same as with regular isolated flyback. It reduces the "huge voltage stress on the power switch" only in so far that additional overvoltage caused by the leakage inductance can be avoided.

Designing a non-isolated flyback for 30V output however forces a very low primary duty cycle and respective high peak current. Although you can probably chose a lower transistor voltage class, switching losses will be most likely higher.

Would you recommend a different topology? This is for driving LEDs so it's a fairly constant load (30V, 150mA).

 

[FvM]

I would use flyback with reasonable transformer windings ratio.

 

[Endymion]

Even if isolation isn't necessary? Thanks. This actually makes it easier because there are several designs out there that I can study. I have been reading up on this for almost a week now. Sometimes you just need to see the bigger picture!

 

[Anna Conda]

Also a sepic converter will give you PFC and an OK characteristic to your load with suitable o/p capacitance. Non isolated, can be controlled from a standard chip, can be high freq (140kHz)...