More methods to salvage leakage energy in a flyback?

I dunno anything about Weinberg converter but I found this transformer assisted snubber circuit in Pressman:

Not sure how useful it would be in your app where supply voltage is not fixed. I am not expert in switch-mode.

piglet

If it does not have to flyback then I have heard of Buck mode PFC. The line current waveform will never be as good as boost topologies but I believe THD can be got low enough to pass regulations. The advantage then is the good efficiency of buck mode.

Here is my sketch:

Depending on catch diode recovery losses it may be necessary to use a non-dissipative current snub scheme around the catch diode but otherwise efficiency could be very good.

Beste piglet

h PFC.

ys

rial/Consumer products

Klaus,

Right you are, this is IEC the 61000-3-2 here in Europe, in addition design ers have to pay attention on a special requirement about noise reinjection

Habib.

This explains it pretty well:

But it has a 50% duty cycle limit which unfortunately won't work for a PFC converter.

There is also a requirement that you can't deploy, say, 20 devices at one location that are each only 50W and not have PFC. They would require PFC even if it was only 10W each and I had that situation with designs before. Otherwise everyone would try to go for such loophole and that would defeat the whole purpose.

It's about 70W and we have to maintain high efficiency at 50-100% of load but also low standby power at almost no load. It's ok, but I think a flyback solution won't be in the cards this time.

Snubbers all burn energy (here in R1) which we can't do. The only other option in a PFC is a high rail that gets clamped against and from where a secondary switcher recycles the energy to some otehr rail where energy is needed, such as the input rail. One issue with PFC is that this causes "bumps and lumps" and thus a hit in the PF because the input cap must be very small.

Thanks, but unfortunately this doesn't seem to help in achieving a power factor of 0.95 or better.

Aahhh in that situation you're absolutely right Joerg. Blessed your heart thinking about energy efficiency because we can't amuse ourselves with Energy.

Habib.

Ok i see. Don't forget about Joerg requirements focused on PF = 0.9 ... Don't know if that structure fulfill PF exi ; This kind of circuitry needs some math demonstration.

For energy efficiency it should be interresting if the Leakage inductance energy could be steered towards the entry source ... i guess, otherwise no need to place 2 MOSFETS for doing the same job as only one.

Habib.

PS : Is there on the Earth a controller for that topology ?

A single transistor flyback requires rating to at least twice supply voltage. The big advantage of the two transistor version is neither transistor is exposed to more than the supply voltage.

No special controller is needed. The high side driver is driven just the same as the low side - both transistors get switched at the same time.

piglet

Great - many thanks. piglet

Yes, just google it, plenty of hits:

Cheers

Klaus

Ok it's basically a forward controller ... I meant it's always tricky to co ntrol the top switch MOSFET in a semi-bridge structure (IR have some nice d river for that) Ok for the snubberless adavantages ... but it requires two MOSFET's and fin ally the overall efficiency may be the same than traditionnal Flyback's.

Habib.

On Friday, February 6, 2015 at 11:39:10 AM UTC+1, snipped-for-privacy@gmail.com wrote :

=0

control the top switch MOSFET in a semi-bridge structure (IR have some nice driver for that)

inally the overall efficiency may be the same than traditionnal Flyback's.

Well, not in my experience

The switches has lower voltage rating, and thus for the same money have les s RDSon. For at 70W supply, the power for the gatedrive is in the noise flo or

Cheers

Klaus

te:

=0

o control the top switch MOSFET in a semi-bridge structure (IR have some ni ce driver for that)

finally the overall efficiency may be the same than traditionnal Flyback' s.

ess RDSon. For at 70W supply, the power for the gatedrive is in the noise f loor

When you say not in your experience ... Is that mean not only based on your experience but generally ... definitively ... with math.

When you say "the power of the gatedrive is in the noise floor" you probabl y mean that Cgs for Low Vds MOSFET's is much less than other with high Vds ? Noise Floor is a bit excessive word in that case ... Anyway if so, i doubt that even if Cgs is divided by 2 the overall efficien cy will increase.

I agree Two switch topo have the great advantage when Input voltage is high and voltage ratio Output/Input high(Vflyback high) An excuse for using a m ore complex controller (Top Switch drive circuitry).

Habib.

rote:

dl=0

to control the top switch MOSFET in a semi-bridge structure (IR have some nice driver for that)

nd finally the overall efficiency may be the same than traditionnal Flybac k's.

less RDSon. For at 70W supply, the power for the gatedrive is in the noise floor

ur experience but generally ... definitively ... with math.

Math can get iffy, only as good as your equations. Sort of just using spice , only as good as the models. I did it with measurements, supported by roug ht calculations

bly mean that Cgs for Low Vds MOSFET's is much less than other with high Vd s ?

No, I mean the gate charge. Cgs doesn't tell the entire story

A little exaggeration doesn't hurt to make a point

ency will increase.

Instead of doubting, rip out the soldering iron and do some tests, you migh t learn something along the way. Over the years I have learned to go to the lab is better earlier than later, sometimes...

Cheers

Klaus

Perhaps you're using the wrong type of PFC controller. Most have bells and whistles that just get in the way.

The active clamp/switched snubber timing also requires more close attention than most are willing to give it, for stability over a wide input range.

RL