New Switching Converter Topology?

Tim, I'm not much of a switcher guy. But I don't see how you get current to flow in in that circuit. The two lines going off the screen are just for fet switching, I assume.

George H.

Another of Tim's wet dreams. ...Jim Thompson

If the switches were reverse polarity, it would be a DC fransformer with fixed voltage ratios. As is, it does nothing.

Swapping things means thatr they're still present. There's no capacitor in your drawing.

RL

Normally I am open to new topologies, just not any with the ?novel ? in the title

But I cannot really see the point of your topology

Cuk is famous for talking wonders about his in his words superior topology, but I have never heard of anyone using it and never seen a case where it w ould be close to come into consideration

To understand your topology, can you name a case in which it outperforms an y current topology?

Cheers

Klaus

We use the coupled Cuk topology in several of our products. It has some nice features.

When the transistors turn on, the inductors are connected in series, and current ramps at dI/dt = (V1 - V2) / 4L (taking L as primary inductance and the windings 1:1). When off, the inductors flyback, and one is clamped with a diode, so the inductor voltage is -min(V1, V2) until fully discharged or the transistors turn on again.

Tim

But I didn't put "novel" in the title. (Fine, "new" is the same meaning, different root, so what. :^) ) I'm expecting it's been done before, and that's what I want to know. :)

Anywhere you want a buck, that bucks in either current direction (whereas a synchronous buck is boost in reverse), where the source and load may be swapped freely (V2 > V1 or vice versa), and you don't need voltage boosting capability (where you'd need SEPIC or such).

Tim

Bidirectional flybacks, forwards and others exist:

They can be practical with systems that need a higher degree of failure safety where either side can feed the other if it fails, but otherwise remain electrically separated. Also for supercap storage or electric vehicles that are supposed to act as storage devices when needed.

Here's my first switching regulator design:

Power schottky diodes were just becoming available from Motorola, and I could only get 20 volt parts, so I hung the catch diode on the inductor center-tap. But N*I is preserved, and N changes by 2:1 when it switches, so the ripple current is big and sharp into the output cap. It worked, but the wet-slug tantalum got warm.

I think your circuit has similar high output ripple current.

The inverse of the "extended boost". :)

My circuit, as shown, has "arbitrarily low ripple" on the lower voltage side, and "full ripple" on the other. So, waveforms like a SEPIC. The two-terminal version of course has "ALR" on both, but dumps the extra power somewhere else, rather than "stirring" it back into a supply (could be useful for charging caps and starting motors?).

Tim

Ah, I drew the waveforms wrong -- an occasional downside of late night / early morning paper-drawing.

Your analogy is spot on, actually:

Tim

One does not need to go to SMPS topologies to do that, a 50Hz trafo with a synchronous bridge rectifier has all what is needed for bidirectional power transfers (+an appropriate controller, of course).

Best regards, Piotr

There are some interesting minor discoveries. For example, I have never seen an active clamping forward converter with a depletion mode MOSFET used in the place of the PMOS switch. Instead people do heroic things to steer the PMOS with negative voltage or use high-side NMOSes which obviously need an insulated gate driver.

It is my invention and I have breadboarded it with success in a 10W switcher just to see if it works.

Best regards, Piotr

That would be a bit on the heavy side for large power levels. Not so great for people with lower back pain :-)

Thanks... I didn't see the the transformer driving itself.. or whatever you call that. I've used and wound various transformers but have always been somewhat mystified.

George H.

Depends how large is large. A bidirectional power supply for a tram would definitely need to be a switcher. OTOH, my bidirectional 500W PSU based on a toroidal transformer was about 3kg, so not that bad.

Best regards, Piotr

It also depends on location and whether you are married. "What's this ugly thing doing here? Does it really have to be here?"

Does she ask whether the refrigerator "really have to be here"? Perhaps the bottom line is that you could have done a better job selecting a mate. ;-)

It's called the "Amateur's Wet Dream Topology" >:-} ...Jim Thompson