ltspice simulation of royer converter

Hi there - I am looking at using a royer converter in an upcoming design. Jim Williams' work has been incredibly helpful. Anyways, I was hoping to simulate the thing to understand it better. A friend and I tried to get a simulation working yesterday, but I don't think we got it just right. Specifically, the output was fine - but I think our efficiency was shot because both transistors were always conducting. So we were seeing base currents oscillate from something like 0.5A to

1A. I would think you'd really want them full on and full off, not partially on and fully on!

Anyways - does anybody have a ltspice simulation of a full royer converter that works?

Additionally, I was a little confused about how to make the transformer in ltspice. So I was trying to simulate the Coiltronics CTX210605-R. Problem is - they don't seem to spec the feedback coil in the datasheet!

formatting link

I am guessing that the feedback coil is the same inductance (and thus windings) as the primary, but is just a really fine gauge wire to keep size down (and thus would have a very high resistance like the secondary). But I can't find a spec for that anywhere!!

Any help would be greatly appreciated. Hope everybody's having a great weekend!

-Michael

The feedback winding requirement for the Royer is dependant on the volts per turn developed on the primary, the drive requirements of the type semiconductors being used as switches, he drive method intended and the biasing method intended for start-up.

An inquiry to the transformer vendor is in order, for part information or an application example. Zetex used to offer some useful app notes for simple self-oscillating Royer circuits for CCFL apps, some running on simple battery voltages for portable and marine use. One example:

formatting link

As the Royer operating frequency is determined by the saturation period of the power transformer, the primary Et must be known in order to determine the correct input voltage/frequency relationship, prior to fixing the turns ratio for the desired output voltage.

More efficient variations of the self oscillating push-pull converter will use another smaller saturating element to determine operating frequency and to enforce cleaner switching.

Modeling is another issue - it will require a transformer model that provides a realistic saturation performance. The SWCADIII package includes a model labelled Royer in it's default installation '\\examples\\educational' folder. This manages to stick an IC in there somewhere and does not use the characteristic transformer saturation that identifies the Royer.