Interesting self-oscillating half-bridge article

[profdc9]

I found this interesting self-oscillating half-bridge article

Bal, G., Oncu, S., & Ozbas, E. (2013). Self-Oscillated Induction Heater for Absorption Cooler. Elektronika Ir Elektrotechnika, 19(10), 45-48. https://doi.org/10.5755/j01.eee.19.10.5894

https://eejournal.ktu.lt/index.php/elt/article/view/5894

It uses a gate drive transformer in series with the load to drive the gates of the half bridge, with the two upper and lower gates being driven in opposite phases.  An inductor Lss (see figure 1) is included so that the oscillation frequency is somewhat higher than the resonance frequency, achieving ZVS.

It looks like a particularly simple setup for a self-oscillating half bridge.  If anyone has experience with this perhaps they could comment?

[davekni]

First time I've seen that circuit implemented with IGBTs.  I've used it in both half and full bridges with BJTs.  For BJTs, that simple circuit is convenient.  CT ratio defines the forced-beta.  With a 10:1 CT, base current is always 10% of collector current.  I've heard that some CFLs use the BJT version.

[profdc9]

For a full-bridge version of this, does one do the obvious thing and just feed the two halves of the bridge with opposite phases of the gate transformer?  And is the current limited in the circuit mostly by shifting the oscillation frequency to be further above resonance, as it does not appear that a series inductor with the load can be used to limit the current because the resonance frequency and therefore oscillation frequency will shift down accordingly?

First time I've seen that circuit implemented with IGBTs.  I've used it in both half and full bridges with BJTs.  For BJTs, that simple circuit is convenient.  CT ratio defines the forced-beta.  With a 10:1 CT, base current is always 10% of collector current.  I've heard that some CFLs use the BJT version.

[davekni]

"R" of the induction coil is the only current limiting.  It is a series-resonant circuit.  A high-Q load will draw high current.

Yes, I think adding two more GDT windings is all it would take for this design.  (For the BJT version, another CT is needed, with primaries in series.  That is because BJTs are current-driven.  Two more windings on the same core would make equal Vbe, which won't necessarily insure sufficiently-equal base currents.  I've built full-bridge BJT bridges like this, but not even a half-bridge using IGBTs, yet.)