UCC28780: Operation without auxiliary winding for VDD

Hello Yair and Nir,

Thank you for your interest in the UCC28780 ACF controller.
I believe that it is conceivably possible, but unfortunately not very practicable to operate this controller without an auxiliary winding. I can't say that it is impossible, because one can always implement at set of complicated circuits that emulate the functions of the Aux winding. But it may not be worth the trouble.

If the Aux winding is eliminated, you would need to provide:
1. a VDD source that could rise to at least 18.2V to guarantee start-up. It could stay there or later fall to 12~15V to reduce bias power.
2. the SWS and HVG circuits involving the switched-node sampling FET (Qs) are still necessary. The controller uses the SWS signal to maintain ZVS switching and control the primary negative peak current. It would result in significant switching and conduction losses if eliminated, and defeat the purpose of this ACF controller.
3. The VS pin is used to sense the bulk voltage to determine brown-in and brown-out conditions, to determine the line compensation current for the OPP function, to detect the zero-voltage crossing moment of the output winding waveform, and to detect an output overvoltage condition.
These are all dynamic signals, sampled at different moments within a switching cycle. Fixed-threshold comparator circuits are completely inadequate to emulate these. Essentially, the primary switched-node voltage needs to be faithfully scaled and level-shifted down to the VS input signal level, at very high frequency, in order to replicate the missing AUX waveform. A simple Aux winding does all this for you without requiring a military budget to finance an emulator.

Regards,
Ulrich

Hello Yair,

Sorry for my delayed response. The well-known opto-coupled CV/CC feedback arrangement will generally work. I'm certain that operating from 48V all the way down to 0V will present some special challenges to be overcome. The IC was not specifically designed for this purpose, so work-around circuits will have to be generated to address some concerns.
These include (but not limited to):
- having sufficient amplitude at the VS input to reliably detect zero crossings (at the 30mV "zero" threshold) in the presence of noise from switching at the maximum current and the minimum aux voltage.
- dealing with extended demagnetization times at very low Vout, and operating within the maximum 40-us burst period (25kHz minimum frequency) in a stable manner. This may be overcome by setting the burst threshold to minimum when Vout declines.
- dealing with the resonant current ringing effects when the demag time is stretched out widely at low Vout.

The UCC28780 ACF controller is configured to facilitate high efficiency from full load to light load, but it is unclear what the efficiency performance curve looks like when operating in constant current at reduced output voltage, where output power decreases yet secondary conduction losses stay high.
It might be best to simulate your application using the Simplis model available from the Tools & Software tab of the UCC28780 product folder on TI's website: www.ti.com/.../toolssoftware
This will give you a good idea of what can be expected in terms of current and voltage wave shapes, and operating modes.

Regards,
Ulrich