UCC28780: UCC28780

Hello Robin,

Thank you for your interest in the UCC28780 ACF controller.

At present, there are a couple of SIMPLIS models for the UCC28780 but no TI-TINA model. I believe that there are no plans to produce a TINA model.

An isolated ACF converter can certainly work from a 380Vdc rail, and can be optimized for efficiency at that voltage. One consideration is whether the flyback must also operate at low input voltage if/when the PFC stage is disabled. Another is whether you want to design one converter that handles all combinations of output voltage and current, or a separate converter for each combination.

In general, absolute highest efficiency can be achieved when the input and output of the converter are limited to fixed or narrow ranges. Wider inputs and multiple output options erode the achievable efficiency because the components cannot be optimized for all operating conditions. The efficiency can still be higher than non-ACF, just not the absolute highest.

An evaluation module (EVM) is useful to observe the various operating modes of the controller under different line and load conditions. The two models available (one with GaN Fets and one with Si Fets) are single-output designs of 20V at 45W capability. It is possible to modify them, if you so choose and have the capability, to explore some of the “what-ifs”.

But be aware that some modifications may be more complicated than at first expected. For instance, changing the 20V output to 5V sounds like a simple feedback resistor change. However, the internal auxiliary winding for the primary side bias voltage doesn’t have enough turns to sustain VDD_PRI at low output voltage. So either a new transformer design would also be necessary or an isolated external bias source must be coupled into the VDD net to maintain the bias. It is not always obvious what all of the consequential changes are needed to accommodate any particular “what-if” modification that you’d like to investigate. A smaller change, like dropping Vout to 18V, can help reveal the trends of what happens at other points in the circuit without requiring extensive additional modifications yet indicate what such modifications may need to be for larger spec variations.

Regards,
Ulrich

Ulirch:
you are right on for changing the xmfr. Indeed, since our xmfr is no ordinary typoe, that is #1 plan. Yes, if we change to any voltage, I will redo calculations of all parts implicated & do so on the eval board. We will buy a few so as not to rework on 1.
Appreciate your explanations.
Let us then begin .
this closes the p[ost.
thnx