UCC28070A: Advanced Current Synthesizer Current Sensing

Hi, to answer your questions regarding the reasoning for the SYNTH feature, you must first understand that this controller operates in Continuous Conduction Mode (CCM).

Because this is a CCM controller, the current doesn’t go to zero between cycles in normal operation and therefore, the transformers would require a periodic reset signal to prevent them from eventually saturating. If there were only one CT in the inductor path, it would require a reset interval to be injected which would result in a current signal that doesn’t accurately represent the boost current.  A common approach is to use one CT in the diode leg and one in the FET leg and sum the two outputs thereby being able to reset one while the other is conducting.  The SYNTH function allows the use of the single current transformer in the FET path only, eliminating the one in the diode path.  This is even more important when phase interleaving is used which would require four CTs rather than just two with SYNTH.

To your point of using a Hall-Effect sensor in the inductor path, it can probably be done as it wouldn't suffer from the saturation issue of a transformer.  However, I would expect there to be a cost hit as the Hall-effect sensors are relatively expensive, plus you would likely have to add scaling to the output to fit into the CSx voltage range.

I think the more common option is to use sense/shunt resistors in the FET source paths and disable the SYNTH function by connecting RSYNTH to VREF.  In this mode, the CSx signals will bypass the SYNTH function and feed directly into the CSx comparators.  Without CTs to boost the signal level, however, losses in the sense resistors will be high.