Tool/software:
My application using the LM3445 is a buck controller driving a 40V LED array.
My circuit is very similar to the one shown in T.I. application note and data sheet: SNVS570L –JANUARY 2009–REVISED MAY 2013.
The circuit works well with 120VAC applied. However, with 240VAC applied, the dissipation in the switching NFET overpowers the stand-alone heat sink which works fine for 120VAC.
The equations that I have found for total loss in the NFET indicate that with the shorter FET on-time at higher Vbuck voltage, the switching loss and conduction loss are actually LESS at 240VAC than at 120VAC.
Yet, I see the loss and dissipation RISE significantly. I do not have the space for an larger heatsink in my application to allow for this 240VAC surprise.
While I try to sort this all out with my oscilloscope, I could use some hands-on help from a T.I. application engineer, such as Irwin Nederbragt -- who helped me get started with the LM3445 several years ago.
I can supply the schematic and my analysis to such an engineer.
