UCC28780: ucc28780

Hello Robin,

I have scrutinized your .xlsx file and found the following items that contribute to the error results, some more so than others. This calculator was originally designed based on typical AC inputs and the DC option was added later. I do concede that the DC inputs are not all clearly defined as to how they relate “equivalent “ AC inputs, and can be confusing to fill in.  

On the “Input Here” sheet:
a) Cell C17 should be an equation result of C18 input, but is improperly allowed to be overwritten. This will have to be corrected in a revision. C17 = 0.83*C18, and is not adjustable.  It should be a result, not a user input.
b) C19 user input should be the highest expected input voltage, such as the peak of PFC output ripple.
c) C20 user input should not be higher than C19.
d) C21 user input should be lower or equal to C22.
e) C23 user input (minimum line frequency) should have no meaning in a DC case.  Since this term is used in sizing a bulk capacitor, I suggest to fill in “1000000” Hz to minimize the recommended bulk cap value for a DC input.
f) C25 user input asks for expected overall efficiency in %. Your entry of 0.93 was translated into 0.93% and the follow-on calculations using this value were greatly affected.
g) C33 user input asks for the % of output power threshold at which to begin the 160ms OPP timer. Your input of “30” translates to 30% of the full rated power of 15.75W.  I believe you intend it to be 130%.
h) C69 result recommends a negative capacitance value for Cbulk due to an internal assumption of -2V from diode bridge voltage drops. This is not appropriate for a DC case and the equation needs to be revised.

On the “Calculation” sheet:
i) D38 and D39 results spread is wider than usual because the SR-Fet Vds voltage rating entered in C83 is higher than necessary for a 5-V output.
j) D40 user input chosen at high end of recommended Nps range results in very high magnetizing inductance at D49 and high reflected voltage at D47.
k) D86 user input was entered as “6..8” ohms and the double decimal resulted in invalid follow-on calculations. This affects at least 11 follow-on calculations.
l) D92 result recommends a negative resistance for Ropp. Essentially, this means the projected offset voltage at CS is higher than the expected Vcst level and it needs a negative offset to work. Since this is not possible, some other value(s) is(are) also not right for DC. I tried following up to find the source(s) of error and became mired in a number of hidden cryptic Excel equations that are difficult to decipher. For now, I suggest to use the AC equivalent condition to the DC level to find an appropriate value for Ropp.

To provide a timely answer, I’m listing these items now and will continue investigation into the remaining errors. It is evident that the tool needs another revision, but I can’t say when that will be finished and available.

Regards,
Ulrich