The windows application auto calculates the negative bleed current, even if FRAC1 and FRAC2 are not 0, but from the ADF4355-2 Rev C page 25, :
"Use negative bleed only when operating in fractional-N mode,
that is, FRAC1 = FRAC2 = 0. Do not use negative bleed for f PFD
greater than 100 MHz."
Which is correct?
(Is it possible to get the source code for the Windows Application)