Honored Contributor
Honored ContributorSection 3-1 defines N as the "magnitude precision," and with that definition it is more or less correct if you consider that magnitude is separate from the sign. An M-bit 2's complement sample has a maximum range of -2^N - +(2^N)-1, where N is M-1. However the UG is not entirely consistent because later on the value M refers to all bits of the output, including sign, and they still call it "magnitude precision."
The take-away is that the core will give you a full-amplitude signal, no bits are wasted, so you have to use a bit of thought and not take the UG too literally. My $0.02 anyway. -Fred
