Some clarifications on IIP3 computation

Hello,


I have some problems in understanding how IIP3 computation works.
I follow the following book, pages 476 to 478: https://books.google.be/books?id=281...page&q&f=false

My goal is to find k3 knowing k1 and IIP3, so eq. 4.7.4 seems just right.

However, I do not understand the results I get.
Looking at the first line of the table at the bottom of page 477, an IIP3 of 1.37 dBm is found (OIP3 of 19.65 dBm).
Always looking at this first line, k1 is equal to 18.28 dB. k3 = Pout3(dBm) - Pin(dBm) = -44.48 dB can also be found.
Obviously, these gains are hugely different.

However, putting the value of IIP3 = 1.37 dBm into eq. 4.7.4 after putting things back in linear scale, I get a ratio k1/k3 close to 1, which I cannot understand.

What do I miss?
--> Is it wrong to mix dBm (power) and dB, even if I cannot see how to avoid this mix?
--> In any way, knowing k1 and knowing the IIP3 value in dBm, how do I find k3?

Thanks a lot for your help,
Arkys