Has anybody tried the default passive values for the SXA's?

[powerpants]

As has been discussed on this board before, some of the values of the components in the gain stage section are way different than those given on the application circuits on the SXA's datasheet, although the circuit topologies are pretty much the same.

Anybody know why this is, or tried using the "default values" given on the datasheet?

[inh]

in a word - impedance matching. If you look at the Sirenza datasheets, the Z pieces are stripline (0.54mm on 0.032" board, of a certain type). Unfortunately (at shorter wavelengths in particular) this stripline acts as an inductor (mainly) and so becomes part of the circuit. Ladyada's design does not conform to the appropriate Z lengths, which as you can see are expressed in wavelengths.

This is why RF design is described as magic. Unfortunately there are also issues such as solder fillet thickness, and constructive/destructive rf reflections going on at multiple places, which can only be measured accurately with a $40,000 network analyser.

So in the end - trial and error with a spectrum analyser on your unit is the best (cheapest) way to improve signal strength - I found slightly different values than Ladyada, but maybe that's because I used different soldering technique, or maybe I was focussing on slightly different fqs.

hope this helps.