I say you chaps...
I've added two LNA (MAX2634 - sorry!) and active filters with gain at baseband. It's a tough one to keep stable!
A SPICE model for the AD8348 would have been nice, to check the bias. I have built it, it does work, -100dBm sensitivity at 250MHz. A re-layout needed for more stability to give even better sensitivity.
I've got the model for the AD8692, (I use the dual for best layout) so at least I got the active filter sorted first. A big C on IOFS and QOFS to slug the higher gain feedback loop.
Here's a crude model for the AD835. Got a better one?
*AD835 by Richard Mayo
.subckt AD835 1 2 3 4 5 6 7 8
BV1 VZ 0 V=V(Z1,0)
R2 Z1 4 1000
G1 VZ 5 value={(V(1,2)*V(8,7)+V(4))/33.33}
Rout 5 VZ 33.33
R1 3 6 1000
Cout 5 3 636p
.ends AD835
TTFN
Richard.
