Today I finally synced up with my friend Doug and collected from him quantity five Coilcraft 0805 10 uH chip inductors. Tiny things. Had little caffeine today, so by the time I’d returned home I was pretty steady.
Brought the LNA down from roof, opened lid, clamped assembly down to bench so it wouldn’t move, put a drop of 60/40 on one inductor pad, and with my TU-10b tweezer I picked up the part and set it in place, then tapped the one end with the soldering iron. It was harder than I thought; the part weighs nothing and has no surface friction with the tiny bead of molten solder, so it instantly moved on me.
After some re-approaching of the problem, I got the part securely attached. Checked continuity, everything looked good! Lid back on. Connected it to network analyzer, and everything did NOT look good. 20+ dB gain above about 200 MHz. The whole 10-200 MHz output level was badly attenuated, and there were strange artifacts in the low end of the spectrum. Removed the inductor and everything returned to normal. Posted a note to the designer over on gpio.com and am awaiting a response. 10 uH at 100 MHz is 6k ohms impedance, so it should be fine. The LNA was drawing its typical current (~ 160 mA).
Took another identical inductor and pressed it down on the pads with a plastic tuning stick, and did exactly the same thing as soon as it made contact. There’s something about that output circuit that doesn’t like the chip inductor.