[Radio Wrangler]

I've had just that sort of fun with the old Watkins-Johnson AH312. The available gain ramps up dramatically at the LF end, and although the strip begins with a nice impedance stabilising pad, the match that the device's s11 forces on me at the intended operating frequency prevents the device from seeing the nice real impedance at LF. If I try to engineer an LF connection, it involves relatively large inductance and all hell breaks loose. Getting a proper match on the wanted frequency is easy, doing it while also presenting the part with a damping impedance across all the frequency range it can oscillate over is a bit awkward.

Qorvo have just made the AH312 NRND, so I get to pick something less challenged on the stability front.

The moral in the story, for any innocent bystanders, is that exotic, fast, parts can be big trouble at lower frequencies.

People used to talk about "VHF construction techniques" and UHF ditto. These continue into styles of PCB layout and circuit design. The sting in the tail is that the level of technique needed to use any device is NOT set by the frequency you intend to use it at, but instead has to be good over all the frequency range the device has enough gain to make oscillation possible.

This applies in the valve world as well. Attempts to use bottles usually seen in TV tuners for audio circuits can oscillate at VHF/UHF. Using RF power valves in audio amplifiers is similarly fraught. The audio world keeps trying this in response to what its own demand for NOS audio valves has done to their prices.

David