Dear Rob, LF Group,

Some years ago my work involved a project to develop a set of tuned loop
measuring antennas for the 100kHz - 30MHz range. To save time, we hoped, we
used "modamp" ICs similar to yours. Although they eventually gave quite good
gain and SNR at LF, the big problem was that they would oscillate
relentlessly at about 1GHz if connected directly into the loop antenna
circuit. With the use of LF tuning and matching components and a layout
suitable for them, there was obviously plenty of opportunity for instability
when using amplifiers with gain into the microwave range. The difficulty was
not immediately obvious, because the receiver connected to the loop output
could not detect the output at 1GHz, and the amplifiers still gave some gain
at LF while they were oscillating - but the linearity and noise figure was
awful, as you might expect. We solved the problem in the end by adding a
pi-section low-pass filter directly at the amplifier inputs, with a cut-off
frequency of about 100MHz and surface-mount components to get good
attenuation at higher frequencies. This looked a bit incongruous next to all
the big transformers and tuning capacitors.

As a wideband LF/MF preamp, I have made a lot of use of the circuit in the
article on G3YXM's web site

http://www.wireless.org.uk/bploop.pdf

The low-frequency transistors used give good performance in the LF/MF range,
are less likely to be unstable at UHF, and are easy to get.

Cheers, Jim Moritz
73 de M0BMU