You know, ISTR that during WWII, some military receiver specs called for two, well isolated, RF stages to reduce the possibility of local oscillator signal from reaching the antenna terminal. Posing the possibility of enemy radiolocating devices pinpointing location. A low level signal can be radiated for all to hear.
In the late 1940's a QST "Hints and Kinks" article described using an all band receiver as a signal generator. If you knew the IF frequency (single conversion was King in those days), it was simple to calculate (and use) the local oscillator's output for a variety of purposes. (Think modern transceivers)
Some of us, probably many of us, used that technique as a pretty stable and fairly accurate signal source working on homebrew (and surplus modification) projects.

Old Chief Lynn, W7LTQ

"lw1ecp" wrote in message ...
Michael is right. For the entire HF spectrum, you don't need RF stages in order to conceal mixer's noise: noise picked up by the antenna is stronger. Wes Hayward W7ZOI has been publishing articles on this since the 80s.
And you don't need active stages to get selectivity against image frequencies either, a double or triple tuned passband can be placed right begore the mixer.
On the other side, adding RF stages before the mixer decreases the receiver's dynamic range: makes it easier to be overloaded by strong signals some kHz or tens of kHz apart from the desired one.