I was thinking about radar speed guns the other day (No, I didn't get a ticket). My understanding of these is that they produce an IF frequency between the transmitted carrier and the Doppler shifted reflection in the receiver diode. This is AC coupled(?) to an amplifier with a passband in the audio region. Most traffic radar operates with Doppler shifts in the audio band and, as a result, the signal may be monitored through a speaker in addition to using a frequency counter.
What does the passband of the IF section look like in terms of its lower and upper limits, dB/octave slopes, etc.? The initial AC coupling and positive gain vs freq. would appear to create a bias toward selecting the faster target over the larger one. However, at some point, there must be a high frequency roll-off where the opposite is true.
Also, how would one characterize the rejection of amplitude modulation in the receiver/mixer (if any)? The common technique for 'calibrating' speed radar seems to be to use a tuning fork whose frequency corresponds to some known speed Doppler shift. Unless radar guns are susceptible to AM interference, it would seem that a tuning fork would be useless. Unless one threw it, the average 'speed' of its tines would be zero whereas the peak would depend on the amplitude of its vibration. Using a tuning fork might verify the accuracy of the device's counter, but it would also indicate that could register something other than a Doppler shifted signal.