LO phase noise!?

This simple oscillator was connected directly to the input of a Perseus receiver. The oscillator was built "quick-and-dirty ugly 3-D style" without any shielding so it picks up some 50Hz hum which can be seen as +/-50Hz AM sidebands at about -90dBc. However, the interesting thing is that the "noise trumpet" is -120dBc at +/-100Hz in ~4Hz BW (10dB/div vertical, 200Hz/div horizontal). I have no idea how much the oscillator is contributing to the total noise (N.B. the oscillator output signal is BP filtered through the XTAL) but, anyway, the result is quite impressive! Direct sampling with a clean ADC clock and low ADC aperture jitter...