[evlatests] T302 Linearity Test

[Rick Perley]

    The solar attenuators present in the T302 gives us another way of 
checking system linearity. 

    In the most recent 'solar test', we observed 3C84 (calibrator) in 
normal mode, then switched in the T302's 20 dB attenuator, while 
continuing to observe the source.  The T304's logic adjusted its 
internal attenuators to keep the power level to the samplers in the 
'happy zone'. 
    If everything works as it should, the switched power system should 
compensate nicely for this 100-fold change of power (seen only within 
the T302, as the T304 has corrected itself).   The apparent 
(post-correction) fringe power on 3C84 should not change after the 
changes have taken place. 
   
    Result:

    Everything works as planned!  The apparent change of gain between 
the 20db in and 20db out states is no more than 1%!  The system (from 
the T302 through to the correlator) appears beautifully linear, as it 
should be. 

    So where is the apparent PDif compression coming from?