[evlatests] 8-bit PDif/Psum compression meeting

[Rick Perley]

    The 'PDif Compression' problem, which is notable at all bands, but 
is particularly troublesome at P, L, and S bands, is still wanting an 
explanation.  By all accounts, 'it can't happen'.  But it does. 

    I've reserved Rm 280 at 9AM tomorrow (Friday), with the purpose of 
reviewing the characteristics of this effect, and with the hope of 
generating some useful tests for its origin.   If this time presents a 
real problem, let me know ASAP, with suggestions for an alternate time 
(not date).

    Note that we're focusing on the problem as seen with the 8-bit 
system.  The '3-bit compression' problem, while much worse by some 
measures, is understood to arise with the expected overflow of the 3-bit 
samplers, and is hence correctable (at least in theory).  It may well be 
that the additional problem plaguing the 8-bit system may exist within 
the 3-bit system, but this is not yet known. 

    Rick