[evlatests] 3-bit status

[Vivek Dhawan]

Michael got some preliminary data, enough to show that the amplitude
dropouts are now gone, as are the delay changes with time.

ea07 was autostowed, oversensitive to wind. ea12 a2c2 not fringing.

ea14A1C1 show a 20MHz standing wave. The pattern is aligned on both
A1 and C1, so I'm guessing that a cable on the T304 or LO might be
the cause.

Raw correlations from A1C1 show the amplitude of A1 > C1 by 10-60%
for many baselines. A2C2 was not so discrepant, except ea10 was the
most imbalanced in both A1C1 and A2C2. BD was 8-bit, balanced +_5%.
I mention this because the levels, equalizers, requantizers were all
adjusted and should have removed gross imbalances.

Post-calibration things look better, within about 1-2%. More serious
data with multiple subbands is expected tomorrow.


>>
>> * New samplers in ea10, ea12; all old samplers removed from array
>>
>>    We currently have all the first article sampler boards out on
>>    the array, with 4 samplers (2 modules) each in:
>>
>>        ea06  ea07  ea10  ea12  ea14  ea26
>>
>>    ea06, ea07, ea10, ea14, and ea26 have new T304s
>>    ea12 is still using old T304s -- we expect no fringes for A2C2, and
>>      possibly low amp on A1C1