[evlatests] 3-bit sampler status

[Ken Sowinski]

I have started to verify that the 3-bit samplers are working well
enough for astronomical tests.  Here is a summary of the state
this morning based on looking pnly at power levels, state counts and
autocorrelations.  It was too windy to do much with correlated data.

There are 11 antennas with 3-bit samplers:
All four      Only D301, D303
   ea05                ea06
   ea17                ea07
   ea19                ea10
   ea21                ea14
   ea23                ea18
                       ea26

1.  A general problem is that the sample multiplier used to choose
the three bits to be used defaults to 64 instead of 32.  A value
of 64 chooses the three LSB bits, rather than the desired bits
1, 2 and 3 resulting in an exaggerated DC offset.  We can easily
set the multipier, but it reverts to 64 whenever a sampler reset
is done.

2.  Many of the 3-bit samplers had to be turned on.  Should we
     routinely be sure that they are on after installation and
     whenever there is an antenna reset just as we do for the
     8-bit samplers?

3.  Problems noted realted to the AC (D301, D303) samplers.
-  ea05C (ea-5-d303) reauired a formatter reset before the three bit
    samplers would work correctly.
-  ea10-C1 has especially funny state counts.  There might be
    CW seen by the sampler.
-  ea19 A2/C2 have larger than usual DC offsets.

4.  Problems noted related to the BD (D302, D304) samplers.
-  ea17-D2 has too little power fed to the sampler.
    D1 is OK, so likely culprits are T304, wiring, or D304.
-  ea21-B2/D2 have too little power; B1/D1 are OK.
-  ea23-D2 attenuator monitor point is stuck at 31.  There
    might be other modules with this problem in a channel 2
    output; this is the only one that I verified.