[evlatests] Nov 8 gain - level tests

[Ken Sowinski]

The pad between the deformatter and T4 for 14A was changed
from 10 dB to 6 dB and the tests of last week perfomred
again.  The results are more or less as expected but there
are some disturbing details.

1.  Low signal case.  Set the Gain parameter to 85 and use the 
downconverter attenuator to control input power to the sampler.
2.  High signal case.  Set the downconverter attenuator so that 
RMS of the sampler  values is between 9 and 10 and then vary the 
Gain parameter of the deformatter.
In both cases SD voltage and correlation coefficient are recorded
as a function of the free parameter.

In the low signal case, the correlation coefficient response is
unchanged from last week; the SD response is a little more extended 
on the low power side.  The peak in the SD repsonse still appears
after requanizer overflow occurs.

In the high signal case, the correlation coefficient response is
unchanged from last week.  The SD response is still in the form
of a plateau but now extends perhaps 3 dB farther toward the low
power side.  The plateau begins about the time the ALC levels and
decays when DAC overflow occurs.  It is disturbing that the SD voltage
is not stable but varies randomly at about the ten percent level
with small changes in Gain setting.  This was apparant last week also,
but with a wider plateau there is more opportunity to see this.  

The conclusion I draw is that we must think about what this means and 
what we want from the system.