[evlatests] Sensitivity of 3-bit vs. 8-bit

[Michael Rupen]

> As I've said before, I consider the phase noise in a single channel
> as the gold standard.  This method probably doesn't, but might, couple
> in spectral anomalies through the calibration process.  We are going
> for 1% after all.
>
> Anybody got any ideas about the 'noisy' 8 bit digitizers?

We have been carefully setting the subband filter gains for the 3-bit
data, but *not* for the 8-bit data.  According to the one test we've done
so far this could be a ~5% effect.  We plan to observe again, with
filter gains set for *both* 3- and 8-bit, hopefully later today.

Note too that the standard 8-bit input power level (-30 dBm) seems
to be heavily over-driving ea22-A -- not enough to saturate the sampler
but this could mean the default subband filter gains are less ideal
for that one.

           -- Michael