[evlatests] Mystery Moon Test

[Ken Sowinski]

On Thu, 17 Dec 2015, Rick Perley wrote:

>    Rob ran the 'Moon compression test' last month at X-band.  The results 
> are bizarre:
>
>    1) We expect a power increment of about a factor of 5 to 8. This is what 
> was observed in the 'PSum' digital power.  So far, so good.
>
>    2) The 'set and remember' regimen was applied when the source was on the 
> moon.  The digital power level should be about 15, when on the moon.  This 
> was also the case.  Still good....
>
>    3) We then expect a 'PDif' compression (PDif on moon)/(PDif off moon) of 
> about 0.9.  In fact, this ratio was about between 0.1 and 0.4 (!!!!!)  That 
> is, for every antenna and every IF channel, the PDif when on the moon was 
> only 10 to 40% of the value when on cold sky.
>
>    This is a level of compressin that I have never seen before.
>
>    It may be worth noting that the PDif values, which are normally hundreds 
> of millicounts, where only 10 -- 50 millicounts.
>
>    Something is wrong.  Any ideas?
>
>    Rob reports that the TPD values (recorded in parallel) were also weird.

This was X band, so the T301 attenuators are irrelevant.

I've looked at the script and compared it with the one I had prepared
for previous tests.  There is a difference in the protocol.  In
the past we did set and remember on "cold" sky" and used that for
the moon; here we did just the opposite, set and remember on the moon,
and use that for the "cold" sky.  That under drives the sampler when
looking at cold sky so much that it is close to being used as a 1-bit 
device.

Might this have something to do with it?