[evlatests] C-Band Switched DAta

[Ken Sowinski]

On Wed, 22 Jun 2011, Rick Perley wrote:

>    Same experiment as described earlier.  Cygnus A at 6cm doubles the
> system temperature.
>
>    About half the antennas show neither a zero offset (when switched
> power is turned off), nor PDif compression when observing Cygnus A.  The
> other half show varying degrees of the following:
>
>    A) Most show PDif compression, typically a few percent, but as large
> as 30% for antenna 24 in RCP.

I reported long ago that a 10 Hz signal is seen in total power
measured in the downconverter for 24 RCP at C band when the noise
tubes are turned off.

I looked for a source for this this morning.  Turning off the noise
tubes for each receiver controlled by an F317 one at a time made no
difference at all to the 10 Hz signal.  Turning off the all noise tubes
at the F320 cleaned up the C band signal.  I tested each one in turn
and no single band was the culprit; rather the more bands with noise
tubes on the greater the amplitude of the 10 Hz seen in the RCP signal.

At ea24 only the K and Q band receivers are controlled by the F320,
but turning on noise tubes for the four possible receivers produced
a larger amplitude signal than when turning on just the noise tubes
for the receivers that are in fact present.

the leakage mechanism proposed to explain this must explain why the
10 Hz seen in the C band RCP output is in antiphase with the noise tube
switching pattern.

An experiment not yet done is to verify that the C band signal is
clean with the F320 noise tubes off and the other F17 noise tubes on.