[evlatests] U-Band Failures to Tune

[Ken Sowinski]

On Mon, 17 Jan 2011, Rick Perley wrote:

>    There were 7 fringing antennas:  7, 10, 12, 14, 20, 22, 24 and 26.
>    The run began on December 26 at 16:42 IAT.
>
>    The following 'failures to tune' (defined as no fringes detected,
> when there should have been) were noted.  In all cases, the entire scan
> is affected.
>
>    ea10, AC+BD at 22:27
>    ea24, AC+BD at 1/01:11  (i.e., on the 27th at 01:11 IAT)
>    ea26, AC only, at 1:03:01
>    ea07, AC only, at 1/03:22
>    ea10, AC+BD at 1/06:55
>    ea22, AC only, at 0/08:27
>    ea12, AC only, at 1/08:47
>    ea07, AC+BD at 1/11:25
>    ea10, AC+BD at 1/13:26
>    ea12, AC+BD at 1/16:26
>    ea12, AC+BD at 1/16:36 (this is the next Ku-band scan following the
> previously listed failure).
>    ea10, AC+BD at 1/18:59
>    ea20, AC only, at 1/19:14
>    ea07, AC+BD, at 1/19:31
>    ea22, AC+BD, at 1/21:19
>
>    Thus, the 'failure rate' can be computed as 15/(105*7)=2.0%.  This
> is of course much too high.

Monitor data was recorded for ea01 and ea10 during this time.
I find that for ea10 every case corresponds to an instance of
the M301 getting a command to switch to Ku band, but the monitor
point reports '0' instead of '4'.  There were no other instances
of the bandswitch monitor reporting '0' when Lu band was commanded.

The case for ea07 is more interesting.  There are many zero monitor
reports at all bands.  In many cases '4' appears, but is replaced
later by '0'.  Sometimes '4' returns before the next bandswitch
command, sometimes not.  It is true that all the instances Rick
reports report consistently a bandwswitch monitor of '0' rather
than the expected '4'.

I suppose that the M301 looks at all the switches and reports '0'
if any switch does not report itself to be correctly set.  In ea10
this correctly and exactly predicts the "failure-to tunes" at Ku
band.  In ea07, many other examples are erroneously reported by the
M301.