[evlatests] 3-bit update

[Vivek Dhawan]

Report on data of April 10th. The intended bandpass stability test is
pointless with these data as the weather was poor for K-band. So these
are general comments.

1. 9 antennas had a full set (i.e. 8) 3-bit samplers,
    ea1, 5, 6, 13, 16, 17, 19, 21, 23, 25.

    6 more antennas were half-populated with A1C1 A2C2:
    ea7, 10, 14, 18, 26.

    ea7:A2C2 and ea16:B2D2 gave no fringes - the samplers were off or
    did not get turned on. Some antennas just do this. I'll try and keep
    track if they are the same ones (and if their delays need updates).

2. All working channels had good delays, from Michael's effort (below).
    8 of the 9 antennas which had previously had their delays set, were
    still good.

3. DC-removal was enabled and works - no significant DC in any of the
    post-FIR data. Except ea19A is showing a curious spike at the lowest
    (highest) channel of every odd(even) subband. This is most likely a
    station-board problem, this one was known to be flaky; also has bad
    switched power. (might be fixed by now, it was rebooted.)

4. The following channels have flaws, as noted here. If they are
    repeatably bad, based on more tests to come, we can discuss if they
    need fetching or some kind of big reset.

Higher noise by > 50%: ea07A1, 13A2, 16B1B2 (older data), 18A2, 25A2.

Bad bandpass ripple: ea10A1C1 (I think this is a DTS issue, or perhaps
                      another st.board problem, not samplers. We'll see.




-------- Original Message --------
Subject: 3-bit delays
Date: Tue, 10 Apr 2012 13:44:32 -0600 (MDT)
From: Michael Rupen <mrupen at aoc.nrao.edu>
To: Ken Sowinski <ksowinsk at aoc.nrao.edu>
CC: Michael Rupen <mrupen at aoc.nrao.edu>, Vivek Dhawan <vdhawan at aoc.nrao.edu>

Hi Ken/Vivek --

   I had to reset one or more 3-bit delays for:

      ea07  ea10  ea14  ea16  ea17  ea18  ea26

The delays were all OK for:

      ea01  ea05  ea06  ea13  ea19  ea21  ea23  ea25

ea07-A2C2 and ea16-B2D2 gave no fringes.  The input power for ea07-AnCn was quite
low, but ea16-BnDn looked fine.  Both antennas gave the same results in multiple
observations, so this is not a simple one-off problem with the gain settings.

VIVEK: the file to look at is

   k3eq_allband_p3C48_003.56027.78745050926

This bops between 3C84, 3C84-offset, and 3C48.  Unfortunately 3C48 was just about at
zenith during this run (grrr...).  Sorry...

       -- Michael