[evlatests] Wobbles at X-band

[Rick Perley]

    The 'Cygnus A' experiment run last week has clearly revealed the 
presence of 'wobbles' at X-band. 

    These are seen only seven baselines (out of 325 in the array at the 
time).  The characteristics are the same as has been reported earlier 
with the prototype correlator.  As noted there:

    a) Periods are typically a few seconds. 
    b) On the baselines which show both amplitude and phase wobble, the 
phase period is half that in amplitude. 
    c) The effect is typically evident for an hour or two.  During this 
interval, the period of the phenomenon changes, reaching a maximum in 
the middle.  There is good evidence that the wobble phenomenon is 
*always present*, but is only visible when its period is long enough 
that it is not averaged out by the integration period. 
    d) The effect is on both IF pairs, in both polarizations, with equal 
amplitude. 
    e) The frequency of the wobble is proportional to the observing 
frequency. 
    f) The phenomenon does not close -- it is not possible to factor out 
the effect by antenna.
    g) There is no relationship of the presence of the wobble, or its 
frequency, to baseline or source geometry. 

    The baselines on which this is seen are given below, with some notes:

    4 x 20:  (W1 x N5)  amplitude and phase.  Seen in the middle of the run.
    4 x 28:  (W1 x N8)  Phase only.  Seen in the middle of the run. 
    2 x 10:  (E2 x N3)  Amplitude and phase.  Seen in the beginning of 
the run.
    2 x 14:  (E2 x E5)  Phase only.  Seen in the beginning of the run.
    5 x 25:  (W8 x N2)  Amplitude and Phase, seen at the end of the run.
    8 x 24:  (N1 x W5)  Phase only, seen at the end of the run.
    5 x 15:  (W8 x W6)  Phase only, seen at the end of the run. 

    I have not yet checked the C-band or L-band data (taken at the same 
time) for the presence of this phenomenon.