[evlatests] Effectiveness of New Switches

[Rick Perley]

    We're a bit more than half way through replacement of the VLA's 
switches.  Time for a check. 

    To do this, I used a test, run last week, where the four high 
frequency bands were observed in sequence, about 3 minutes each.  Four 
cycles around were completed -- a bit less than one hour. 

    Many curious phenomena are seen -- nearly all are very likely 
connected with real, short-term (tens of seconds) gain changes in the 
receivers or IFs.  Nearly all of these short term variations are small 
-- less than 2%.  These are very likely unconnected with switch issues. 

    To isolate issues which are more likely  due to switch failures, I 
tabulated all changes in PDif greater than 5%.  These were divided into 
two types -- (a) those which occur between scans -- very likely a switch 
issue, and (b) those occuring within a scan -- could be a bad switch or 
a receiver gain change.  A simple table -- shown below was drawn up, 
color-coded by severity.  There are four classes: 

    A)  Changes greater than 40%
    B)  Changes between 20 and 40%
    C)  Changes between 10 and 20%
    D)  Changes between 5 and 10%. 

    The table is broken into two halves -- those with new switches, and 
those with old.   The numerals in the matrix give the number of 
occurances.  The maximum for any one antenna is 16 (four IFs * four 
bands). 

I)  Antenna with new switches

Antenna         Class A          Class B            Class C           
Class D
---------------------------------------------------------------------------------------
     3                     0                   0                      
0                       1
     5                     0                   0                      
2                       5
     7                     0                   0                      
2                       2
     9                     0                   0                      
0                       0           Perfect!
    10                    0                   1                      
2                       2        IF 'D' is bad
    12                    0                   0                      
1                       3
    13                    0                   0                      
0                       3
    14                    0                   0                      
3                       1
    16                    0                   0                      
0                       0
    19                    0                   0                      
5                       0
    21                    0                   0                      
0                       0
    22                    0                   0                      
0                       0
    24                    0                   0                      
0                       0
    26                    0                   0                      
1                       2
    27                    0                   0                      
0                       0
--------------------------------------------------------------------------------------

    Note that 6 of the 15 antennas have no noted variations. 

II)  Antenna with Old Switches
------------------------------------------------------------------------------------------
    1                     0                     1                     
3                       0
    2                     0                     1                     
1                       2
    4                     0                     0                     
0                       0  (No Ka-band rcvr)
    6                     0                     0                     
3                       0
   11                    0                     0                     
0                       2
   15                    0                     0                     
0                       2 (No Ku-band rcvr)
   17                    0                     1                     
1                       0
   18                    0                     0                     
0                       0
   20                    3                     3                     
1                       0  The Worst by far!
   23                    0                     0                     
0                       0
   25                    0                     3                     
1                       1
   28                    0                     0                     
2                       0
------------------------------------------------------------------------------------------------------
Two antennas out of 12 are 'perfect'. 

    There is a clear difference in the number of severe (> 20%) changes 
in gain,  but very little difference in the rate of 'small' gain 
changes.  A possible conclusion is that the new switches are much better 
in preventing large gain changes. 

    It seems likely that the smaller gain changes are not related to 
switch issues, but rather to short-term gain changes in the signal chain.