[daip] 512 Mbps IF setup and AIPS

[Eric Greisen]

Yuri Y. Kovalev wrote:
> Dear colleagues,
> 
> We observe in BL149 these days with the following frequency setup:
> 
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> ! U band
> ! Central frequency for SEFD 15369 MHz
> !
> setinit = 'BL149-512-2cmv.set' /
> ! NON STANDARD setup created by Yuri Kovalev on 08/27/2008
> !
> ! based on Standard setup file: v2cm-512-16-2.set                       
> !     (Produced by MAKESETUP)
> !      For the the 2 cm band.
> !        Radio Astronomy allocation:  15350-15400
> !        Radio Astronomy footnote(?): 14470-14500
> !
> !        There are 4 filters in the VLBA IF converter.  They are
> !        11820-12980, 12820-13980, 13820-14980, and 14770-15630 GHz.
> !        The filter is chosen based on the frequency of the first channel
> !        assuming an upper sideband first mix.  The splits are at 12.9,
> !        13.9 and 14.9 Ghz.
> !
> !        VLA band 14650 - 15325 MHz for 0.9 sensitivity
> !        Methanol has a rest frequency of 12178.595 MHz
> !
> !     Format      Tape length  Density  Tape speed  Time/pass
> ! VLBA/MkIII/MkIV    17600      High     160.00        22:00
> !    MkIII/MkIV       8800      Low      270.00         6:31
> !      512 Mbps - 1 pass/headpos - 2 tapes (VLBA) or 2 heads (Mark IV)
>   nchan    = 16
>   bits     = 2
>   bbfilter = 8.0
>   freqref  =  15360.99
>   freqoff  =  -27.50, -27.50, -27.50, -27.50, -11.50, -11.50, -11.50, -11.50,
>                4.50,  4.50,  4.50,  4.50,  20.50,  20.50,  20.50,  20.50
>   netside  =  L, L, U, U, L, L, U, U, L, L, U, U, L, L, U, U
>   pol      = dual
>   /
> endset  /
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> 
> I noted recently that AIPS (31DEC08) VLBASUM shows the following for 
> such a dataset:
> 
> Frequency Table summary
> FQID IF#      Freq(GHz)      BW(kHz)   Ch.Sep(kHz)  Sideband
>    1   1      15.32555250    8000.0005    500.0000      1
>        2      15.33349000    8000.0005    500.0000      1
>        3      15.34155250    8000.0005    500.0000      1
>        4      15.34949000    8000.0005    500.0000      1
>        5      15.35755250    8000.0005    500.0000      1
>        6      15.36549000    8000.0005    500.0000      1
>        7      15.37355250    8000.0005    500.0000      1
>        8      15.38149000    8000.0005    500.0000      1
> 
> I would expect that different IFs would have dufferent Sideband
> (1 and 2?) since we observe in both, U & L.
> Could you please comment on this?

So long as the frequencies are correct I suggest that you do not worry 
about it.  AIPS tries to simplify things and so sorts the channel order 
and frequencies to make it more uniform without the uncertainty of what 
"lower sideband" actually means.  In fact, it is unclear to me really 
what sideband means at this stage in the data handling.  What one 
actually has are N channels at some frequency and frequency increment.

Eric Greisen