[evlatests] [widar-wg] WIDAR/EVLA Birdies

[Brent Carlson]

If the birdies appear after fshift is introduced, they will be strongly
attenuated in the cross-corr output, but will of course show up in the
autocorrs.  In case anyone forgot this...
Brent

> There is a known birdie at 8192MHz  produced in the T304
> downconverter (probably the 4096 fixed LO x 2).  This will likely
> show up in all bands since the T304 is used for all bands. The
> location will probably move around depending on L302 settings.  I
> don't remember if this will also show up in the 2-4 GHz (3-bit) path
> or is limited to the 1-2GHz (8-bit) path where you are seeing it
> now.  We knocked it down as far as we could without a major redesign
> of the T304 circuit boards and at that time everyone seemed to be in
> agreement that it was too low to cause any major problems.  I think
> it was close to or even within the original spec of the module but
> it's been at least 3 years since we looked at it.  We'll take a quick
> look at it again but at this point in the game, we're probably beyond
> investing any major time or money into it.
>
> Jim Jackson
>
>
>
> At 11:49 AM 9/18/2008, Michael Rupen wrote:
>>                    Summary of Widar (and EVLA) Spurs
>>                            September 18, 2008
>>
>>We have looked at enough data now to have noticed persistent birdies
>>generated in the antenna and, perhaps, in station board 2 which is
>>currently connected to ea23.  So far there is no evidence of birdies
>>generated in station board 1 (ea18).
>>
>>The spurs seem for ea23 appear in the center channel of subbands
>>0, 2, and 6.  Subband 4 shows no obvious birdie in the central channel,
>>but there is a strong birdie around channel 290.  These birdies have
>>all been seen in both C and X bands, and do not move when the antenna LO
>>is shifted.  They must be generated in the sampler/DTS or in the station
>>board itself.
>>
>>In X band there is large birdie seen at 8192 MHz in both antennas, and
>>also seen in Walter's autocorrelation plots.
>>
>>In C band there is a large birdie seen at 4648 MHz in both antennas.
>>This moves in the wrong direction for a sky signal if the LO is retuned
>> so it
>>must be inserted between two mixers when the signal is lower sideband.
>> This
>>is also seen in Walter's autocorrelation plots.
>>
>>This table summarizes what we have seen.  The third column lists
>>the sky frquency for the X band test and the fourth column indicates
>>whether a birdie was seen at the same *channel* at C band.  The final
>>column indicates whether the line moved when the antenna LO was offset
>>at C band.
>>
>>SB   Ch    Xband?  F_sky  Cband?    moves on retuning LO at C band?
>>0   1024     yes    8936    yes       no
>>
>>1           nada            nada
>>
>>2    320     no     ----    yes       no
>>      1024    yes     8680    yes      yes
>>
>>3           nada            nada
>>
>>4    290    yes     8470    yes       no
>>      1024    yes     8424    yes       no
>>
>>5           nada            nada
>>
>>6    640    yes     8192     no       --
>>      1024    yes     8168    yes       no
>>
>>7           nada            nada
>>
>>[Note: each subband is 128 MHz wide, and has 2048 channels.]
>>
>>================================================================================
>>_______________________________________________
>>evlatests mailing list
>>evlatests at listmgr.cv.nrao.edu
>>http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>
>
> ________________________________________________________________________________
> There is a web page, used as a document repository, associated with this
> list.
> The URL for the web page is:
> http://www.aoc.nrao.edu/doc/vla/EVLA/Computing/Widarwg
>