[evlatests] Strange X-band birdie

[Dan Mertely]

The real question then is whether the 4092 clock
harmonic is conductively leaking into the signal
path, or radiating out from the digitizer & finding
it's way through the feeds into the RF signal path.

We've taken great pains to hold-in these harmonics. Mike
& Kerry have applied careful PCB design rules to try to
avoid radiation and cross talk.  Then, the digitizer
boards are installed in shielded modules.  The D30x modules
are the tightest we have, and are installed in an (open) bin
lined with absorbing foam.  Finally, the FE/Utility rack
was selected for it's shielding properties, and also
includes absorbing foam on the walls.  The total system
shielding should be in the order of 100 dB!  On the outside
chance that the strongest harmonics are escaping above the
EVLA detrimental levels, it shouldn't be strong enough to
bother any but the few frequency bins at 4096*N.  We hate
to lose any bins, especially during spectral line sleuthing,
but that may be the cost of having such strong signals in
the processing chain so close to our observing ranges, and
so close to the inputs to our ultra-sensitive front ends.

-Mert



Rick Perley wrote:
>     And Ken is right!  It is indeed at 8192 MHz (apparently I can't do 
> simple arithmetic). 
> 
>     I reviewed all 351 cross-power spectra, and confirmed that the 
> birdie is seen only on baselines amongst the antennas listed below, and 
> that it is strongest *by far* on baselines with antenna 16.  Start 
> investigating with that one!
> 
> 
> 
> Ken Sowinski wrote:
>> On Wed, 15 Sep 2010, Rick Perley wrote:
>>
>>>    The X-band HercA data have shown an odd birdie, not previously known
>>> to me.
>>>
>>>    All four IFs show a strong, unresolved (2MHz wide channel) birdie at
>>> 8188 MHz.  But it's visible only on antennas 13, 16, 19, and 26.  Of
>>> these, it is, by far, strongest on antenna 16.
>> 8192 is a multicple of 4096 and all its sub-harmonics.
>> Any chance you are off by two channels?
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests