[evlatests] 4band, interference and ea14

[Dan Mertely]

Allen Woo, Dennis Mobley & I checked around ea14 and ea10 with a
handheld scanner today.  Dennis helped switch off ACU components
one-by-one while Allen & I listened at various VHF frequencies
with the scanner.

While on ea10 (old style ACU) We verified that the old "C-rack"
puts out the strongest birdie at 50 & 55 MHz by far!  The MIB/ACU
interface chassis is the next noisiest.  Lastly, the "Data
Readout" chassis and the 48 VDC switching power supply are also
noisy at 50 MHz, but quite a bit quieter than the C-rack.

On ea14 we do see the 50 & 55 MHz, but it is very weak--perhaps
coming only from the 48 VDC switching power supply, or an adjacent
antenna(s).

So it appears that last year's guess that the 5 MHz comb was coming
from the old ACU components is mostly likely correct.

Earlier in the day Allen & I verified that the 50 & 55 MHz RFI is
clearly visible on ea14 at DE08 using the EVLA bandpass plotter tool.
ea14 has the new ACU components.  We also saw a mostly populated,
5 MHz VHF comb on ea12 & ea19 (the other MJP-equipped antennas, but
which have the old ACU components).

What we'd like to do is take a bandpass plot of 4-band on ea14, then
turn-off the *old* ACU stuff on the adjacent antennas (ea10 at DE09,
ea03 at DE07, & ea15 at DE06), then check with the bandpass plotter
tool again & see if the 50 & 55 MHz RFI goes away, or, at least gets
substantially weaker.  If so, we will know that ea14 with the new ACU
components is not the major source, and most of the VHF RFI junk is
coming from the adjacent antennas.

We can't do that today or tomorrow because we will be prepping for, and
installing the next MJP equipment on ea06.  However, if we can shutdown
the ACUs on the 4 end antennas on the east arm simultaneously on Thursday,
we could run the test then.  If not, we wait for the next week maintenance
day for this definitive test of the source of the VHF comb.

-Mert



On 6/23/2014 3:49 PM, Frazer Owen wrote:
>       The autocorrelation spectrum for antenna 14 shows no narrowband
> feature at 60 MHz. The correlated interference is at 60 MHz.
>
> ---Frazer
>
> On 06/23/2014 02:22 PM, Keith Morris wrote:
>> There are no intentional 60MHz signals in the API.
>>
>> Dan -- have you sniffed the API shack lately?
>>
>>
>>
>> On 6/23/2014 1:56 PM, Rick Perley wrote:
>>>         Mike Revnell just showed me VLITE spectra from ea14.  There's a
>>> super-strong resonant feature ('birdie') right around 60 MHz.  Looks
>>> like external RFI.
>>>         We note the API is nearby ...  any chance it's from there?
>>>
>>>
>>> On 06/23/2014 11:33 AM, Frazer Owen wrote:
>>>> Details of 60 MHz interference:
>>>>
>>>> 1) My test was looking at an 8minute average.
>>>>
>>>> 2) The test that shows the strongest cross-correlation at 60MHz is
>>>> looking at the NCP (North Celestial Pole).
>>>>
>>>> 3) The antennas are ea12 at N9 , ea14 at E8, ea19 at W1
>>>>
>>>> 4) Except for the NCP, the signals when averaged over 8 minutes
>>>> decrease in amplitude relative to 2 second averaging.
>>>>
>>>> 5) The strength of the correlated 60MHz signal changes with different
>>>> pointing positions (Cas A, 3C48, NCP). Especially the NCP strength is
>>>> very different from the other two.
>>>>
>>>> 6) ea14 at E8 has no internal 60 MHz interference based on the total
>>>> power and is near the end of the east arm. However, ea14-ea19 appears
>>>> to have stronger 60MHz interference than ea12-ea19. These signal have
>>>> been through the requantizer step but it seems clear that whatever
>>>> signal ea14 is seeing that correlates with ea19 is not coming from
>>>> ea14 itself.
>>>>
>>>> ---Frazer
>>>>
>>>>
>>>> On 06/23/2014 09:51 AM, Rick Perley wrote:
>>>>> In the (distant) past, a similar effect was commonly noted at
>>>>> 4-band.  The accepted explanation was either that:
>>>>>
>>>>>        1) An exceptionally bad antenna is radiating to the others. This
>>>>> would be most prominent in D configuration.
>>>>>        or
>>>>>        2) The supposedly incoherent LOs are sufficiently close in
>>>>> frequency that, for periods of seconds to tens of seconds, there is
>>>>> insufficient phase slip between them to destroy the cross-correlation.
>>>>>
>>>>>        I recall that #2 was in fact the theory I came to embrace, as it
>>>>> was usually noted that the baselines with strong combs were
>>>>> independent of antenna location.
>>>>>
>>>>>        You could test this by looking at the phase of the line
>>>>> cross-correlation.  If from a single source (i.e., theory #1), the
>>>>> phase should slip with the fringe rate.  But if from different LOs
>>>>> with slightly variable phases, the phase relation will be
>>>>> 'different'.  (I'm not prepared to define 'different', other than it
>>>>> won't be strictly the same as the fringe phase).
>>>>>
>>>>>        Rick
>>>>>
>>>>>
>>>>> Interestingly the 5MHz comb frequencies correlate on baselines to 14
>>>>> (as they do on 12-19). Since the 5MHz clock is supposed to be
>>>>> incoherent from antenna to antenna, there must be a general source of
>>>>> the comb frequencies, especially 60MHz, which is being broadcast
>>>>> around the site. Could this be one very bad antenna ? ---Frazer
>>>>> _______________________________________________ evlatests mailing
>>>>> list evlatests at listmgr.cv.nrao.edu
>>>>> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>>> _______________________________________________
>>> evlatests mailing list
>>> evlatests at listmgr.cv.nrao.edu
>>> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>>>
>> _______________________________________________
>> evlatests mailing list
>> evlatests at listmgr.cv.nrao.edu
>> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>
> _______________________________________________
> evlatests mailing list
> evlatests at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/evlatests
>