[evlatests] 10-second tuning failures at X-band

Ken Sowinski ksowinsk at nrao.edu
Mon Aug 31 11:00:00 EDT 2015


On Mon, 31 Aug 2015, Walter Brisken wrote:

>
> If it is a correlator setup issue, it probably is happening within scans as 
> well, correct?  If so, things would probably not look too bad (the previous 
> model would just be extended another 10 seconds I presume), but there would 
> be differences and this would break accountability of the delay model.  Might 
> show up best at high frequencies on long baselines pointing near zenith where 
> the second derivative of the delay model is greatest.
>
> -Walter

The delay and phase information sent to the hardware will "flywheel" if
no new delay polynomial has been received at the station board when
the previous one expires.  An alert or log message might help us see
if that is the problem.

Ken


> On Mon, 31 Aug 2015, Ken Sowinski wrote:
>
>>  On Mon, 31 Aug 2015, Vivek Dhawan wrote:
>> 
>> > 
>> > 
>> >   Sounds lke L302's failing to lock for 10 sec, a fairly common transient
>> >   effect on 1% of the 224 units. But why it is only at X band I cannot
>> >   explain
>> >   - I assume you changed bands many times after the reference pointing?
>>
>>  I don't think so; that would break two IFs at a time.  My money
>>  is on the distribution of current delay and phase models and
>>  sideband information within the station board.  Tracking it down
>>  will probably not be easy.  For Rick: when it fails for an IF,
>>  is it true that all subbands for that IF behave the same way?
>>
>>  Ken
>> 
>> 
>> >   On Mon, August 31, 2015 08:26, Rick Perley wrote:
>> > |        The calibrator survey run, taken last October, has revealed the
>> > |   continued presence of short-lived tuning failures.  The 
>> > |   characteristics
>> > |   are quite simple:
>> > | 
>> > |        1) No fringes at the beginning of a scan, lasting exactly 10
>> > |   seconds.  After this interval, full-valued fringes abruptly occur.
>> > |   Hence, this is not a pointing problem.
>> > | 
>> > |        2) Roughly 20% of scans are affected, nearly always a single
>> > |   antenna, occasionally two.  This seems consistent with a random
>> > |   process.  There was no clear tendency for any particular antenna to 
>> > |   fail
>> > |   more often.
>> > | 
>> > |        3) The four IFs failed independently, with equally likelihood --
>> > |   failures are always on a single IF.
>> > | 
>> > |        4) So far as I can determine, this problem only occurs at 
>> > |   X-band.
>> > |   (I have checked P, L, S, and C, so far).
>> > | 
>> > |        5) The on-line flagging system (unsurprisingly) does not flag 
>> > |   these
>> > |   failures.
>> > | 
>> > |        These X-band observations followed X-band referenced pointing,
>> > |   which utilized a different tuning and correlator configuration.
>> > | 
>> > |        There was also a single incidence (out of 182 scans) where the
>> > |   *last* 10 seconds abruptly lost fringe power.
>> > | 
>> > | 
>> > |   _______________________________________________
>> > |   evlatests mailing list
>> > |   evlatests at listmgr.nrao.edu
>> > |   https://listmgr.nrao.edu/mailman/listinfo/evlatests
>> > | 
>> > 
>> > 
>> >   _______________________________________________
>> >   evlatests mailing list
>> >   evlatests at listmgr.nrao.edu
>> >   https://listmgr.nrao.edu/mailman/listinfo/evlatests
>> > 
>> _______________________________________________
>>  evlatests mailing list
>>  evlatests at listmgr.nrao.edu
>>  https://listmgr.nrao.edu/mailman/listinfo/evlatests
>> 
>


More information about the evlatests mailing list