[daip] FRING Problem

[Thomas Krichbaum]

Dear AIPS-Friend,
Dear Eric,

when fringing 3mm-VLBI data from the Bonn correlator, I realize a possible problem
with FRING in versions NEW and TST (Dec. 04).

If I fringe the full time range of the experiment (setting TIMER 0) and with the usual parameters, FRING aborts
after the 3rd VLBI scan, saying:
....
hp218 > FRING1: Standard RMS errors (deg, mHz, nsec):
hp218 > FRING1: Ant(01): Phas=  0.16 rate=     0.003 delay=     0.049
hp218 > FRING1: Ant(04): Phas=  0.23 rate=     0.004 delay=     0.070
hp218 > FRING1: Ant(05): Phas=  0.69 rate=     0.011 delay=     0.203
hp218 > FRING1: SOLUTION INTERVAL IS TOO LONG. TRY LESS SOLINT.
hp218 > FRING1: OR INCREASE PREAV TIME APPLYING UVAVG.
hp218 > FRING1: Purports to die of UNNATURAL causes
hp218 > FRING1: hp218        31DEC04 NEW: Cpu=      94.5  Real=     239

When looking at the SN-table, one however sees that phase, delay and rate 
of the previous scans and on that particular scan, were Fring aborted, were correctly
written into the table.

If instead of versions NEW/TST I use FRING from version 'OLD' (Dec. 03),
the whole experiment is fringed without major problem.


In order to find possible reason for the  behaviour of FRING in version NEW/TST I did some more testing.
Fringing the data with different solution intervals in the range of 1-5 mins or other search windows (dparm) does not help.
If I however specify the time range exact around the scan, on which FRING aborted,
that scan is fringed correctly. If I inrease the time range parameter, eg. including also the following
scan, fringe aborts again. 

I also looked at the CL and NX table, but could not find anything obviously wrong.

I also looked in 2 other experiments (from the same correlator), the problem remains the same, 
of course at other scan times. So what I now do is to fringe the experiments until
fring aborts the first time. Then I fringe that particular scan again, specifying its exact time range and
write the solution in the existing SN table. Then continue from the next scan until the end. Sometimes 
FRING fringes another 5 to 10 scans, before it aborts again a 2nd or even 3rd time. So this is very painfull.
In summary, one can always fringe the problematic scans by setting an appropriate timerrange.

Was there any significant change in Fringe between Dec. 03 and Dec. 04 ? Is this a known problem ?

best regards, 
                 Thomas

PS: This are the main FRING PARMS:
DOCALI 1; GAINUS 1; SOLI 4.3; APARM 2 0 0 0 0 2 4.5 0 1 0; DPARM 1 200 200 1 0
I also tried DOCALI 2; APARM(5) 2; different SOLI and DPARM(2) and (3). Always the same
behaviour.

-- 
Thomas P. Krichbaum                       
Max-Planck-Institut für Radioastronomie 
Auf dem Hügel 69                        
D-53121 Bonn / Germany

Tel.:   (Germany +49)-228-525-295
Fax :  (Germany +49)-228-525-229
e-mail: tkrichbaum at mpifr-bonn.mpg.de
URL:    http://www.mpifr-bonn.mpg.de/staff/tkrichbaum