[evlatests] EVLA Tests Meeting
Vivek Dhawan
vdhawan at nrao.edu
Fri Oct 30 14:37:10 EDT 2009
Here is a 'state of play' overview for today's test meeting.
We may not get to all of these, so I thought I'd spam y'all.
-------------------
User-centric Checklist
Fixes put in and tested:
Correction for amplitude non-closure from large ~100ns
delay errors across single channel (Aips/Eric)
F_shift correctly accounted in turning delay into phase
(Barry)
Phase transfer accuracy now:
< 60 deg max error, whole sky, at L,S,C,X.
< 60 deg max error at K, Ka, Q across 35, 25, 15 degrees.
i.e. image position better than 1/6 beam for cal/target
separation closer than quoted number. This is limited by
antenna hardware (aside from weather). Correlator model
errors are unmeasurable now - 5 ps or better on timescales
of a few minutes to a few hours.
Image dynamic range, at tracking center.
No sweat: 30K:1 or better, L, C, X band
without BLCAL
Noise reduction with N_chan: L:not. C: sorta - see next.
Reaching thermal noise
I/V image on a few-Jy point source:
factor 5 or more @ L
Factor ~2 @ C, X.
Blank field: within factor 2.
Possible causes:
Delay rocking - remove with short fringe fit.
probably not the major cause.
Bandpass stability: few parts in 1000, A&P.
slow variation over several minutes
self-cals out, I think.
Beam squint - not important(?) on point source
at phase center, w reasonable pointing.
(?Check via Sanjay's software in CASA)
D*D terms, unpol source: should partly self-cal out.
frequency dependence may not self-cal(?)
Source spectral index: should partly self-cal out.
(each subb is normalized & self-cal'd.)
Wobbles - known to violate closure, so let's make
them the scapegoat till they're gone.
* Period shortens as F_sky increases, LCX.
* Amplitude decreases as F_sky increases.
* Wobbles are much faster than delay rate.
* There is no delay wobble, only A&P wob.
---------------------
More information about the evlatests
mailing list