The new Leach electronics were tested at the Northern 2MASS Facility
on the night of 980925 UT. Several calibration scan sets were taken,
as was a block of survey scans. The calibration scans and three survey
scans were ftp'd to IPAC and processed using 2MAPPS v2.1. The pipeline
output from these runs was evaluated and summarized below.
- Darks and Flats processed normally, although no comparison was
made against current canonicals. The dark frame statistics
are:
Band/ReadType: J/1 H/1 K/1 J/2 H/2 K/2 (evening)
Avg Dark Sigma: 4.38 4.40 4.38 2.37 1.99 2.15
St Dev of Above: 1.09 1.83 2.56 0.90 0.78 10.81
Visual inspection of the individual flattened frames for cal scans
008-011 revealed no unusual source-related artifacts in J, H or Ks.
Known detector-related artifacts such as latent images appear nominal.
However, several people reported noticing a fixed-pattern noise in the H
frames. This is a diagonal hash pattern with about a 10 pixel period
that is very subtle and
intermittent, not visible in all frames. The peak-to-peak amplitude is
probably less than 2-3DN.
Gene Kopan examined the average R1 frame noise and compared them
to frame noise from 980919n, and found the electronics noise
to be slightly lower with the Leach electronics:
980925n Leach electronics test:
hi0140232 / diag noise ~10 cam pix period
at an h background of ~2500 dn
average r1 frame noise:
night J H K
980925n (Leach) 11.929 12.075 12.263
980919n 12.732 12.218 12.672
so the electronics noise seems to be slightly lower.
Coadds appear nominal, except for diagonal fixed pattern noise in
some H coadds. Tracey Evans reported that source related artifacts
are consistent with those traced by MAPCOR, which means they are
consistent with data taken with existing electronics.
Tom Jarret measured the noise in the coadds as reported by
GALWORKS and concluded that the noise levels are consistent with
earlier data. Plots of the backgrounds and background sigmas for
scans 008-013 are
here
Close examination of the H-band coadds revealed a pattern of repeating
holes which are produced by low coverage. The hole occurs at
column 276, and repeats ever 82 pixels in the in-scan direction. Coadds
formed from data taken on the night of 980919n were also examined,
and the holes also appear in those images. Thus, we conclude that
the H-band coadd holes probably result from noisy pixels in the H-array
that are being turned off that appeared after the most recent thermal
cycle. They are probably not associated with the Leach electronics.
The following 3 figures show the photometric repeatability plots
for each of the 3 calibration scan sets taken with the Leach electronics.
Scans 008-013 and 023-028 covered the calibration field P182-E
and scans 017-022 covered P290-D. The SNR=10 levels
are indicated with horizontal lines, and the Level 1 Requirement magnitude
fiducials are indicate with vertical lines.
Scans 008-013 Photometric Repeatability
Scans 017-022 Photometric Repeatability
Scans 023-028 Photometric Repeatability
The Level 1 Requirements for sensitivity are met in all three scans, although
only marginally so in scans 008-013, probably due to
poorer seeing and higher backgrounds right after twilight.
The repeatability plots also show the effective completeness levels
in the three bands, and it appears that nominal completeness is
satisfactory. Again, it is marginal in H band in the P182-E scans.
Photometric precision is 2-3% for all cases.
The photometric zero point offsets
for these scans are -0.031, 0.004 and 0.015 magnitudes at J, H and
Ks, respectively, so the instrumental magnitude scale in
the repeatability plots is very close to the calibrated scale.
Zero point offsets close to zero indicate that the system throughput
is consistent with historical values.
