Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.stsci.edu/ftp/instrument_news/FOC/Foc_isr/foc_isr076.ps Дата изменения: Tue Oct 4 23:21:00 1994 Дата индексирования: Sun Dec 23 05:40:26 2007 Кодировка: IBM-866 Поисковые слова: 47 tuc

INSTRUMENT SCIENCE REPORT
FOCн076
TITLE: FOC f/96 Detector Focus
AUTHOR: P. E. Hodge, M. Miebach DATE: 4 October 1994
ABSTRACT
Observations of NGC 104 (47 Tuc) were used to determine the optimum firstнstage voltage
for the FOC f/96 image intensifier. A voltage of 10.69 KV was found to give the sharpest
images, while the default voltage was 11.25 KV. The brightness in the central pixel was 15
to 20 percent greater at the optimum voltage compared with the default voltage.
0 DISTRIBUTION:
FOC Project: D. Eaton, K. Kalinowski, B. G. Taylor, R. Thomas
IDT: R. Albrecht, C. Barbieri, A. Boksenberg, P. Crane, J.M. Deharveng, M.J. Disney, P. Jakobsen,
T. Kamperman, I.R. King, C. Mackay, G. Weigelt
SIB: D. Baxter, C. Cox, P. Greenfield, W. Hack, A. Nota, S. Osmer, F. Paresce, P. Stanley, All
Instrument Scientists
SCARS: P. Hodge
SESD: M. Miebach, W. Safley
SPD: J. C. Blades, R. Jedrzejewski, F. Macchetto
USB: A. Saha
RSB: D. Gilmore, D. Golombek
ST/ECF: H.нM. Adorf, P. Benvenuti, A. Caulet, R. Fosbury, R. Hook

1 Introduction
Changes in the voltage of the first stage of the FOC image intensifier affect the sharpness
of the detected image. The image sharpness should not depend on the voltages across the
second and third stages; these should affect the event size instead. The voltage for the image
intensifier first stage is varied by changing both the 12 KV and 36 KV values such that the
difference between the two remains constant. This keeps the second and third stage voltages
the same. In this report we describe how we found the optimum voltage for the f/96 relay
by assessing image quality.
The f/48 detector focus was reported on by Towers and Miebach in Instrument Science
Report FOCн055.
2 Observations
Observations with rootnames x1f10101t through x1f1010ft were taken with the FOC f/96
relay in proposal 4195 on 1993 Oct 14. The first observation x1f10101t was taken with an
objective prism, and this image was not used in this study. The other images were taken
in pairs; that is, two images at each of seven different high voltage settings. The exposures
were all in fine lock, filter F220W was used, and the exposure times were 896.625 seconds.
The image format was 512 by 512 centered.
Previous f/96 detector focus observations were taken in proposal 2804 on 1992 May 29. These
observations showed that at that time the default high voltage was close to the optimum
value. Previous f/48 detector focus observations were taken in proposals 3204, 3361 and
4007.
The 12 KV and 36 KV values for the 14 focus images are given in table 1. The columns
labeled ``12 KV'' and ``36 KV'' are the voltages in kilovolts, while the columns ``NI212KVF''
and ``NI236KVF'' are the values from the unique data log (UDL) header.
2

Table 1: Voltages used for focus measurements.
root names 12 KV 36 KV NI212KVF NI236KVF
x1f10102t & x1f10103t 11.25 36.74 11.364 36.878
x1f10104t & x1f10105t 10.97 36.47 11.100 36.615
x1f10106t & x1f10107t 10.69 36.19 10.836 36.352
x1f10108t & x1f10109t 10.40 35.90 10.559 36.088
x1f1010at & x1f1010bt 10.12 35.62 10.295 35.787
x1f1010ct & x1f1010dt 9.83 35.33 10.019 35.524
x1f1010et & x1f1010ft 9.55 35.05 9.780 35.260
3 Data Analysis and Interpretation
Twelve stars of roughly the same brightness in the raw (*.d0h) images were selected to use
for determining the best focus. The highest count rate among these stars was 0.31 c/s/p,
so saturation should not have been significant for 512 by 512 format. The X and Y pixel
coordinates of the stars in image x1f10106t.d0h are listed in table 2.
Figure 1 shows these 12 stars extracted from the 14 images. For each star, a 9 by 9 pixel box
centered on the star was extracted. Each column in figure 1 shows the same star in the 14
images. Each row shows the 12 stars in one image. The bottom two rows were taken with
the default voltage of 11.25 KV. The bottom row is from image x1f10102t, the next row is
from x1f10103t, and so on up to x1f1010ft for the top row.
In the foccs package of stsdas in IRAF, the tasks fextract, fsquares, fradius, fgauss,
and fhyper were used to find the optimum voltage by fitting curves to the sharpness and to
the characteristic radius.
The fsquares task computes the ``sharpness criterion'' for the selected stars, putting the
output in a text file. This file was then edited to put the voltages (see Table 1) in the first
column. The fgauss task then fits a Gaussian to the sharpness as a function of voltage.
The plot of the sharpness data and fitted curve are shown in figure 2. The optimum voltage
based on the sharpness is shown in the upper right of the figure.
Another criterion for determining best focus is based on computing a characteristic radius
of the star images. This is done with the fradius task, and fhyper is then used to fit a
hyperbola. The plot is shown in figure 3.
3

Table 2: Positions of stars in x1f10106t.d0h used for focus measurement.
X Y
349.89 36.82
87.19 129.61
19.76 200.47
288.54 231.04
129.39 232.82
361.22 265.07
32.86 275.47
58.95 298.38
120.07 359.27
265.40 385.59
92.31 399.63
448.92 404.53
For figures 2 and 3 the independent variable is the 12 KV value from table 1, but the unit is
volts. Two points (small squares) are shown at each of the seven different voltages used in
these observations, since there were two images at each voltage. The voltage was decreased
during the test, so the first pair of images is plotted near the right hand side. This pair
corresponds to the first two rows in figure 1. Having a pair of images at each voltage gives
a better feeling for the accuracy of the fit and increases our confidence in the results.
While it is clear that a voltage near 10.69 KV will give sharper images than the current
default voltage, it is of interest to know how much we can expect the peak count rate to
increase. The average was taken of the data values of the four brightest pixels for a set of
five stars. The star positions and the results are listed in table 4, and figure 4 shows these
sums plotted against the variation of the 12 KV power supply output voltage. Judging from
the figure, an improvement of 15 to 20 percent can be achieved by lowering the voltage from
11.25 to 10.69 KV.
The images taken at higher voltage are rotated about a half degree clockwise with respect
to the images taken at lower voltage, with a center of rotation near X=260, Y=180. The
rate of rotation is about 0:3 degree per kilovolt.
In order to determine whether the high voltage value affects the detector sensitivity, the
apphot package was used to measure the brightnesses of ten stars in each image. For the
most part, the same stars were used for apphot as for the focus, but some changes were
4

Table 3: Positions of stars in x1f10106t.c1h used for aperture photometry.
X Y
105.57 130.88
139.81 234.04
284.11 228.48
350.38 261.92
229.47 276.14
50.32 278.99
73.00 301.80
126.18 363.16
257.76 387.67
429.79 406.19
made because of stars in the background annulus. In addition, apphot was run on the
calibrated rather than on the raw images. Table 3 lists the pixel coordinates of the stars
in x1f10106t.c1h. Five pixels was used as the radius of the star annulus and also as the
width of the background annulus. The resulting magnitudes were converted into counts
(with arbitrary normalization), and the counts for the stars within each image were summed
to give one number for each image. These are plotted as a function of voltage in figure 5.
There are two images at each voltage setting, and the two values for each voltage are plotted
with different symbols, a square and a cross. The error bars are the values given by apphot,
and they do not include any systematic effects. It is seen that the sensitivity drops off at
lower voltages, but the drop is not significant except below the voltage that we have found
to give the sharpest images.
5

F/96"
FOC4195
"Sum
of
four
brightest
pixels"
Star
#1
Star
#2
Star
#3
Star
#4
Star
#5
Xнcoord:
92
33
87
289
361
Yнcoord:
399
275
130
230
264
12
KV
Average
Std
Var
X1F10102
11.26
405
337
399
364
346
370
30.69
X1F10104
10.97
412
349
414
471
474
424
51.42
X1F10106
10.69
440
385
485
430
431
434
35.58
X1F10108
10.40
393
343
342
457
389
385
47.11
X1F1010A
10.12
308
289
351
392
349
338
40.33
X1F1010C
9.83
220
233
257
214
203
225
20.72
X1F1010E
9.55
220
241
257
210
199
225
23.48
min
9.55
220
241
257
210
199
225
9.83
220
233
257
214
203
225
10.12
308
289
351
392
349
338
10.40
393
343
342
457
389
385
10.69
440
385
485
430
431
434
10.97
412
349
414
471
474
424
opt
11.26
405
337
399
364
346
370
Table 4: Stars used to get sum of four brightest pixels.
6

Figure 1: Twelve stars extracted from 14 images.
7

9500 10000 10500 11000
.0018
.002
.0022
.0024
.0026
12 KV intensifier HV setting in volts
sharpness
f/96 detector focus (4195) нн> 10706.4
Figure 2: Image sharpness vs 12 KV high voltage.
8

9500 10000 10500 11000
2.35
2.4
2.45
2.5
2.55
2.6
12 KV intensifier HV setting in volts
radius
f/96 detector focus (4195) нн> 10714.5
Figure 3: Characteristic radius vs 12 KV high voltage.
9

9500 10000 10500 11000
250
300
350
400
12 KV intensifier HV setting in volts
average
of
four
brightest
pixels
f/96 detector focus (4195) нн> 10781.0
Figure 4: Average of counts in four brightest pixels vs 12 KV high voltage.
10

9.5 10 10.5 11 11.5
15
16
17
18
Voltage (KV)
Sum
of
Counts
STScI/IRAF V2.10EXPORT hodge@bowline.stsci.edu Thu 14:32:48 09нDecн93
f/96 Gain vs 12 KV
Figure 5: Detector sensitivity vs 12 KV high voltage.
11