WIYN ODI User manual

The WIYN One Degree Imager
User Manual
Version: 4.1 Generated: February 27, 2017 SVN Rev: 183
c
 2012-2016 WIYN Consortium, Inc.

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Summary:
This is the ODI Operating Manual and is a recommended read for anybody using or operating
the ODI instrument.
Signo for Version 2:
WIYN ODI Scientist Daniel Harbeck July 1 2014
Revision History:
Version Date Author Changes
1 5/30/2013 D. Harbeck First public release.
2 9/11/2013 D. Harbeck Updated GUI gures, include bin-
ning, misc.
3 6/1/2014 D. Harbeck & Wilson
Liu Update on calibration stability, guide star prediction, pointing cor-
rection. Overall update.
4 July 2015 D. Harbeck Work in progress for 5x6 ODI.
2

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Notes:
1. no notes available at this time
Title picture:
A 5 millisecond exposure of the moon observed by 5x6 ODI.
3

Contents
1 Instrument Overview 7
1.1 InstrumentWalk-Through.............................. 8
1.1.1 InstrumentBody............................... 8
1.1.2 Major External Components & Infrastructure . . . . . . . . . . . . . . . 10
1.2 TheODIFocalPlane................................. 11
1.2.1 OTADetectors................................ 11
1.2.2 pODI with 13 detectors (2012 - 2014) . . . . . . . . . . . . . . . . . . . 11
1.2.3 ODI with 30 detetctors (2015- ...) . . . . . . . . . . . . . . . . . . . . . . 12
1.2.4 Coordinate Systems in ODI . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.5 Understanding ODI Images - A directory of FITS les . . . . . . . . . . 13
2 ODI Detector & Imaging Performance 17
2.1 ImageQuality..................................... 17
2.1.1 Sensitivity................................... 17
2.1.2 Read Noise, Gain, Full Well . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.3 Amplier and Pass Transistor Glow . . . . . . . . . . . . . . . . . . . . 20
2.1.4 DarkCurrent................................. 20
2.1.5 Crosstalk ................................... 21
2.1.6 Persistent Charge / Ghost Images . . . . . . . . . . . . . . . . . . . . . 21
2.1.7 Fringing.................................... 22
2.1.8 Charge Transfer Eciency of Lot 6 detectors . . . . . . . . . . . . . . . 22
2.1.9 Vignetting................................... 23
2.1.10ImagingQuality ............................... 23
2.1.11 Linearity & Photometry . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.12Astrometry .................................. 26
2.1.13 Calibration Stability and Artifacts . . . . . . . . . . . . . . . . . . . . . 29
3 Observing with ODI 35
3.1 Understanding the data ow in guided & OT corrected images . . . . . . . . . 36
3.1.1 Open-Loop OT Correction vs. Closed-Loop Telescope Guidance . . . . . 37
3.2 OverviewofODITools................................ 38
3.3 TheODIMenu.................................... 38
3.4 Launching Observations with the ODI GUI . . . . . . . . . . . . . . . . . . . . 39
3.4.1 TheLoginScreen............................... 39
4

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
3.4.2 Adhocobservations ............................. 40
3.4.3 Standard Dome Calibrations . . . . . . . . . . . . . . . . . . . . . . . . 45
3.4.4 DitherPatterns................................ 47
3.5 Manual Filter & ADC Control . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.5.1 FilterOperation ............................... 48
3.5.2 ADCOperations ............................... 48
3.6 TheODIFileBrowser ................................ 48
3.6.1 Image Actions and Image Display . . . . . . . . . . . . . . . . . . . . . . 50
3.7 TheOTAListener................................... 51
3.7.1 OverviewImage ............................... 53
3.7.2 Single OTA Display & Guide Star Selection . . . . . . . . . . . . . . . . 53
3.7.3 Correcting the Telescope Pointing Zero Point . . . . . . . . . . . . . . . 53
3.8 Telescope Guider Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
3.8.1 Non-siderealtracking............................. 56
3.9 WhereareMyData?................................. 57
3.9.1 QuickReduceData.............................. 57
3.9.2 Obtaininghelp ................................ 58
4 ODI Observing Strategies & Hints 59
4.1 Starting to Observe: A Walkthrough . . . . . . . . . . . . . . . . . . . . . . . . 59
4.1.1 Afternoon................................... 59
4.1.2 Evening .................................... 60
4.2 GuideStarSelection ................................. 61
4.2.1 Automatic Guide Star Selection . . . . . . . . . . . . . . . . . . . . . . . 61
4.3 ExposureTimeCalculator.............................. 63
4.4 Dithering ....................................... 63
4.5 Telescope Guiding & OT Corrections . . . . . . . . . . . . . . . . . . . . . . . . 63
4.6 Filters......................................... 64
4.7 PointingosetsforODI ............................... 64
4.7.1 Correcting Telescope Pointing using ODI . . . . . . . . . . . . . . . . . . 64
4.8 ObservingOverheads................................. 66
5 Afternoon Calibrations and the ODI Standard Calibration Plan 67
5.1 Summary of standard calibration plan observations . . . . . . . . . . . . . . . . 68
A Software Operations & Known Issues 71
A.1 KnownIssues..................................... 71
A.1.1 Stargrasp Controllers GH become unresponsive . . . . . . . . . . . . . . 71
A.1.2 Readout stalls for a while, but resumes . . . . . . . . . . . . . . . . . . . 71
A.1.3 Readout stalls and does not nish, with error messages such as DAC busy. 72
A.2 Walkthrough: Bringing up the ODI Observing Environment . . . . . . . . . . . 72
A.3 Starting and Shutting Down the ODI Computer System . . . . . . . . . . . . . 74
A.3.1 Restarting the JBOSS Application Server on odiweb ........... 74
A.3.2 Restarting the JBOSS Application Server on odiserv .......... 75
5

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
A.3.3 Starting the ODI Database System (Oracle) . . . . . . . . . . . . . . . . 75
B Starting Observations on the Command Line 77
C When Images Look Wrong 79
C.1 Flat Field: The Tertiary Mirror at Wrong Port . . . . . . . . . . . . . . . . . . 79
C.2 Condensation on Dewar Window . . . . . . . . . . . . . . . . . . . . . . . . . . 80
C.3 VacuumIonGaugeenabled ............................. 80
C.4 Residual Charge in Detectors I: After Powering On . . . . . . . . . . . . . . . . 81
C.5 Residual Charge in Detectors II: Darks . . . . . . . . . . . . . . . . . . . . . . . 81
6

Chapter 1
Instrument Overview
Telescope & Port WIYN 3.5m telescope, permanently mounted at f/6.3
Nasmyth port
Field of View pODI (2012-2014): 240×240
ODI (2015- ... ): 400×480
Optics 2-element eld attener & atmospheric dispersion
compensator
Detectors pODI (2012-2014): 13× STA OTA CCDs
ODI (2015- ... ): 30× STA OTA CCDs
full design: 64× STA OTA CCDs
Pixel Scale 12 µ m, 0.11  / pixel
OTA Cell Size 480 ×494 pixels; 5200 ×5400 on sky
Gaps between cells x: 28 pixels (3"); y: 11 pixels (1.3")
Gaps between OTAs about 21
Fill Factor per OTA 90%
Read noise ≤10 e
−
, 8.5 e
−
typ.
Gain typ. 1.1−1.6 e
−
/ADU, varies by cell
Dark Current 0.006 e
−
/sec/pixel typical
Exposure time Minimum ∼5 ms. Maximum limited by OTA amplier
glow. Linear from 20 ms to 60 min.
Filters Full size lters: g' r' i' z'
compatible w/ KPNO Mosaic lters w/ smaller FoV
Readout time ≤30 seconds sustained bias to bias
Humidity range ≤70%
Table 1.1: ODI in a nutshell.
7

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
1.1 Instrument Walk-Through
ODI consists of the instrument body that is mounted to the Nasmyth port of the WIYN
telescope plus supporting infrastructure throughout the WIYN facility. In this chapter we will
discuss the major components of the instrument hardware, the properties of the detectors and
how they are arranged on the focal plane, and the support infrastructure for ODI.
1.1.1 Instrument Body
The instrument main body is divided into three major components:
The Forward Corrector Element that mounts into the Nasmyth port. Its sole purpose
is to position the rst optical element of ODI, the forward corrector (also called Lens 1, see
Fig. 1.1 (a)).
The Instrument Support Package (ISP) (bolted to the forward corrector element) is the
backbone of the instrument and hosts the following elements (see Fig. 1.1):
1. Two rotatable prisms that constitute an atmospheric dispersion compensator. Each
prism is driven by a stepper motor and has end switches.
2. Nine lter arms that move into the telescope beam like a semaphore. The lter arms
are mounted to three supporting posts (labeled A, B, C). On each post there are three
lter arms mounted (labeled 1, 2, 3, where 1 is the one closest to the telescope). Each
lter arm has its own stepper motor and gear box.
3. The Bonn shutter.
4. Temperature sensors and a temperature multiplexer box.
The dewar is mounted to the ISP. Its main component is the vacuum vessel with the focal
plane. The dewar window is also an optically powered element, and forms the eld attener
in conjunction with the forward corrector lens. The focal plane with up to 64 OTA detectors
is cooled by four closed-cycle helium cryo-heads that are capable of removing 200W of heat
from the focal plane ( ≈120 W for a full set of 64 OTA detectors, plus about 60W radiative
load). You will also nd a turbo-molecular vacuum pump permanently mounted to the dewar.
The CCD controllers (IfA Stargrasp controllers) are mounted on two opposite sides of the
dewar, and each side drives one half of the detector array on the focal plane. Since the
CCD controllers dissipate a signicant amount of heat (1.2kW), they are enclosed in isolated,
glycol-cooled boxes to avoid creating a turbulent air ow in the dome (and thus diminishing
the seeing).
8

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Figure 1.1: Cut through ODI Model. Light enters from the telescope (right) through the forward
corrector lens (a), then passes the two prisms of the atmospheric dispersion compensator (b) and one
of the nine lters (c). Following the lters the shutter (d) is mounted just in front of the dewar window
(e); note that the dewar window also has optical power. In the dewar the detectors are mounted on
the focal plane (f). Noticeably mounted on the sides of the dewar are the two golden Stargrasp CCD
controller boxes (g). The hex panel and cable wrap (h) are mounted on the back of the instrument.
9

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
On the backplane of the dewar the hex-panel is mounted. The hex-panel is the interface for
various cable, glycol, dry air, and helium connections that are routed through a cable warp to
a boom hanging over the instrument.
1.1.2 Major External Components & Infrastructure
ODI depends on external infrastructure and components that either could not be mounted on
the instrument body due to weight constraints, or for other practical considerations. These
components are in particular:
1. ISP control computer : The ISP Control Computer is mounted to the telescope fork.
It controls the helium cryo-heads, the thermal regulation and heating of the focal plane,
overall instrument temperature telemetry, ADC prism and lter arm motors. The box
also provides power for the shutter and monitors the shutter's state. While not critically
dependent on cooling, the ISP box is glycol cooled for thermal management.
2. The Stargrasp Power Supply Box provides the electrical power to the Stargrasp
CCD controllers. Since the power supplies also dissipate a signicant amount of heat, it
is also linked to the glycol cooling system. Indeed, the control node for the glycol system
is located in that box. In order to keep the lines short, the main power lines are routed
directly from the power supply box to the hex panel, bypassing the cable wrap.
3. Heat removal from the instrument is achieved through the ODI Fluid Chiller & glycol
heat exchanger system. This system consists of the uid chiller in the WIYN utility
room, and a heat exchanger outside the WIYN building.
4. The ODI helium compressors in the WIYN utility room supply the cryo heads on
the ODI dewar with pressurized helium to cool the focal plane. The waste heat from the
helium compressors is dissipated in a heat exchanger unit outside the WIYN building.
Helium lines are routed through the telescope pier to the instrument.
5. The ODI data acquisition computer rack & network switches are located in the
WIYN computer cabinet. A dedicated 10-Gigabit switch routes data coming from the
32 Stargrasp ber pairs to the ODI data acquisition computer.
6. The ODI Observer Computer is a Mac Pro computer with two 30 inch displays,
located in the WIYN control room.
10

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
1.2 The ODI Focal Plane
1.2.1 OTA Detectors
A signicant design driver for ODI  in addition to the large eld of view  was to deliver
and actively increase the already excellent image quality that the WIYN site is known for. In
addition to passive measures (eld attener and ADC, thermal management), active tip/tilt
image compensation can be deployed using Orthogonal Transfer Array (OTA) CCDs. This
type of CCD detector allows to compensate for image motion electronically in the detector
(see Fig. 1.2). While regions in a single OTA detector can be used to trace the actual image
motion by monitoring bright stars, the charge in integrating cells can be shifted to follow the
optical image in real-time, and thus compensate for motion blur.
Figure 1.2: Concept of the Orthogonal Transfer Array (OTA) detector: The structure of the single
pixels (left) allows controlled charge movement up and down (as in a conventional CCD), and to the
left or right in a given cell. Thus, the charge image, as it builds up during an exposure, can be made
to follow the moving optical image. Each individual OTA detector (right) is divided into an 8×8
array of individual OT Cells (middle, ODI OTAs have actually 480 ×494 pixels per cell, about 10×10
on sky) that share some clocking and output lines, but with some restrictions can be considered to be
independent detectors. The video signal of cells with a bright star can be used to measure the actual
image motion, that can then be corrected for in the other remaining OTAs.
1.2.2 pODI with 13 detectors (2012 - 2014)
While ODI was originally designed for an array of 64 detectors to cover the full one degree eld
of view, the rst incarnation was only partially (hence p ODI) populated with 13 detectors.
11

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Figure 1.3: In pODI (2012 - 2014), 13 detectors are mounted on the focal plane (left). The imprint
on the sky is shown right; in that image north is up, east is left. Note that the actual pixel coordinate
system of pODI is swapped in the east/west direction. In the imprint map, defective OTA cells are
marked with red crosses; cells that are unable to perform ecient OT corrections are marked with
yellow crosses.
The vacant slots in the focal plane are covered with blackened aluminum blanks to reduce the
possibility of reections (see Fig. 1.3).
1.2.3 ODI with 30 detetctors (2015- ...)
During the Winter of 2014 / 2015 17 additional devices were added to the ODI focal plane.
The new devices originated mostly (16 of them) from a new OTA production Lot ("Lot 7"),
where the low light level CTE problem has been resolved. The new Lot 7 detectors were
arranged into a 4x4 array, with the remaining Lot 6 detectors padding the central square to a
total 5x6 array of detectors, or a eld of view of 400×480 on sky.
During production of the Lot 7 devices it was discovered that aging of the ceramic carriers
caused an issue with epoxy wetting the entire carrier surface. This results in some epoxy voids
in the detectors, that can be seen in images as extended structures. At the time of this writing,
the epoxy underll areas do at-eld correctly.
Whereas the new Lot 7 detectors are not impacted by the low light level problem, the remaining
Lot 6 devices are, and when using the full eld of view of the 5x6 conguration observations
should be planned accordingly.
12

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Figure 1.4: A total of 30 detectors ( a mixture of one Lot3, thirteen Lot 6, and 16 Lot 7 detectors)
is mounted in the 5x6 ODI (2015 - ) focal plane. As in pODI, the logical pixel coordinate system is
swapped with respect to the standard east-west orientation.
1.2.4 Coordinate Systems in ODI
A pixel in ODI can be uniquely identied by three coordinates:
1. The OTA detector, which is identied by its X/Y position on an 8×8 grid on the focal
plate, as identied in Fig. 1.5.
2. The Cell within an OTA, which is identied by its X/Y position on an 8×8 grid on the
OTA, as identied in Fig. 1.6. Note that the cell Y axis is anti-parallel to the OTA Y
and pixel Y axis.
3. The pixel X/Y coordinate within an OTA cell.
Thus, a pixel can be identied by these three coordinates pairs: OTA X/Y, Cell X/ Y, and
Pixel X/Y. The ODI raw FITS format, and all instrument software, work in the natural pixel
coordinate system. Therefore, images at the telescope will display north up, and east right.
1.2.5 Understanding ODI Images - A directory of FITS les
The complexity of the focal plane and detector architecture is reected in the raw ODI images.
The image size of ≥2 GB for a full ODI image puts further constraints on a usable le format,
forcing to split a full readout into several les. As a consequence, each ODI image readout is
stored in an individual directory. Within each directory there are a large number of les, but
at the core there are *_XY.fits les that contain the actual image. There is one MEF FITS
13

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
A1 dev0 A3 dev0 B1 dev0 B3 dev0 C1 dev0 C3 dev0 D1 dev0 D3 dev0
A2 dev0 13946
A4 dev0
13835
B2 dev0
13902
B4 dev0
13792
C2 dev0
8101
C4 dev0 D2 dev0 D4 dev0
A1 dev1 13837
A3 dev1
13880
B1 dev1
13947
B4 dev1*
13879
C1 dev1
13838o
C3 dev1 D1 dev1 D3 dev1
A2 dev1 13901
A4 dev1
17275
B2 dev1
17166
B3 dev1
17167
C2 dev1
17122
C4 dev1 D2 dev1 D4 dev1
H3 dev1 13923
H1 dev1
17144
G3 dev1
17121
G1 dev1
17341
F3 dev1
17187
F1 dev1 E3 dev1 E1 dev1
H4 dev 1 13974
H2 dev1
17297
G4 dev1
17231
G2 dev1
17277
F4 dev1
17190
F2 dev 1
17253
F1 dev0
E2 dev1
H3 dev0 13968
H1 dev0
17189
G3 dev0
17278
G1 dev0
17234
F3 dev0
E4 dev1
E3 dev0 E1 dev0
H4 dev0 H2 dev0 G4 dev0 G2 dev0 F4 dev0 F2 dev0 E4 dev0 E2 dev0
OTA X
OTA Y
01234567
FS1
FS2
0
1
2
3
4
5
6
7
View from behind dewar lid onto detector pins
H G F E
A B C D
ODI BOTTOM
1 2 3 4
5x5 ODI as build Jan 2015
ODI Top
E
N
x
y
* B4, B3 dev1 are currently swapped in pODI
Figure 1.5: Left: pODI focal plane layout. There are 13 OTA detectors populated out of the 64
possible locations. Right: 5x6 ODI focal plane layout. To the left and right, two focus sensors are
mounted. For each of the OTA detectors, the device's serial number and the Stargrasp controller
address are given. The Stargrasp controller address consists of a letter (A through H, identifying a
chassis of four controller boards), the board number in that chassis, and one of the two devices that a
board can drive. Individual detectors are shortly identied by their X and Y coordinate on the focal
plane (numbers with green background).
0/0 1/0 7/0
0/7 1/7 7/7
Connection Pad
x
y
E
N
Looking through the focal plane from behind the instrument towards the telescope
Figure 1.6: Coordinate system in a single back-
side illuminated OTA device. Pixel coordinates
increment from the lower left to the upper right.
Cells are identied by numbers from 0 to 7. Cell-
X coordinates increment from left to right, cell-Y
coordinates increment from the top to the bot-
tom, i.e., opposite to the pixel coordinates in the
y direction.
14

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Figure 1.7: Content of the raw
image directory for exposure
o20121112T030221.0 : For each of the 13
detectors there is one FITS le, a factor-8
and factor-16 binned jpeg preview image
each, and a .txt le containing a dump
of the ts header. The metainf.xml
le contains a directory manifest, post-
observation user comments, and addi-
tional telemetry data.
15

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
le present per detector that has been read out; in the case of pODI there are consequently
up to 13 FITS les. The extension XY denotes the location of an OTA detector on the focal
plane.
The image of each cell within a given detector is stored as an individual FITS extension, i.e.,
each single MEF FITS le has 64 extensions, one per cell.
Auxiliary data are available in the image directory: jpeg images that have been down-sampled
by factors of 8 or 64, extracts of the ts headers in .txt les, and a metainf.xml manifest le
that contains a list of all the FITS les with checksums, additional instrument telemetry data,
and user annotations for an exposure.
Images with guide star readouts can contain additional directories:
1. preimage
This directory contains an entire image readout containing the preimage that was used
to select a guide star.
2. expVideo
This directory will contain the guide star video stored in FITS format, and centroid /
photometry data stored in a FITS table.
3. temp
This directory contains auxiliary les supporting the data acquisition and debug logles.
Details on the image format can be found in the ODI-PPA-ICD-0001.
16

Chapter 2
ODI Detector & Imaging Performance
2.1 Image Quality
2.1.1 Sensitivity
ODI at WIYN's sensitivity is predicted based on modeling all optically relevant components.
The throughput model includes: the atmospheric extinction and telluric absorption lines, the
three telescope mirrors, absorption losses in the ODI optics (two lenses and two ADC prisms),
anti-reection coating performance, and the quantum eciency of the OTA detectors. The
predicted throughput has been veried on sky in the g', r', i', and z' bands within reasonable
error margins, and the dierences we found are included in the throughput model as a fudge
factor. The resulting as-build throughput of ODI is shown in Fig. 2.1.
The blue cut-o in ODI's sensitivity is governed by special glass (O'Hara PBL6Y) in the
ADC prisms. The fall-o in the red is driven by the vanishing quantum eciency of the ODI
detectors. The peak throughput of order of 55% is the sum of all losses in the system; however,
the most signicant peak throughput limit is dened by the losses in the three WIYN mirrors.
2.1.2 Read Noise, Gain, Full Well
Each of the 64 cells on a single OTA detector has an individual output amplier, and depending
on the location on the device, voltage supply lines dier in length, resulting in some voltage
drops. Hence, the variation of the gain of the individual OTAs is signicant. Typical gains
for OTA detectors range from 1.2 to 1.6 e− /ADU. The read noise is typically below 10e− , and
can be as low as ≤6e− . The distribution of read noise is shown in Fig. 2.2.
The gain of detectors varies with temperature; while this is usually not a problem for temperature-
regulated detectors, the ODI detectors dissipate a signicant amount of power (2W per detec-
17

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Figure 2.1: WIYN telescope and ODI throughput at an airmass of one. The throughput curve is
based on modeling the throughput, reectivity, and quantum eciency of the optics and detectors,
and an atmospheric extinction model. The model has then been adjusted to match actually measured
zero points in the ODI lters.
tor), and hence their temperature will oscillate for one hour after powering on detectors. In
order to guarantee gain stability of ODI, there must be an hour wait time between
powering on detectors and scientic (including calibration) data acquisition.
Gain, noise when binning
Binning with ODI has been tested and is ready to be used. However, as this mode is new and
not yet fully characterized, it should be considered as shared risk operations.
With binning turned on, the ODI detector array will be read out in a 1×2 binned mode.
Physical binning at this time is applied only in the parallel clocking direction; physical binning
in the serial register direction is not oered at this time due to increased read noise. However,
for convenience, serial binning will be simulated by averaging in the serial readout direction,
and data are stored 2times2 binned.
Averaging data in the serial direction impacts the read noise and gain of the logical pixel, since
each logical pixel contains two physical read pixels: The read noise will increase by
√2
, and
18

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
Figure 2.2: Top: Heat-map representation of the gain and noise for each individual cell in the pODI
focal plane. Note the typical increase in the gain to the upper left of an individual OTA. The OTA at
the upper left of the central array has a dierent doping of the ampliers, hence the dierence in gain
and noise.
Bottom: Histogram of read noise and gain in the pODI focal plane. The upper and lower halves of the
detectors are served by separate controller chassis; however, no signicant dierence in performance is
apparent.
the eective gain will increase by a factor of 2 (each pixel averages two readouts). The signal
to noise ratio in the read noise limited case will hence increase by
2
√(2) =p(2)
.
The benet of binning is hence the decrease of time needed to the sky noise limit by a factor
of
√2
. A second potential benet under investigation is that the critical background level to
overcome the low light level CTE problem could be achieved twice as fast.
19

ODI-MAN-1001 ODI User Manual
DRAFT Version 4.1, Expires n/a
2.1.3 Amplier and Pass Transistor Glow
Transistor glow in ODI's OTA detectors remains the major limitation. The most signicant
source of glow are the transistors in the output ampliers of each cell. Their glow is pro-
hibitively high, and would dominate every exposure with exposures times above one minute.
A work-around to control the amplier glow is to reduce the drain voltage of the output am-
pliers from typically 24V to 10V during any idle times, and power the ampliers on only
while detectors are read out. Therefore, output amplier glow is reduced to a bias eect; an
example bias is shown in Fig 2.3, left.
In order to facilitate the switch between science integration and active video / OT shifting
mode, the parallel clocks for each cell can be driven by either a standby parallel voltage,
or actively clocked parallel signals. The multiplexing between these two parallel voltages is
handled by pass-transistors, and these also glow, albeit at a lower level than the transistors
in the output ampliers. Yet, in a long exposure, they start to dominate as can be seen in
Fig. 2.3, right.
Figure 2.3: Example of transistor glow in a new implant level Lot 6 OTA device. Output amplier
glow in a bias (left), and pass-transistor glow in a 900 seconds dark (right).
2.1.4 Dark Current
A long dark exposure will have two components: The classical dark current which is generating
charge in all pixels, and localized accumulating low-level amplier and pass-transistor glow.
In this section we describe the classic dark current.
20