Instrument Page:


Overview

Wavelength Coverage

Sensitivity

Integration
Times

ROGUE

MPE-3D Pages

MPE-3D Homepage

References and Links

Data Reduction

SPECIAL NOTES FOR SPECIFIC
RUNS

Public FTP-Site

Recent Results from 3D

 

MPE-3D Instrumental Characteristics:

Basic Overview

The instrument is based on a 2562 NICMOS-3 Rockwell array and can simultaneously obtain 256 H- or K-band spectra at R = 1100 or 2100 from a square 16 x 16 pixel field on the sky. The pixel size depends on the specifics of the telescope, but typically ranges between 0.3" and 0.5" per pixel.  On the 3.9 meter Anglo-Austalian Telescope the nominal pixel scale is 0.4" per pixel (see also comments under the section on ROGUE).

The basic working principle for 3D can best be illustrated by the following diagram (from Thatte et al. 1994)
 
 







In this figure, an astronomical source has a morphology which depends on wavelength (A). Its broadband morphology is illustrated in (B).  3D uses an image slicer to "dissect" the source into 16 slitlets (5 of which are depicted here) and to stack the slitlets on top of each other, forming a pseudoslit (C).  This pseudoslit is then passed through a grism, dispersing it in wavelength and forming a two dimensional "image" of the source (D), which is imaged on the detector.  In post-observation analysis this two dimensional format is converted into a data cube (E), which can be collapsed to recover the broadband morphology (F).
 
 

Wavelength Coverage

3D has a remotely controllable grism slide as part of its design. This slide holds two grisms; thus, at any point in time the observer can choose between the two grisms which have been installed for that run. Although we have five grisms (see Table below), only two are installed for any given run (to avoid loss of observing time resulting from a grism exchange).  Please note that the wavelength ranges in the Table below are not to be viewed as absolute.  Rather, the values used during a run may vary by, typically, ±0.02 µm, owing to last minute adjustments in the observatory prep room.  Our stock of grisms is as follows:
 
3D's Suite of Grisms
Grism
low(µm)
high(µm)
R=/
Hlow
1.48
1.78
1250
Hhigh
1.55
1.75
2100
K
1.94 2.41
1100
Kshort
1.95
2.18
2100
Klong
2.17
2.43
2100

Sensitivity

The total instrument transmission (excluding of the telescope, but including the tip-tilt optics the grism, and the filter is XX%. The quantum efficiency of the detector itself is 65%. Overall, the sensitivity of 3D is given by the values in the Tables below
 
Point Source Continuum Sensitivity
 Point Source magnitude
0.25 pixels
0.4 pixels
Seeing
Seeing
H-band
K-band
0.5 1.0
1.5
0.5 1.0 1.5
 
15.5
15
 
80
160
310
620
690
1380
<60
60
120
240
270
540
 
16.5
16
 
480
960
1900
3800
4280
8560
190
380
740
1480
1670
3340
 
17.5
17
 
3000
6000
12000
24000
27100
54200
1200
2400
4700
9400
10600
21200
 
18.5
18
 
18900
37800
75500



7400
14800
29500
59000
66380

 
19.5
19
 






46700
93400




The numbers represent on source integration times (in seconds) for a source of given broadband magnitude for a S/N of 3. Total times include equal exposure on sky plus atmospheric (and flux, if required) calibrator(s). Therefore, to arrive at a total time one should multiply the above values by a factor which is no less than 2.3 (larger values would be used for larger telescope offsets required to attain blank sky positions, sources for which offset ROGUE tracking may be time consuming, etc.) The numbers in the above Table refer to 1 channel (at R = 1000) at a nominal wavelength of 1.64 µm (H) or 2.15 µm (K).
 
 

Line Flux in Extended Objects
Flux (erg/s/cm2/()2)
0.25pixels
0.4 pixels
H
K
H
K
3 x 10-15
160
80
60
30
1 x 10-15
1400
700
550
275
3 x 10-16
16000
8000
6200
3100
1 x 10-16

70400
55000
27500

The numbers represent on source integration times required to achieve a S/N of 3 in 1 channel (at R = 1000) at a nominal wavelength of 1.64 µm (H) or 2.15 µm (K). Total times include equal exposure on sky plus atmospheric (and flux, if required) calibrator(s). Therefore, to arrive at a total time one should multiply the above values by a factor which is no less than 2.3 (larger values would be used for larger telescope offsets required to attain blank sky positions, sources for which offset ROGUE tracking may be time consuming, etc.)
 
 

Integration Times

The minimum integration time is 0.3 seconds, while the maximum integration is set by the extent to which sky emission lines can be removed. This clearly depends on sky conditions but is typically of order 60 seconds at H-band and 100 seconds at K-band.

ROGUE

3D incorporates a fast tip-tilt image motion corrector, ROGUE (Rapid Off-axis GUider Experiment). This system employs a dichroic to separate the visible light from the NIR, and uses 4 avalanche photo-diodes (APD's) to track a bright compact object in the field. This object can either be a star or the program source itself, depending on the source characteristics. ROGUE is capable of operating at 80 Hz for tracking sources as faint as V=17. The field-of-view for ROGUE is ±1.6' (at the AAT) from the program object, and the movable pick-off mirror allows the observer to use any sufficiently bright, compact source in that field.  In addition, ROGUE incorporates a feature which allows observers to choose between two pixel scales. At the AAT these two scales are the nominal 0.4" per pixel, and the finer 0.25" per pixel.
 

 This page was last modified on 10 November 1998.


 
MPE-3D Homepage
References and Links
Data Reduction
SPECIAL NOTES FOR SPECIFIC
RUNS
Public FTP-Site
Recent Results from 3D