Transcript Cvetojevic_PIMMS_11 - School of Physics

THE INTEGRATED PHOTONIC
SPECTROGRAPH
MULTIPLE OFF-AXIS INPUTS AND
TELESCOPE RESULTS
Nick Cvetojevic1,2, Nemanja Jovanovic1,2, Joss Bland-Hawthorn3, Roger Haynes4, Mick Withford5,
and Jon Lawrence1,2
1. Department of Physics and Astronomy, Macquarie University, NSW, 2109, Australia
2. Australian Astronomical Observatory, NSW, 2122, Australia
3. Sydney Institute for Astronomy, School of Physics, University of Sydney, NSW, 2006, Australia
4.innoFSPEC, Astrophysikalisches Institut Potsdam, Potsdam, 14482, Germany
5.CUDOS, Centre for Ultra-high Bandwidth Devices for Optical Systems, Australia
The Integrated Photonic Spectrograph
A complete “spectrograph on a chip” for
astronomy
Fully integrated photonic platform with
no moving parts, no alignment, high
stability
Mass-producible and small
Current-Generation Spectrographs
Existing spectrographs for astronomy are very large, full of custom built parts,
and very expensive
Component & Spectrograph
Size increases with
Telescope Diameter
Cost ~ Diameter2 !
Bland-Hawthorn & Horton
(2006)
http://www.astronomy.com/asy/default.aspx?c=a&id=2863
Non-Monolithic Designs
Why not use a multitude of smaller, cheaper, replaceable spectrographs to do
the same thing?
Ideal for fiber-fed multi-object
spectroscopy
VIRUS
Identical modules combine to form
one large spectrograph
25% of the cost of the monolithic
design
Still very large!
http://www.as.utexas.edu/hetdex/
The Spectrograph Chip
‣ Silica chip with an lithographically written Arrayed Waveguide Grating structure
‣ Typically used in Telecommunication Networks
The Spectrograph Chip
Arrayed Waveguide Grating
Array of Waveguides
Input
Free Propagation
Zone
Output
Free Propagation
Zone
Focal Surface
Input Fibre
6
The Photonic Lantern
Converts a Multimode fibre into multiple Single Mode fibres for efficient
interfacing with a telescope
1x MMF with N
modes
Photonic
Lantern
7
N x SMF
Simultaneous Multi-Fibre Input
By interfacing multiple SMFs to one chip we can increase its observational
efficiency and reduce the total amount of chips used
Arrayed Waveguide
Grating
Photonic Lantern
Multiple Off-Axis Fibre Launch
So what happens to the spectral output when inputting multiple fibres?
Waveguide Array
Free Prop. Zone
1550 nm
We would assume that if the fibres are offset enough for the FSR not to overlap
we could get separate spectra on the output
Top View
Front View
Blue
Fibre #1
Fibre #2
Fibre #3
Red
10
Unfortunately this is not the case!
This causes the spectra to be superimposed regardless of the fibre input
position
Top View
Front View
Blue
Fibre #1
Fibre #2
Fibre #3
Red
11
However, if we use a crossdisperser we can uncouple the
spectra from the different fibres
If cross-dispersed, we can simultaneously record the spectra from multiple
fibres. We can fit as many as the gap between the orders allows.
Front View
Higher Orders
Blue
Fibre #2
Blue
Fibre #1
Fibre #3
Fibre #1
Cross
Dispersed
Fibre #2
Fibre #3
Red
12-14 Fibres at 125 um spacing
Red
12
Red
Blue
The AAT
The Demonstrator Instrument
Lenslet Array
Multimode Fibres
Photonic Lantern
12x SMF
The initial IPS setup
• 3 different setups on one assembly.
• Designed to be interchanged on the night
The initial IPS setup
Laser @ 1550nm
The initial IPS setup
Setup #1 – Wide wavelength window, Medium resolution
• R ~5000, full H-Band, 12 SMF, 1 MMF
Setup #2 – Highest resolution, Small wavelength coverage.
• R ~7000, 50nm wide band, 14 SMF, 1 MMF
Setup #3 – 2 Chips on one detector, Anamorphic optics
• R ~2000, full H-Band, 24 SMF, 2 MMF
The Boss supervising
The IPS going on
the AAT
The initial IPS setup
Unfortunately, initial tests showed we were not getting enough light
through and approaching the noise floor of our detector.
Our detector was not sensitive enough
We decided to use IRIS2, and MacGyvered together a new interface
between the IPS and IRIS2
The Raw Results
Antares
Different Orders
1450 nm
1780 nm
Spectra from individual Fibres
The Raw Results
alf Ara (Be Star)
V* Pi 01 Gru (Cold red giant)
Conclusion
We have demonstrated simultaneous input of multiple single mode
fibres directly into an AWG chip is possible and practical for
Astronomy
If used, cross dispersion is all but essential
We have successfully demonstrated the first IPS-like device on a
telescope, with spectra taken from 3 different types of stars.
Currently, redesigning the AWG chips to improve FSR, R, Wavelength
Looking at using AO systems to directly couple into SMF
THANK YOU