A Comparative Review of some
Commercial Spectrographs
|
|
This short article is intended to assist the newcomer to amateur astronomical spectroscopy who is contemplating the purchase of a ready-made commercial spectrometer or spectrograph. It provides some guidelines and advice with a critical review of some spectroscopic equipment that can be found on the commercial market and that may be adapted and modified for use as astronomical spectrographs. The information provided is in no way meant to promote or recommend one manufacturers equipment relative to another; it is simply based on my personal experience working with some of the commercially available instrumentation in my professional work. Most of the equipment that I shall be describing is not advertised in magazines that the typical amateur astronomer reads, being aimed principally at the industrial analytical chemistry market. This is a highly competitive multi-billion dollar business with numerous suppliers, both large (as in immense!) and small (as in start-up), and examples of the
instruments that I shall describe cover both these extremes. For the sake of space and time, only a limited number of devices are described, and include only those likely to have any potential applications to amateur astronomical spectroscopy, and to have any chance of falling within the budget of amateurs.
It is hoped this information will guide the complete beginner to astro-spectroscopy, prevent he or she from making the wrong choices, and encourage them to look further. I also hope that F.A.A.S. members will also find some things of interest. A list of Web sites and other useful references is provided at the end, where much more information such as company product specifications, data sheets, colour brochures, demo disks, etc...can be requested.
There are essentially two ways of "connecting" a spectrograph to a telescope in order to perform astronomical spectroscopy:- either by employing an optical fibre feed to guide as much light as possible to the spectrometer, or by direct coupling of the spectrograph at the focal plane of the telescope. Both methods exhibit their respective advantages and drawbacks, and both are mentioned below.
|
The Ocean Optics Product Range
|
|
10 years ago Ocean Optics Incorporated
(OOI) was a tiny start-up company based
in Florida, USA, formed by two state
university researchers working on the
development of fibre optic pH sensors
for oceanographic applications. The
work led to the introduction of a
palm-sized fibre-optic spectrometer that
was orders of magnitude smaller and
cheaper than more conventional
laboratory bench systems. It is now ten
years on and the company has won several
industry awards for innovation and has
significantly increased its product
range. Two "typical" optical fibre
spectrometers that are marketed by OOI
are the PC1000/2000 series and the S2000
series. The former has the spectrometer
physically installed directly onto a
small format ISA interface card that
slots into the PC. The latter is a stand
alone model that can either connect to a
desktop PC or to a laptop computer via a
PCMCIA card. Both models are pictured
below.
|
Figure 1. The Ocean Optics S2000 Fibre-Optic Spectrometer and NI DAQCard 700 Interface Card
|
Figure 2. The Ocean Optics PC1000 Spectrometer on a card!
|
|
The PC1000 spectrometer (Figure 2) is basically similar to the S2000 but is a first generation device having 1024 pixel elements. Chronologically the PC1000 preceded the S2000. OOI have now introduced a 2nd generation PC series instrument, the PC2000, with a 2048 pixel count. The software for data acquisition and processing was initially DOS based, but quickly moved up to the Windows™ platform with an application called OOIBase. The latest version is V1.5. A much improved 32 bit package was introduced this year called OOIBase32 with vastly more data handling functions. A demo disk is available from the OOI Web site; it will give you some idea of capability, but is not much use without the spectrometer!
The prices for these instruments start around $1,500 to $2,000 for the spectrometer, interface card or PCMCIA card and a couple of fibres. A recent "tag sale" that the company advertised had some of their earlier models selling for under $1,000. A large range of accessories is available such as UV/Vis cuvette holders, reflection probes, immersion probes, even a spectrometer for Raman studies, none of which is particularly useful for astronomy. What must be appreciated is that the target market for these devices lies in the analytical sciences and environmental testing industries. The tiny footprint of these instruments, their low weight, and their exceptional portability lend themselves ideally to these businesses. So how about the spectrograms?
Typical Spectra
Some typical spectra obtained with these instruments can be seen below in Figures 3 and 4. The quality is surprisingly good, with excellent noise rejection and acceptable resolution (grating dependent) at the very short exposure times employed. However, we are dealing here with spectrally bright sources.
|
Figure 3. Hg(Ar) calibration lamp, integration time 10 ms; zero averaging, raw data.
|
Figure 4. Spectrum taken through a 12in. LX200, 50 micron slit, 100 ms integration, zero averaging, raw dat. (S2000 spectrometer)
|
|
Fig. 4 shows the spectrum of a high pressure sodium arc lamp about 1 km distant from the telescope. It was taken from the deck of my back garden in Massachusetts. A 50mm fibre was fed between the telescope and the S2000. Integration time was 100msec. Strong self-absorption of Na atoms due to pressure broadening is readily apparent in the doublet emission line due to the high pressures at which these lamps are operated. (This is all very well, I hear you say, but what about applications to astronomy where the ambient light is much lower?)
|
Potential for Astro-Spectroscopy?
|
|
Although attractive from the point of
view of size and weight, the potential
for astronomical spectroscopy with these
spectrometers is unfortunately limited.
The image below, recorded recently,
shows the spectrum of Jupiter which,
other than the Sun and Moon, was the
brightest object in the evening skies
during October/November of 1999.
|
Figure 5. Jovian spectrum from 400-700nm. Integration time 2000 ms, boxcar smoothing, zero averaged.
|
|
The spectrum of Jupiter identifies some of the principal solar Fraunhofer lines, but the data is beginning to require smoothing After the planets, I attempted some of the brighter stars. Unfortunately, the resulting spectra are essentially too noisy to be worth showing here. Vega and Altair, for example, revealed little or no features beyond time-integrated noise.
|
|
OOI equipment is not ideally suited to astronomical spectroscopy, which is no criticism of the company, of course, since astronomy is not their target market. At best these spectrometers are limited to the Sun and the brighter planets. The advantage of these instruments is their small size (too small in fact when one considers the tiny focal lengths that are being used) and their low mass. Optical fibre coupling is a great convenience but comes with a considerable reduction in light throughput, up to ca. 50-60% in these examples. As always with optical fibres great care also has to be taken with f/number matching, which is why OOI factory-install collimator lenses tailored to their spectrometers and fibre range. But the greatest drawback here, I feel, is that the CCD chips are un-cooled and thus S/N suffers considerably with the prolonged integration times necessary for astronomy. Some natural cooling of the chip may be experienced outdoors in winter which could improve matters, but will
not be significant and will certainly not compare to the controlled cooling obtained from multi-staged Peltier coolers available in other equipment.
|
|
Located in Putnam, CT, USA, CVI Spectral Instruments is part of CVI Laser Corporation, and market a series of portable and lab-bench spectrophotometers of various focal lengths from several centimetres to ½ metre. The SM-210 (Fig. 6) appears to be a direct competitor to the Ocean Optics devices. Choice of gratings, portability, attainable resolution and photo-sensitivity are all comparable.
|
Figure 6. CVI's Hand-held Spectrophotometer SM-210
|
|
In the same way, CVI also provide a fairly complete range of filters, gratings, wavelength calibration sources and bundled or single-strand optical fibres to use. The SM-210 is of a similar configuration to the OOI S2000 in that it is a stand alone instrument but requires an ISA interface card (provided in the package).
Typical spectra achievable with the SM-210 are closely similar in sensitivity and resolving power to the OOI S2000 and PC1000 spectrometers. Figure 7 is a spectrum of a quartz halide lamp about 15 min. after start-up with the fibre held close to the light source. A mix of Hg line and band emission can easily be seen. (The presence of lines and bands in a heavy metal source spectrum such as this can be readily observed visually by using a direct reading spectroscope such as Edmund Scientific’s prism train unit, [product No. 42586]).
|
Figure 7. Spectrum of quartz-halide lamp obtained with the SM-210 spectrophotometer (re-imaged from my observer's notebook).
|
|
Figure 8 is an image of the solar spectrum (reflected full Moonlight) that I made about a year ago with the SM-210 using a Celestron C90 spotting scope and with a 200mm fibre hand-held against the eyepiece. Results are fairly low resolution, and not as good as in Fig. 5 that was taken with a Meade12" LX200.
|
Figure 8. The solar spectrum taken with the SM-210 (notebook image).
|
|
CVI also market a range of larger bench-top spectrographs and monochromators for UV and visible spectroscopy, such as the Digikrom ¼ metre units, the CM110 1/8 metre spectrographs and monochromators, the SM-100 and the SpectraMatch GT-2 fibre-based spectrophotometer. The Digikrom CCD spectrograph systems employ an OEM version of the SBIG ST-6 camera, and run Rhea Corporation Kestrelspec image acquisition software. A demo version of this software was available the last time I looked at Rhea’s Web site. My experience with these other CVI instruments is much more limited. They are significantly more expensive than the hand-held unit, starting at several thousand dollars. The SM-210 can be purchased for $1,000-$1,500, depending on accessories. As with the OOI range, the SM-210 comes with Windows based software for data acquisition and manipulation but, as is the case with the Ocean Optics products, the potential for low light level astro-spectroscopy is severely restricted. For much more
information on their products you can contact CVI Laser Corporation Development Group at http://www.neca.com/~cvilaser.
|
|
We now move up a step in size and focal length and consider some instruments with more potential for astronomical spectroscopy. ISA (Instruments S.A. - Instruments Société Anonyme) is the company that incorporates Spex Instruments and Jobin-Yvon. The parent company is located in Longjumeau, in the southern Paris suburbs, but was recently acquired by Horiba Instruments, a large Japanese corporation specialising in particle sizing instruments and other analytical equipment. Both Spex and Jobin-Yvon are well known in the chemistry/physics community as manufacturers of high quality reflection gratings, monochromators and spectrographs. ISA offers a complete range of UV/Visible spectroscopic equipment some of which can be directly employed for astronomical spectroscopy. If you can get hold of a copy, request their "Guide to Spectroscopy" if it is still in print. Although understandably concentrating in large part on their own product range, this little book is also a mine of
information on the theory and practice of light sources, monochromators, aperture matching theory for fibre optics, detectors (CCD and non-CCD) and spectrographs. Product specifications on ISA spectrographs and monochromators include small 0.2 metre focal length units and increase in size up to 1.5 metre focal length "behemoths" that weigh over 100kg! Also not to miss is an excellent tutorial article called "The Optics of Spectroscopy" written by J.M. Lerner and A. Thevenon of the Optical Systems Division at ISA. This article contains all the relationships and equations you need to know for those of you who may be thinking of designing your own instruments. The last time I looked this tutorial was still downloadable from the ISA Web site (http://www. Isainc.com/systems/theory/oos/oos.htm).
|
Potential for Astro-spectroscopy?
|
|
Probably the best suited unit for amateur astro-spectroscopy from ISA is the CP200 (Fig.9), a solidly built imaging spectrograph. Its weight of over 2kg will restrict its use to fiber optic coupling for all but the most massive amateur telescopes and mountings, but I know that CP200 units are in use in astronomy departments at several universities and colleges in the New England area. A choice of interchangeable gratings is offered, from a groove density of 65 lpm (dispersion 76 nm/mm) to 360 lpm (dispersion 16nm/mm).
|
Figure 9. Fact sheet on the ISA CP200
|
Reverse side.
Contact ISA directly for more information on these
compact well-made spectrographs. In the USA their
office is located in Edison, NJ at 1-800-438-7739
toll-free, but their Web site should be your first
port of call.
|
|
Oriel Instruments are a well known company in the spectroscopy field based in Stratford, CT, USA. They have selling companies worldwide and compete directly with ISA and other manufacturers. The company markets a vast array of spectroscopic and radiometric instrumentation ranging from lasers and fibre optics to equipment for photolithography and modular Fourier Transform IR spectrometers. Their series of spectrographs and monochromators is fairly extensive, and a full range of supporting accessories likely to interest the amateur astronomer is available. Two models that I shall briefly consider with which I have some practical experience are the MS-125 spectrograph and the FICS (Fixed Imaging Compact spectrograph).
|
Figure 10. The Oriel MS-125 1/8m spectrograph.
|
Figure 11. Oriel's FICS(TM) Spectrograph
|
|
Figure 10 shows the MS125 connected to an Oriel Instaspec IV™ linear CCD detector (more about the CCD later!). The spectrograph configuration is a crossed Czerny-Tuner bench which is very popular these days permitting a compact design. As the ‘125’ designation suggests, the focal length is 1/8 metre. The body of the spectrograph measures roughly 6 inches by about 5.5 inches in a distorted hexagonal shape. The figure clearly shows the slit entrance compartment (red cover), with an interchangeable slit inserted to the left. The micrometer is used for wavelength range adjustment and grating change-overs are simple and reproducible. Available gratings (both ruled and holographically etched) range from high density 2400lpm (80nm band-pass measured at the blaze wavelength) to a low density 300lpm (675nm band-pass). Typical spectral resolution ranges from about 1Å to 11Å (measured with a10µm slit and 14 x 200µm pixel array of 2048 pixels). Figure 12 shows the insides of the MS125, showing it to
be well baffled for minimum stray light and the rejection of higher diffraction orders.
|
Figure 12. Details of the MS-125(TM)
|
|
The cost of the MS125 is between $1,500 and $2,000 depending on the accessories that are purchased. Additional gratings and slits can add significantly to the total price. Nevertheless, the MS125 may be a competitive alternative to the recently introduced SBIG spectrometer. Where the cost takes a big leap, however, is with the addition of the detector. The CCD detector shown coupled to the spectrograph is an Oriel InstaSpec™ MkIV unit that accommodates linear CCD arrays. [N.b. Photodiode arrays are also common in analytical laboratories with HPLC equipment and UV/Vis spectrophotometers. Do NOT consider them. They are NOT suitable for astronomy being much less sensitive than the CCD]. The detector that I use has a Hamamatsu linear array of 64 x 2048 pixels, specifically designed for spectroscopy. Pixel area is 24µm2. The chip is always cooled with a single stage Peltier down to a minimum temperature of -28C. Water/ethylene glycol circulation can be use to provide additional cooling if
required. Performance testing has measured total readout noise at 16.6 electrons RMS with a system gain of about 10 photoelectrons per count.
A close-up of the detector head can be seen in Figure 13.
|
Figure 13. The instaSpecIV head showing the linear CCD sensor. The O-ring gasket forms a light-tight seal up against the mounting flange
of the spectrograph. The CCD compartment has been flushed with dry argon to prevent moisture condensation. The two connectors at the top of the image are for water cooling.
|
|
Computer interfacing and data acquisition is provided by a full size16 bit ISA card and comprehensive software. The InstaSpec IV software on my unit is DOS based (yes, the old "dinosaur" is still around!) but it provides all the necessary functions and also includes routines such as 3D kinetics data collection, internal/external synchronization for the capturing of spectra of transient chemical species, etc. More recent models of the InstaSpec detectors are now Windows™ based. The spectrograph/CCD ensemble is in fact an "imaging" system in that the vertical column height of the chip has 64 pixels. Larger heights up to 256 pixels are available, with the same 2048 pixel width, but they are much more expensive! The imaging capability is usually exploited in the chemistry laboratory by using this "multi-track" imaging mode and employing multiple fibre-optic feeds into the spectrograph so that the slit actually samples different physical areas of an extended object
(e.g. for flame and plasma studies), and records the spectra of each vertically distinct slice. None of these capabilities are of much use for time-exposed astronomical spectroscopy, but as is usually the case the primary market for this instrumentation is the mass chemical spectroscopy community. The software, however, allows for vertical binning of the 64 pixels in LIS (Linear Image Sensor) mode for very fast readout and increased sensitivity. This is the usual mode that I use for astro-spectroscopy.
|
|
Steve Dearden is a British chemist who specialises in analytical chemistry and polymers, and is currently living in Massachusetts, USA. He has had a long standing amateur interest in the applications of spectroscopy to astronomy over the last 15-20 years.
He is presently "between jobs" (which is how he managed to complete this article in reasonable time!), so if you know of anything available in the chemistry/spectroscopy field he'd be pleased to send you his résumé/CV.