Years ago, I realized the
limitations of simply measuring light intensity within an
aquarium and longed for a way to measure spectral quality. Then,
the cost of a spectroradiometer equaled that of purchasing a new
luxury sedan. Times have certainly changed. If you have deep
pockets and a true desire to tinker with and know practically
everything about your lighting system and then some, then the
relatively low-cost spectrometers and accessories from Ocean
Optics just might be for you.

One may wonder why one would want to invest about a thousand
dollars in a ‘big boy’s toy.’ After all,
dedicated hobbyists, such as Sanjay Joshi, have analyzed the
spectrum of hundreds of lamps and have freely distributed this
information. Why invest in a spectrometer? While monitoring
spectral quality of lamps is likely the primary goal of most
hobbyists, the Ocean Optics spectrometers are capable of doing
much more (provided necessary accessories and additional software
– all at extra cost – are available). Of these,
reflectance, transmittance, absorbance, color analysis, relative
irradiance and absolute irradiance are of possible interest.

This article will relay some of my personal experiences with
two Ocean Optics spectrometers in my lab (a USB-2000FL – a
model designed to measure low level fluorescence and a USB-2000
capable of measuring ultraviolet radiation and visible radiation
from 200 nm to 850 nm) as well as touch upon some exciting new
introductions. The USB-2000 has just recently been replaced by a
newer model – the USB-4000 – however, the
observations of the older models are generally valid for the 4000
series, as well as the new, low-cost ‘Red Tide’
USB650 spectrometer.

The following article will also describe various options
available from Ocean Optics; however, it is not a substitute for
consultation with an Applications Scientist. The few minutes
required for a phone call to Ocean Optics is time well spent, and
will ensure that your spectrometer is correct for your
application. There are some things you’ll need to know
before calling, and this article can assist you with that.

Spectrometers

Ocean Optics revolutionized the spectrometer field when it
introduced the diffraction grating-based spectrometer using a
charged coupled device (CCD) for light collection. The early
models were just a fraction of the cost of spectrometers then on
the market. Combined with fiber optic technology, the Ocean
Optics equipment was, and continues to be, a powerful and
relatively inexpensive research tool. The new Red Tide
spectrometer is $999, while the new USB-4000 starts at $2,199.
The required SpectraSuite software is available for an additional
$199. See a partial list of specifications in Table 1.

The spectrometers are very compact (about the size of a deck
of playing cards) and light weight. These specs utilize USB
technology, making then ideal for field use (in fact, many
researchers have their spectrometers in underwater housings for
work with corals in their natural habitats). All OO spectrometers
are the crossed Czerny-Turner design.

Table 1. A partial listing of spectrometer specifications.

	USB2000	USB4000	Red Tide USB650
Optical Bench	Asymmetrical crossed Czerny-Turner
Detector	Sony ILX511	Toshiba TCD1304AP	Linear silicone CCD
CCD Pixels	2,048	3,648	650
Grating	Choice of 14 holographic gratings		Preconfigured for 350-1000 nm
Entrance Slit (microns)	Choice of 5,10,25,50,100,200w X1000h, or none		Preconfigured – 25µ slit
Input Fiber Connector	SMA 905	SMA 905	SMA 905
Focal Length	42 mm (input), 68 mm (output)
Integration Time	3 µsec – 65 sec	10 µsec – 65 sec	3ms to 65 sec
Effective Range	Depends on slit and grating, ranges 200 -1000nm		350 – 1000 nm
Optical Resolution (FWHM)	Ranges from 0.3-10.0 nm, depending upon grating		~2 nm
Signal to Noise	250 to 1	300 to 1	250 to 1
Power Source	USB	USB*	USB
Software	N/A	SpectraSuite	SpectraSuite
Spectrometer Price	No longer available	$2,199	$999
SpectraSuite Software	N/A	$199	$199
Operating System	N/A	Windows 98/Me/2000/Xp, Mac OSX & Linux w/USB port
* Use of serial port requires 5v power supply at additional cost

As mentioned, these spectrometers utilize diffraction gratings
to split incoming light into its spectral components. This
diffracted light falls upon the CCD array and specialized
software analyzes and reports spectral characteristics. No single
diffraction grating is efficient for broadband analyses (200nm
-1100 nm, except in the case of the HC-1 grating available in the
CG ‘composite grating’ series spectrometers, starting
at $3,999), so Ocean Optics offers a choice of 14 gratings for
the USB4000 (the inexpensive USB650 is preconfigured for
analyzing wavelengths between 350 and 850 nm). Analyses of UV
(down to ~220nm) and visible radiation will require a
specific grating, while work with visible wavelengths and
near-infrared will call for a different grating. At a minimum,
one should decide what part of the spectrum is of interest for
analysis. Even with the proper grating, the reported spectral
quality is not 100% correct. A particular grating is efficient
over a given spectral range (Ocean Optics uses 30% efficiency as
the cutoff). This means these spectrometers generally
under-report some wavelengths. The #3 diffraction grating is
probably the best option for most. It is most efficient at 500nm
and under-reports other wavelengths, particularly blue and red
wavelengths. There are various remedies, but perhaps the best is
free spectral correction software from Dr. Charles Mazel at
www.nightsea.com. Charlie
hopes to have this posted on his website ‘soon’ and
will include it along with his new products for viewing
fluorescence in aquaria. Ideally, one would use a reflectance
standard along with the software to correct the
spectrometer’s measurements, but it is not absolutely
necessary. The data is easily exported to a spreadsheet program
for further analyses (though this will require some time and
effort invested in writing a few lines of code).

Another important consideration is of expected light
intensity. Six variously sized apertures (called
‘slits’) are available for the USB4000 (5, 10, 25,
50, 100, 200 µ and ‘none’- the fiber optic size
acts as the light regulating mechanism). Low light applications
(such as fluorescence measurements) require larger slits (say,
200µ), while higher light intensity requires a smaller
slit. The Applications Scientist will walk you through these
choices. The question of suitability of a particular setup over a
broad range of light intensities is obvious – can a
spectrometer designed for low-light work in measuring spectra of
‘bright’ metal halide lamps? The short answer is
‘yes’ – but discuss this with the Ocean Optics
Applications Scientist (a few ‘tricks of the trade’
are listed in the Helpful Hints section, below). Optical
Resolution is a function of slit width and holographic grating.
An Optical Resolution of just a fraction of a nanometer is
possible. However, the Optical Resolution will be generally
between 2 and 10nm.

These specs are compatible with USB 1.1 and USB 2.0; Use of a
serial port is possible with the available 5v power supply.

Upgrades

Particular applications may require an upgrade to your
spectrometer which is usually at additional charge. For instance,
work with shortwave ultraviolet radiation will require special
UV-transmitting optics (optional UV2 quartz window for work with
200 – 340nm) and fibers resistant to UV damage
(‘solarization’). For low-light work (such as
fluorescence) consider the L2 lens upgrade.

Fiber Optics

It is not absolutely necessary to use fiber optic patch cords
if the goal is simply to measure lamp spectra – one can
merely remove the protective cover from the input aperture and
point the spectrometer at the light source. However, fiber optic
cords do offer advantages in that they can be tightly attached to
the aperture (via SMA 905 terminated fiber) thus protecting the
internal works of the spectrometer (the thought of a drop of
saltwater or debris entering the spec housing is frightening).
The fiber optic cords also offer other advantages. While it is
apparent that cords are a must for connecting optional
accessories, it should be noted that the size (diameter) of the
cord is also a critical consideration. Very simply, the larger
the diameter of the fiber optic cord, the more light transmitted.
This is useful to know if high light intensity
‘saturates’ the CCD array and causes the reported
intensity to be above the maximum allowed – Use of a smaller
diameter cord could attenuate (‘weaken’) the signal,
thus allowing measurement. Fibers are available in the following
diameters (in µ): 8 (VIS/NIR only, range of 450 –
1000nm), 50, 100, 200, 300, (for use with UV <250nm) 400, 600
and 1000.

Fibers are usually 2m in length, and custom lengths are
available. Fibers terminated with SMA 905 connectors can be
spliced with bushings (available from OO).

Software and Applications

SpectraSuite: Ocean Optics has upgraded its
older OOIBase32 software to the new SpectraSuite software.
SpectraSuite is Java-based software that operates with Windows
98/Me/2000/XP, Mac OSX and Linux w/USB port. Light measurements
are usually reported as ‘counts’ (a generic term) but
it is possible to measure absolute irradiance. See
comments on OOIIrrad software below. Graphical data can be
copied, and numerical data exported to spreadsheets such as
Excel. Integration time is programmable, as are functions such as
averaging and boxcar smoothing. Two spectral charts can be open
at one time thus allowing simultaneous use of two spectrometers
(especially useful when observed fluorescence excitation and
emission wavelengths). Languages include English, German,
Japanese, Russian, Chinese, French and Spanish. Lists for
$199.

OOISensors: Software specifically for use with
fluorescent pH and dissolved oxygen probes. Available for
$199.

OOIIrrad:Software for measure of relative and
absolute irradiance. OOIIrrad also requires the purchase of a
LS-1-Cal Calibrated NIST-traceable Light Source for an additional
$749. OOIIrrad software sells for $299.

OOIIrrad-C: In addition to the functions of
OOIIrad software (above), this package also analyzes reflective
and emissive color. $399.

Note- Ocean Optics is changing their software to for
compatibility with various operating systems (Windows, Mac, and
Linux). The prices listed are current as of early May 2006.

Accessories

Ocean Optics offers a wide range of assorted accessories for
assorted applications.

Specialized sensors utilizing fluorescence are available for
monitoring of pH (TP300 pH probe, available for $750) and
dissolved oxygen (‘FOXY’ probes, starting at $499).
OOISensor software is also required (see above). These probes are
of modest diameters and can test pH or dissolved oxygen levels
with the interstitial spaces of ‘bushy’ corals.
We’ll likely see important information generated with these
sensors in print soon. (Jake Adams experimented with corals and
an oxygen sensor and made a very interesting presentation at the
2006 IMAC.)

A variable neutral density filter is available for light
attenuation (but see ‘Helpful Hints’ below). The
neutral density filter is $499. If one is needed, consider
purchasing the UV or Vis Fiber Optic kits. They are $999 each
(specify ‘ultraviolet’ or ‘visible’) and
include one variable neutral density filter, four 74-series
optical collimating lenses, 4 fiber optic patch cords of various
diameters, CC-3 cosine corrector, bulkhead and splicer fittings,
a SEM wrench and some odds and ends. If purchased individually,
the assortment of items in the Fiber Optic Kits is ~ $1,500.

Reflectance standards (diffuse and specular) are available and
should be used for reflectance determinations.

Various excitation and calibration light sources are
available, including deuterium, tungsten-halogen and
light-emitting diodes (LEDs) of various ‘colors.’

Scratch and Dent Surplus Equipment

Ocean Optics offers trade show items for sale at bargain
prices (I got a ‘like new’ Spectralon 99% diffuse
reflection reference for $199 – it normally sells for
$329). Check
http://www.oceanoptics.com/PriceList/Scratch&Dent_FiberPriceList.pdf
for available inventory.

Helpful Hints

Ocean Optics suggests using this formula for determining
approximate optical resolution:

Optical Resolution = Dispersion X pixels in slit or
fiber, where:

Dispersion = Spectral Range of the
Grating/Number of Pixels in CCD array.

Resolution – Use Table 2. (Results are in
pixels, typical value of slit width or fiber diameter).

Table 2. Optical Resolutions

Slit Width / Fiber Diameter (µ)	Pixels
5	3.0
10	3.2
25	4.2
50	6.5
100	12.0
200	24.0

If spectral analysis of lamp(s) is your only concern, then
perhaps observations of percent spectral change would be of most
use. Spreadsheets, such as Excel, allow for relatively quick and
easy manipulation of exported data.

One of the most common challenges facing users of these
spectrometers is one of excess light. Attenuation is possible by
obvious means (such as moving the spectrometer or the fiber optic
cord away from the light source). Reducing integration time is
also useful as is use of a smaller fiber optic cord. Fiberglass
window screen material, acting as a neutral density filter can
also be replaced over the aperture or fiber cord terminal. Light
intensity can also be reduced by applying a layer (or two) of
transparent tape over the end of the patch cord or over the light
entrance aperture.

Attach the spectrometer to a small piece of plywood with nylon
cable ties. This will protect it from those inevitable bumps (See
Figure 1).

General Impressions

I learned more than I had ever known about lamp spectra
(especially in the UV range) in a few minutes’ use of my
first OO spectrometer. Their relatively low cost instruments have
made possible many experiments concerning lighting and its
effects on zooxanthellae photosynthesis. These
spectrometers’ small size, light weight and USB power have
made field research easy. The Ocean Optics equipment is
relatively inexpensive, but by no means cheap. Workmanship and
overall quality of the hardware is first class.

There are a few areas in which products could be improved. The
new software, SpectraSuite, is not as user friendly as the older
OOI32Base software. The ‘Help’ section is on a CD
(which is of course the rule today). Perhaps I’m just
old-fashioned, but I much prefer a printed User’s Manual
straight out of the box as opposed to printing a hardcopy
myself.

Fully read and understand the directions for connecting the
spectrometer to your computer. Resist the urge to plug the USB
and spectrometer in! Read the directions! Plan to devote an hour
to installing the software and configuring the spec.

Still, the equipment is not for everyone. Personally, I would
recommend the purchase of a quantum or PAR meter before
purchasing a spectrometer – a PAR meter will be of much
more use. However, if one has a PAR meter, then spectral analyses
are the next frontier. Hobbyists should honestly evaluate the
return on investment as well as the amount of time one has to
dedicate to instrument use and experiment set up. Sanjay Joshi
did (and is doing) a wonderful service to the hobby by providing
spectral data for hundreds of lamps (see archived articles in
Advanced Aquarist, or visit Sanjay’s website:
http://www.reeflightinginfo.arvixe.com).

On the other hand, Ocean Optics has earned the reputation of
quality equipment at reasonable prices and any really serious
hobbyist or small budget-minded laboratory would benefit from the
spectrometer’s flexibilities. There are many unanswered
questions about artificial light quality and its effects on
zooxanthellae photosynthesis and coral host pigmentation.
Especially with the introduction of the ‘Red Tide’
USB650 spectrometer and software package costing a total of only
$1,200, Ocean Optics equipment will be an important key in our
quest of unlocking mysteries and unknowns the hobby still
faces.

See
http://www.oceanoptics.com
for an online catalog, contact information and more technical
data.

I’m in the process of writing an article or two around
the practical applications using Ocean Optics equipment, how to
interpret the results and what they tell us.