How to choose a spectrometer

 

The basic parameters you need to know before choosing a suitable spectrometer are:
  • Wavelength range
  • Resolution

For instance, if you need to analyze color you need a spectrometer that covers the visible spectrum from approx. 400 – 700 nm.

The resolution is the spectrometer’s ability to distinguish between two closely spaced wavelengths. So, if your spectrum has some sharp peaks that are separated by say 0.5 nm or more, you need a spectrometer with at least 0.5 nm resolution.

Some applications (like Raman and IR spectroscopy) will not list the wavelength range and resolution in nanometer (nm) directly but rather in terms of inverse cm (cm-1). For Raman you can use our RamanShift converter to convert back and forth between nm and cm-1.

Even though you know your wavelength range and resolution there are still a lot of possible spectrometer design options. So your final choice of spectrometer will depend on the importance of parameters such as overall Size, Cost, Speed, Sensitivity, Signal-to-Noise Ratio, Dynamic range, Linearity, Thermal Stability, Robustness. Below, you can find some general guidelines that may help you determine what kind of spectrometer parameters you should focus on for you application.

  • Size
    If you need a compact spectrometer you should in general go for:

    • High grating dispersion
    • Small detector size
    • Low numerical aperture/high f-number
  • Cost
    If you need a low cost spectrometer you should in general go for:
    • Small size spectrometer
    • Low-end, non-cooled CCD/CMOS-based detector or a Photo-Diode array
    • Low resolution spectrometer
  • Speed
    If you need a fast spectrometer you should in general go for:

    • High through-put spectrometers
    • Transmission grating based
      High NA / Low f-number
      Few optical elements (preferably lenses)

    • High-speed, non-cooled cameras
  • Sensitivity
    If you have very little light from your sample you need a highly sensitive spectrometer and you should in general go for:

    • High throughput spectrometers

    Transmission grating based
    High NA / Low f-number
    Few optical elements (preferably lenses)
    A wide and tall input slit

    • Cooled camera/detector that enable long integration time
    • Highly sensitive CCD (or CMOS) detector with tall pixels
  • Signal-to-Noise Ratio
    If signal levels have to be very stable over time you have several options depending on the actual signal level and balance of cost versus performance:

    If your signal level is very weak and tends to drown in the noise:

    • High through-put spectrometers

    Transmission grating based
    High NA / Low f-number
    Few optical elements (preferably lenses)
    A wide and tall input slit

    • Cooled CCD/CMOS detectors with tall pixels that allow long integration time with low noise

    If your signal level is very weak but, you also need a low cost solution:

    • High through-put spectrometers

    Transmission grating based
    High NA / Low f-number
    Few optical elements (preferably lenses)

    • Non-cooled CCD/CMOS detectors where you average over many short integration periods

    If your signal level is stronger:

    • High through-put spectrometers
    • NMOS detectors with very low noise
  • Dynamic range
    If there are large (several orders of magnitude) variations in your signal levels you should in general go for:

    • Low stray-light spectrometers

    Holographic, Master gratings rather than replicated, ruled gratings
    As few optical surfaces as possible
    Bandpass filters to block un-wanted light

    • NMOS detectors with very large dynamic range or Non-cooled CCD/CMOS detectors where you average over many short integration periods.
  • Linearity
    If you require a linear relation between signal level and integration time you should in general go for:

    • NMOS/BT-CCD detectors with very good linearity or CCD/CMOS detectors where you make a software correction for linearity.
  • Thermal stability
    If your spectrometer should operate under varying temperature conditions you should in general go for:

    • Spectrometers with low wavelength and power shift vs. temperature

    Lens-based, tele-centric rather than mirror-based spectrometers
    All-dielectric, transmission grating based rather than reflection grating based spectrometers
    Thermally stable mounting methods

    • Short integration time or temperature controlled detectors if you need longer integration time
  • Robustness
    If your spectrometer should operate under demanding environmental conditions you should consider some of the following:

    • Spectrometers with low wavelength and power shift vs. external impacts

    Lens-based, tele-centric rather than mirror-based spectrometers
    Transmission grating based rather than reflection grating based
    Environmentally qualified mounting methods for optics and gratings

    • Short integration time and/or temperature controlled detectors if you need longer integration time