UV spectroradiometers and optical spectrometers

UV spectroradiometers and optical spectrometers capture radiation resolved by wavelength. This makes it possible to assess the spectral distribution, peak position, bandwidth and spectral changes of a UV source. Absolute irradiance additionally requires a radiometric calibration of the complete measuring chain. The selection is governed above all by wavelength range, spectral resolution, dynamic range, stray-light behaviour and measuring geometry.

The portfolio comprises the portable UV spectroradiometers UVpad and UVpad E as well as the array spectrometer SR900 for research, development and quality assurance. The spellings spectroradiometer and the generic term spectral measuring instrument are also common – the same instrument is meant.

All three are absolutely calibrated; calibration takes place in our laboratory accredited to ISO 17025 and is traceable to the PTB. The corresponding quantities and units are set out under radiometric quantities, the range boundaries under spectral ranges and standards.

Both terms describe the same optical design but two different measuring tasks. The difference lies not in the optics but in the calibration and the entrance geometry: a spectrometer disperses light and outputs a relative curve. A spectroradiometer is an absolutely calibrated spectrometer – it delivers spectral irradiance in W/(m²·nm) and from it mW/cm², lux or PPFD.

Anyone who needs an irradiance or a dose needs a spectroradiometer; everything else measures relatively. Every instrument in this category is a spectroradiometer. For measuring transmission and UVT in the process – a photometric rather than a radiometric task – Opsytec offers the process photometer ZPM.

A spectrometer delivers counts – numbers that depend on the detector, the integration time, the entrance slit and the light guide. They say where the source emits, not how much. Only absolute calibration against a standard assigns a physical value to each wavelength and turns the curve into a measurement.

Three things are needed for that, and a plain spectrometer has none of them:

• A traceable calibration against a standard that itself goes back to a national institute – at Opsytec to the PTB. Without it the value is not comparable, not even with your own instrument from last year.
• A defined input optic. Irradiance is power per area, so the instrument has to collect a known area with cosine-correct weighting. A light guide without a diffuser does not do that.
• Stray light correction. In the UV-C the signal is often orders of magnitude weaker than in the visible. Without correction what you measure there is mostly stray light from inside the instrument.

A broadband radiometer measures faster, costs less and is more rugged in daily use; it does, however, weight the radiation with its own spectral responsivity and is correct only when the source resembles the calibration source. For continuous monitoring of a known installation it remains the right choice – that is what broadband radiometers for process and dose checks are for.

As soon as the source changes, several wavelengths have to be assessed separately or an action spectrum has to be applied, there is no way around the spectrum. The full side-by-side comparison of both designs, including their error sources, is shown under when a spectroradiometer is necessary.

First question: do you need an absolute value? If it is only about the position of the peak, the FWHM or comparing two batches, a relative measurement is enough – a spectrometer will do. As soon as a figure in mW/cm², lux or J/cm² has to come out, there is no way past absolute calibration.

Second question: which wavelength range? Pure UV to 440 nm, UV plus the start of the visible to 480 nm, or the full optical spectrum to 1,100 nm – that decides the instrument.

Third question: where is the measurement taken? No cable leaves a closed chamber; there the instrument has to measure and store autonomously. At the workplace or on a running line, a sensor head on a cable is right. On the optical bench what counts is range and resolution, not size.

The last row is not a contradiction but the most common answer in practice: measure spectrally once, then monitor broadband. The spectral measurement establishes how the broadband sensor deviates at this source; after that the fast instrument suffices for recurring checks – as long as the source stays the same. Change the lamp and the chain starts again.

A spectrometer splits light into its wavelengths and measures the intensity of each one. The entrance optic is irradiated first. A dispersive element – a prism or a grating – separates the wavelengths spatially; photodiodes or CCD arrays then detect them. The resulting spectrum allows conclusions about a lamp, an LED or a sample.

The most common design is the Czerny-Turner spectrometer: two mirrors and a grating, arranged so that optical aberrations stay small and spectral resolution stays high. The design is compact enough for mobile instruments and small laboratory benches.

Three designs have to be distinguished.

Array spectrometers use a row of detectors, usually CCD or CMOS, which capture the whole spectrum at once. That makes the measurement fast and needs no moving parts. All Opsytec spectroradiometers work this way.

Monochromators have a slit and a rotating grating and pass one wavelength at a time. They measure very precisely and with high spectral purity, but they need time – one wavelength after the other. Typical for research and quality control where wavelength selection is what matters.

Double monochromators place two monochromators in series. That lowers stray light further and is needed wherever very weak signals sit next to strong ones – in the UV range and in high-accuracy laboratory metrology.

A single spectrum yields several quantities:

• Illuminance – the photometrically weighted light intensity
• Correlated colour temperature (CCT) – describes how warm or cold the light appears
• Colour rendering indices – how accurately a source renders colours
• CIE colour coordinates – the colour locus in the standard colour space
• Spectral power distribution – how the energy is spread across the spectrum
• Peak and dominant wavelength – where the highest and the visually dominant energy sit

A broadband radiometer returns one single figure instead.

What the individual quantities mean is explained under colorimetric quantities; symbol, unit and definition of every optical quantity are in the glossary of optical quantities.

The word spectrometer reaches far beyond light measurement. Three instrument types regularly appear in the same search and measure something entirely different:

Mass spectrometers separate ions by their mass-to-charge ratio. They identify the smallest quantities of a substance and are used in chemical analysis, biochemistry, environmental science and the pharmaceutical industry.

Atomic absorption spectrometers measure how vaporised metal atoms absorb light and determine metal concentrations in solution from it – environmental analysis, toxicology and materials science, above all for heavy metals.

Spectrophotometers, often called UV-Vis spectrometers, measure the absorption or transmission of a sample and derive concentrations from it. Widespread in clinical diagnostics, water analysis and food monitoring. For transmission and UVT in the process, Opsytec offers the ZPM process photometer.

A spectroradiometer, by contrast, measures spectral irradiance against an absolute calibration. Its input optic is a diffuser, usually connected through a light guide – as in the SR900 spectroradiometer.

What distinguishes a spectrometer from a spectroradiometer?
A spectrometer delivers relative values; a spectroradiometer is absolutely calibrated and delivers spectral irradiance. Every instrument in this category is a spectroradiometer.

Which wavelength range and resolution do I need?
The range must cover the emission and any secondary bands in full – UV only up to 440 nm, UV plus the start of the visible up to 480 nm, or the full optical spectrum up to 1,100 nm. The resolution decides whether closely spaced peaks or narrow lines are still separated.

Which measurement errors occur with array spectrometers?
Stray light above all: parts of strong wavelengths appear where they do not belong. In the UV-C the signal is often orders of magnitude weaker than in the visible range – without stray-light correction what is measured there is mainly the instrument's own stray light. Dark current, non-linearity and temperature drift add to this.

Which quantities can be calculated from a spectrum?
Spectral irradiance yields almost any figure of merit by integration and weighting: irradiance per spectral range, UV dose, illuminance via V(λ), PPFD and action-weighted values such as those to IEC 62471. That is precisely the advantage over broadband radiometers for process and dose checks.

Find the right sensor: The UV Sensor Finder proposes the three best-matching UV sensors and meters by application, wavelength and form factor – with key values, system partner and reasons.