UV monitors

Peter Lundskow <plundskow [at] dnr__state__in__us> writes

>Recently, we purchased a more accurate meter, the UVX Radiometer
>from UVP Ultra-Violet Products (this meter comes with certification
>of calibration)it takes readings in microwatts per cm squared.  I
>have not been able to find any information on how to convert one
>measurement to the other and what safe readings would be for museums
>when measured in microwatts per cm squared. Also, I would mention
>our new meter has two sensor heads, one in the 250nm range and the
>other in the 360nm range.  I have been told by the same source that
>readings with both sensors are needed to get a more complete picture
>of UV exposure.

You may want to consider the Hanwell Instruments ULM Universal Light
Meter for taking field measurements <URL:http://www.hanwellusa.com>.
It will display light levels in lux and uv levels in both
microwatts/lumen and milliwatts/sq meter. This would eliminate the
need to take a separate lux reading and perform the calculation to
obtain the microwatt/lumen measurement. Price is $1195 USD.

Following is an excerpt from the ULM manual written by Dr. Martin
Hancock of Hanwell Instruments concerning design considerations and
limitations of such an instrument.

    "The measurement of Lux and particularly UV is subject to many
    potential errors when performed outside laboratory conditions, a
    fact that is usually glossed over. This note explains some of
    the potential problems and gives some methods which can be used
    to improve the accuracy of measurements. When using equipment
    short of that found in a Physics laboratory, reliable
    measurements require a good degree of knowledge on the part of
    the user.

    Lux is a measurement of the brightness of the incident light but
    measured in such a way as to be comparable with the subjective
    opinion formed by the human eye. In practice this means choosing
    a detector that has a similar spectral response. Hanwell
    instruments use a photocell which gives an accuracy of
    approximately 1% fsd.

    It is however not a all clear that this is what is required for
    conservation purposes: the human eye response has no
    significance for objects. In the event of the spectrum of the
    incident light source being greatly different to the calibration
    source, very large errors could arise in using Lux measurements
    for assessing cumulative damage.

    Another issue is the spatial response of the sensor, in a sense
    that the Lux measurement should be independent of the intensity
    distribution of the source. This is known as cosine-correction
    and in practice is attempted by fitting a diffusing element in
    front of the photocell. In the general case of multiple light
    sources with different spectra, it is clear that this is not a
    complete solution.

    The problems above are even more applicable to UV measurements
    when there is also a requirement to provide a ratiometric
    measurement (which by its very nature introduces uncertainty)
    the end result must be interpreted with great care. To
    illustrate this consider the following points:

        *   The spectral response of a UV filter is typically
            Gaussian-like in the spectral range 250...395 nm. This
            is inevitable if a 'breakthrough' form visible light
            (above 400 nm) is not to be included in UV measurements.

        *   The spectral response of a typical UV detector diode is
            not flat in the UV region.

        *   The pass characteristics of a suitable filter material
            is not flat in the UV region.

    These facts mean that the calibration factor varies from source
    to source although fortunately not by much for typical museum
    lighting types.

    The currents generated by such heavily filtered diodes at low
    intensities are comparable to bias and leakage currents in
    typical electronics, making the measurement potentially very
    noisy.

    The circuitry responds to UV power. The ratio to Lux (uW/Lumen)
    is then deduced afterward by hardware or software. This can lead
    to huge problems in practice. As an illustrative extreme example
    consider a Lux measurement of 1 (+/- 1 unit) and a UV power
    measurement of 1 (+/- 1 unit) When you take the ratio you can
    get any answer you like between 0 and infinity!

    With these problems in mind Hanwell Lux/UV instruments have been
    designed with the following features:

        *   All calibration is performed in software with the
            original figures obtained using a 'black-body' source
            (Tungsten bulb). The software allows multiple
            calibration data sets to be entered and recalled for
            spectra corresponding to different colour
            temperatures/spectral characteristics. The relative
            magnitudes of these values can be found in standard
            texts but it is the user's responsibility to choose
            appropriate settings for the situation.

        *   The circuitry uses ultra low-leakage amplifiers and
            capacitors to ensure a highly linear characteristic over
            a very wide range of UV intensities. At the lower end,
            the measurement (the uw/lumen ratio) is aborted when the
            charging rate is so low as to make significant the ratio
            errors described above.

        *   As standard, a Optilas Series and SuperUV diode with a
            DUG11 filter material is fitted, ... but other
            combinations can be supplied on request to suit
            individual requirements."

Greg Basso
VP Applications
Hanwell Instruments, USA 516-624-2900


                                  ***
                  Conservation DistList Instance 17:18
                  Distributed: Tuesday, August 5, 2003
                       Message Id: cdl-17-18-003
                                  ***