A sphere measurement setup is an ideal measurement system to compare lamps, if the lamps to be compared a) are of exact the same size and form, b) have the same spectral power distribution, c) have the same light distribution and d) show the same time dependency. If one of these boundary condition is not fulfilled - as it is typically the case when a test laboratory calibrates its sphere setup with a standard lamp and uses this calibrated sphere to measure customer lamps - corrections must be applied and/or the expected uncertainty of the comparison may be drastically increased. The reasons for the need of corrections is the different weighting of the emitted radiant flux by the sphere setup due to manifold of reasons, which comprise the non-uniformity of the sphere coating with respect to its absolute spectral and geometrical reflectivity, the influence of the size and geometry of baffles, the type of auxiliary lamp used, the capability of the detecting system, the capability of the power supply, connectors and the electronic measurement system, and finally the repeatability of the sphere lock.
To apply corrections or to determine the estimated increase in uncertainty depending on the DUT, at least some information about the influence of the sphere setup on the measurement result with respect to the influencing parameters a) to d) is needed. A straight forward approach to solve this problem is, to build a couple of different transfer standards with different spectral, geometrical, distributional and temporal behaviour, where every single standard will be traceably calibrated to National Standards. If these transfer standards are used in an ideal sphere, the sphere calibration factor calculated from every single of these different transfer standards will be the same. However, using an imperfect sphere, the sphere calibration factor will differ between the standards used, depending on the sensitivity of the sphere with respect to spectral, geometrical and temporal variations of the radiant flux.
Taking into account the needs of test laboratories, we develop a the transfer standards which is capable not only to calibrate a sphere setup under one defined radiant condition but also to characterise the measurement setup with respect to various conditions.
Our development targets:
• easy to use for test laboratory (even for semi-skilled employees)
• used as simple as a single classical standards
• affordable price for test laboratories.
To get information about the sensitivity of a particular sphere with respect to deviations in the spectral and spatial distribution of radiation emitted by a source the project ENG62 MESaIL developed a so called Multiple Transfer Standard (MTS). The MTS is a LED-based source capable to emit different spectra with different shapes and different spatial distributions and the ability to run under different modulation conditions.
It was decided to build one single transfer standard device, which is capable to operate under different defined stable and programmable operating conditions providing the variety of different spectral, geometrical and time dependent properties. In this way, the operator manipulation of the test setup is restricted to some mechanical re-configurations for characterising measurements with respect to the influence of the shape of the source.
However, even in this case it will be not necessary to remove the source from the sphere, but simply to add some shaping parts.
To accomplish this concept, the core of the MTS is based on an LED-Cube which consists of five square-cut Printed Circuit Boards (PCBs) which are connected to each other such that they build a thermally controlled one side open cube.
In addition, the prototype device can be supplemented with mechanical accessories to allow for the simulation of extended sources of different shapes (e.g. flat panels, In-Line sources and spherical bulbs).
The operating voltage of an MTS is 24 Volt DC. The typical current at maximum current setting of the LEDs is about 6 Ampere. During operation, the MTS is computer controlled via wireless connection using an eRIC4 module (433MHz). An internal multi-master bus using RS-485 (8N1,19800 baud) can be used for services and change of default settings. The stabilisation temperature of the PCBs is set by default to 35 °C to allow the system to stabilize even if all LEDs of the cube are operating at about maximum power. Temperature control can be linked to only one PCB-board or to the averaged reading of all PCB-board on the cube to avoid unbalanced thermal operation.
Depending on the set values, the LED-driver can operate the LEDs in continuous mode as well as in modulation mode. The highest frequency of the step function used to modulate the LED current, i.e. the light output, of the LED is restricted to 20 kHz. A screen dump of the MTS control software is shown in the figure.
The spectra of the five chosen types of LEDs are a follows:
The spatial distribution of the light output of the MTS if all centred white LEDs are operating (without additional mechanical accessory) is already very uniform.