Lock-in thermography (LIT) is an imaging method that depicts radiated heat and
its diffusion in manifold samples. LIT offers versatile possibilities for the characterization
of solar cells and modules since the radiated heat is proportional to the dissipation
of electrical power. Up to now, the quantitative correlation of detected heat and dissipated
electrical power has been known for silicon solar cells only. For many other types
of solar cells and modules – especially thin film solar cells – LIT has been used as a
qualitative measurement tool for depicting the location of defects, for example. Thus,
the potential of LIT in terms of the calculation of power generation and dissipation in
thin film solar cells has not been exploited. This visualization and calculation of power
flows leads to a better understanding of the influences of defects on the efficiency of
solar modules. Furthermore, it enables the evaluation of potential improvements, which
results in solar modules with higher efficiencies, produced to lower costs.
In order to interpret LIT signals accurately, the lock-in algorithm and particularly
its limits have to be understood. The present thesis shows the evaluation of the lockin
algorithm and its algebraic complex result with simulations. It is found that the weak
points of the lock-in algorithm lie in the sampling of the acquired heat signal. Sampling
moments that are not uniformly distributed in a lock-in period produce unreliable results.
A low sampling at high measurement frequencies shows significant deviations
distorting the LIT result. The findings allow for the development of user-friendly LIT
systems that automatically avoid sampling errors and produce reliable LIT results.
The comprehension of LIT measurements of thin film solar cells needs a theoretical
thermal model for the solar cells that can be used to solve the differential heat
diffusion equation. The solution describes the surface temperature distribution that is
acquired in LIT measurements. By the evaluation of the frequency response of a point
heat source in a thin film solar cell, a simple thermal model representing a solid body is
found to adequately reproduce LIT measurements.
LIT investigations in the scale of the thermal diffusion length are hampered by
the diffusion of heat that leads to a blurring of heat sources. With the description of the
thermal model and a Fourier transform technique, it is possible to successfully deconvolute
the heat generating sources from the heat diffusion, meaning the removal of the
thermal blurring. This leads to the unimpeded visualization of the dissipated power of
small heat sources such as shunts or the series interconnection of cells in a thin film
solar module.
Max Henrik Siegloch
LIT Lock-in thermography Solar Cells