For the quantification of the solar influence on the climate change on earth, it is needed to extrapolate the measured total solar irradiance variations, which cover 'only' 26 years, to climate time scales, e.g. the 300 years which separate us from the Maunder minimum. The existing extrapolations assume a long term trend of the background sun irradiance in between the 11 year solar minima. We will review what is the assumed value for the long trend and what is and what can be achieved in terms of instrument repeatability in order to be able to measure this trend. We will conclude that the measurement over one 11 year solar cycle with the side by side cavity radiometers DIARAD and TIM will provide the limit of repeatability that is needed to detect the anticipated trend.