Working on a test setup for LED parameter extraction over temperature. I'm using a peltier element to heat/cool a block of aluminum that the LED is mounted to. The LED is powered using a fixed 10mA supply. The most important bits are probably the change in peak intensity (which correlates to lower efficiency), and the Vf drop. I need to work out a faster way to measure IV curves so that I can measure higher currents without adding too much self-heating.