Experiments with Peltier Junctions Pulsed with High Transient Currents

Abstract
It was shown in a previous article by Landecker and Findlay that when transient currents are passed through a cooling Peltier junction the resulting transient temperature depends strongly on the current waveform. In particular, an extremal of the current as a function of time exists for which the resulting transient temperature is much lower than for any other current waveform. With the aid of an experimental method developed for the purpose, temperatures well below 100°K were in fact observed with bismuth telluride junctions. Since the occurrence of temperature transients producing such low junction temperatures is rather difficult to reconcile with some notions of solid‐state theory, these results have been reexamined for current densities of up to ∼4000 A/cm2 using a modified technique that eliminates possible sources of systematic errors in the previous method. The results may be summarized as follows. Up to the highest current densities that could be produced a monotonic increase of the temperature drop with final current density was observed. The final temperature drops ranged from 205° to 120° measured from a reference temperature of ∼20°C (room temperature) depending on the efficiencies of the junctions. These efficiencies were difficult to predict from the performance of the junctions with stationary currents. For the highest currents and the very low transient temperatures an error of up to 150° may arise from imperfect heat transfer from the hot junctions to the water‐cooled heat sinks unless special probes are provided to measure the Seebeck voltage. It is very likely that the accuracy of the measurements cannot be expected to be improved much further as long as grown crystalline n and p branches are used. It is hoped to produce sintered junctions of more suitable shape and greater mechanical strength for further experiments.

This publication has 3 references indexed in Scilit: