Fifteen miniature fixed-point cells made of three different ceramic crucible materials (\(\hbox {Al}_{2}\hbox {O}_{3}\), \(\hbox {ZrO}_{2}\), and \(\hbox {Al}_{2}\hbox {O}_{3}(86\,\%)+\hbox {ZrO}_{2}\)(14 %)) were filled with pure palladium and used to calibrate type B thermocouples (Pt30 %Rh/Pt6 %Rh). A critical point by using miniature fixed points with small amounts of fixed-point material is the analysis of the melting curves, which are characterized by significant slopes during the melting process compared to flat melting plateaus obtainable using conventional fixed-point cells. The method of the extrapolated starting point temperature using straight line approximation of the melting plateau was applied to analyze the melting curves. This method allowed an unambiguous determination of an electromotive force (emf) assignable as melting temperature. The strict consideration of two constraints resulted in a unique, repeatable and objective method to determine the emf at the melting temperature within an uncertainty of about \(0.1\,\upmu \hbox {V}\). The lifetime and long-term stability of the miniature fixed points was investigated by performing more than 100 melt/freeze cycles for each crucible of the different ceramic materials. No failure of the crucibles occurred indicating an excellent mechanical stability of the investigated miniature cells. The consequent limitation of heating rates to values below \({\pm }3.5\,\hbox {K}\,\hbox {min}^{-1}\) above \(1100\,{^{\circ }}\hbox {C}\) and the carefully and completely filled crucibles (the liquid palladium occupies the whole volume of the crucible) are the reasons for successfully preventing the crucibles from breaking. The thermal stability of the melting temperature of palladium was excellent when using the crucibles made of \(\hbox {Al}_{2}\hbox {O}_{3}(86\,\%)+\hbox {ZrO}_{2}(14\,\%)\) and \(\hbox {ZrO}_{2}\). Emf drifts over the total duration of the long-term investigation were below a temperature equivalent of about 0.1 K–0.2 K.