Both Differential Thermal Analysis (DTA) and Differential Scanning Calorimetry (DSC) are concerned with the measurement of energy changes in materials. They are thus the most generally applicable of all thermal analysis methods, since every physical or chemical change involves a change in energy.

DTA is the older technique. The principle of the "classical" arrangement is readily explained with reference to the following figures.

S and R are containers holding the Sample and an inert Reference material. In these are thermocouples measuring their respective temperatures. By connecting the thermocouples in opposition, the difference in temperature (DeltaT) is also measured. If S and R are heated at the same rate, by placing them in the same furnace, their temperatures will rise as in the middle figure. T_R rises steadily, as the reference material is chosen to have no physical or chemical transitions. T_S also rises steadily in the absence of any transitions, but if for instance the sample melts, its temperature will lag behind T_R as it absorbs the heat energy necessary for melting. When melting is complete, steady heating is resumed. The right-hand figure shows the DTA curve - a plot of DeltaT against time, or more usually, sample temperature. The curve shows an endothermic (heat-absorbing) peak. If an exothermic (heat-producing) event had occurred, the curve would show a peak in the opposite direction. The area A on the curve is proportional to the heat of the reaction:

The constant K comprises many factors, including the thermal properties of the sample, and varies with temperature. The generation of quantitative data using the "classical" arrangement above is laborious.

Nowadays the thermocouples are rarely, if ever in the sample itself, but are placed below the container, which has the effect of reducing the influence of sample properties on the area of the DTA peak. With such designs, it is easier to determine the variation in K with temperature, and quantitative data are more readily obtained. This approach led to the development of heat-flux DSC (see below).

DTA instruments are still valuable, particularly at higher temperatures (>1000°C), or in aggressive environments, where true heat-flux DSC instruments may not be able to operate.

Most DSC instruments are of the heat-flux design, a schematic of which is shown below. There is another type of instrument, "power-compensated DSC", which is discussed in standard texts, and for most practical purposes gives equivalent results to good heat-flux designs. The figure most closely resembles the TA Instruments design of cell, but the features are common to most. Small, flat samples are contained in shallow pans, with the aim of making a good thermal contact between sample, pan and heat flux plate. Symmetrical heating of the cell, and therefore S and R, is achieved by constructing the furnace from a metal of high thermal conductivity - silver in the case of the TA Instruments design. Note the provision for establishing a gas flow through the cell, to sweep away volatiles, provide the required atmosphere, and to assist in heat transfer.

The control of the furnace, signal acquisition, and data storage and analysis are of course handled by a computer.

The primary signals from the cell are of the order of mV for the temperature, and µV for DeltaT. Low noise high gain amplifiers are necessary to boost these signals before data logging. Reproducible construction results on a known variation in sensitivity to heat flow with temperature, and software correction results in an effectively constant sensitivity over the working range, which is typically up to 700°C, and down to ca. -140°C with a liquid nitrogen cooling system.

Temperature calibration is carried out by running standard materials, usually very pure metals with accurately known melting points. Energy calibration may be carried out by using either known heats of fusion for metals, commonly indium, or known heat capacities. Synthetic sapphire (corundum, or aluminium oxide) is readily available as a heat capacity standard, and the values for this have been accurately determined over a wide temperature range. The absolute accuracy for measurements of heat capacity and transformation enthalpies are more often limited by the lack of appropriate standards, and difficulties in assigning a baseline construction, than by limitations of the instrument itself.

A variety of sample pans can be used for different purposes. The best quantitative results for polymers are obtained from thin samples crimped flat between the pan and a lid. Hermetically-sealed pans capable of holding a few atmospheres pressure are used for liquids, or when it is necessary to retain volatiles. Very high-pressure seals can be achieved using O-ring or screw-threaded seals. For materials that react with aluminium, or for higher temperatures, pans may be made from stainless steel, inconel, gold, alumina, graphite, silica or platinum.

Typical purge gases are air and nitrogen, though helium is useful for efficient heat transfer and removal of volatiles. Argon is preferred as an inert purge when examining samples that can react with nitrogen. The experiment can also be carried out under vacuum or under high pressure using instruments of the appropriate design.