Some electronic components develop lower resistances or lower triggering voltages (for nonlinear resistances) as their internal temperature increases. If circuit conditions cause markedly increased current flow in these situations, increased power
dissipation may raise the temperature further by
Joule heating. A
vicious circle or
positive feedback effect of thermal runaway can cause failure, sometimes in a spectacular fashion (e.g. electrical explosion or fire). To prevent these hazards, well-designed electronic systems typically incorporate current limiting protection, such as thermal fuses, circuit breakers, or
PTC current limiters. To handle larger currents, circuit designers may connect multiple lower-capacity devices (e.g. transistors, diodes, or
MOVs) in
parallel. This technique can work well, but is susceptible to a phenomenon called
current hogging, in which the current is not shared equally across all devices. Typically, one device may have a slightly lower resistance, and thus draws more current, heating it more than its sibling devices, causing its resistance to drop further. The electrical load ends up funneling into a single device, which then rapidly fails. Thus, an array of devices may end up no more robust than its weakest component. The current-hogging effect can be reduced by carefully matching the characteristics of each paralleled device, or by using other design techniques to balance the electrical load. However, maintaining load balance under extreme conditions may not be straightforward. Devices with an intrinsic
positive temperature coefficient (PTC) of electrical resistance are less prone to current hogging, but thermal runaway can still occur because of poor heat sinking or other problems. Many electronic circuits contain special provisions to prevent thermal runaway. This is most often seen in transistor biasing arrangements for high-power output stages. However, when equipment is used above its designed ambient temperature, thermal runaway can still occur in some cases. This occasionally causes equipment failures in hot environments, or when
air cooling vents are blocked.
Semiconductors Silicon shows a peculiar profile, in that its
electrical resistance increases with temperature up to about 160 °C, then starts
decreasing, and drops further when the melting point is reached. This can lead to thermal runaway phenomena within internal regions of the
semiconductor junction; the resistance decreases in the regions which become heated above this threshold, allowing more current to flow through the overheated regions, in turn causing yet more heating in comparison with the surrounding regions, which leads to further temperature increase and resistance decrease. This leads to the phenomenon of
current crowding and formation of
current filaments (similar to current hogging, but within a single device), and is one of the underlying causes of many
semiconductor junction failures.
Bipolar junction transistors (BJTs) Leakage current increases significantly in
bipolar transistors (especially
germanium-based bipolar transistors) as they increase in temperature. Depending on the design of the circuit, this increase in leakage current can increase the current flowing through a transistor and thus the
power dissipation, causing a further increase in collector-to-emitter leakage current. This is frequently seen in a
push–pull stage of a
class AB amplifier. If the pull-up and pull-down transistors are
biased to have minimal
crossover distortion at
room temperature, and the biasing is not temperature-compensated, then as the temperature rises both transistors will be increasingly biased on, causing current and power to further increase, and eventually destroying one or both devices. One rule of thumb to avoid thermal runaway is to keep the
operating point of a BJT so that Vce ≤ 1/2 Vcc. Another practice is to mount a thermal feedback sensing transistor or other device on the
heat sink, to control the crossover bias voltage. As the output transistors heat up, so does the thermal feedback transistor. This in turn causes the thermal feedback transistor to turn on at a slightly lower voltage, reducing the crossover bias voltage, and so reducing the heat dissipated by the output transistors. If multiple BJT transistors are connected in parallel (which is typical in high current applications), a current hogging problem can occur. Special measures must be taken to control this characteristic vulnerability of BJTs. In power transistors (which effectively consist of many small transistors in parallel), current hogging can occur between different parts of the transistor itself, with one part of the transistor becoming more hot than the others. This is called
second breakdown, and can result in destruction of the transistor even when the average junction temperature seems to be at a safe level.
Power MOSFETs Power
MOSFETs typically increase their on-resistance with temperature. Under some circumstances, power dissipated in this resistance causes more heating of the junction, which further increases the
junction temperature, in a
positive feedback loop. As a consequence, power MOSFETs have stable and unstable regions of operation. However, the increase of on-resistance with temperature helps balance current across multiple MOSFETs connected in parallel, so current hogging does not occur. If a MOSFET transistor produces more heat than the
heatsink can dissipate, then thermal runaway can still destroy the transistors. This problem can be alleviated to a degree by lowering the
thermal resistance between the transistor die and the heatsink. See also
Thermal Design Power.
Metal oxide varistors (MOVs) Metal oxide
varistors typically develop lower resistance as they heat up. If connected directly across an AC or DC power bus (a common usage for protection against
voltage spikes), a MOV which has developed a lowered trigger voltage can slide into catastrophic thermal runaway, possibly culminating in a small explosion or fire. To prevent this possibility, fault current is typically limited by a thermal fuse, circuit breaker, or other current limiting device.
Tantalum capacitors Tantalum capacitors are, under some conditions, prone to self-destruction by thermal runaway. The capacitor typically consists of a
sintered tantalum sponge acting as the
anode, a
manganese dioxide cathode, and a
dielectric layer of
tantalum pentoxide created on the tantalum sponge surface by
anodizing. It may happen that the tantalum oxide layer has weak spots that undergo
dielectric breakdown during a
voltage spike. The tantalum sponge then comes into direct contact with the manganese dioxide, and increased leakage current causes localized heating; usually, this drives an
endothermic chemical reaction that produces
manganese(III) oxide and regenerates (
self-heals) the tantalum oxide dielectric layer. However, if the energy dissipated at the failure point is high enough, a self-sustaining
exothermic reaction can start, similar to the
thermite reaction, with metallic tantalum as fuel and manganese dioxide as oxidizer. This undesirable reaction will destroy the capacitor, producing
smoke and possibly
flame. Therefore, tantalum capacitors can be freely deployed in small-signal circuits, but application in high-power circuits must be carefully designed to avoid thermal runaway failures.
Digital logic The
leakage current of logic switching transistors increases with temperature. In rare instances, this may lead to thermal runaway in digital circuits. This is not a common problem, since leakage currents usually make up a small portion of overall power consumption, so the increase in power is fairly modest — for an
Athlon 64, the power dissipation increases by about 10% for every 30 degrees Celsius. For a device with a
TDP of 100 W, for thermal runaway to occur, the heat sink would have to have a
thermal resistivity of over 3 K/W (kelvins per watt), which is about 6 times worse than a stock Athlon 64 heat sink. (A stock Athlon 64 heat sink is rated at 0.34 K/W, although the actual thermal resistance to the environment is somewhat higher, due to the thermal boundary between processor and heat sink, rising temperatures in the case, and other thermal resistances.) Regardless, an inadequate heat sink with a thermal resistance of over 0.5 to 1 K/W would result in the destruction of a 100 W device even without thermal runaway effects.
Batteries When handled improperly, manufactured defectively, or damaged, some
rechargeable batteries can experience thermal runaway resulting in overheating. Sealed cells will sometimes explode violently if safety vents are overwhelmed or nonfunctional. Especially prone to thermal runaway are
lithium-ion batteries, most markedly in the form of the
lithium polymer battery. Lithium-ion batteries are often found in everyday consumer electronics and vehicles. Reports of exploding cellphones occasionally appear in newspapers. In 2006, batteries from Apple, HP, Toshiba, Lenovo, Dell and other notebook manufacturers were recalled because of fire and explosions. The
Pipeline and Hazardous Materials Safety Administration (PHMSA) of the
U.S. Department of Transportation has established regulations regarding the carrying of certain types of batteries on airplanes because of their instability in certain situations. This action was partially inspired by a cargo bay fire on a
FedEx airplane. One of the possible solutions is in using safer and less reactive anode (lithium titanates) and cathode (
lithium iron phosphate - LFP) materials — thereby avoiding the
cobalt electrodes in many lithium rechargeable cells — together with non-flammable electrolytes based on ionic liquids. ==Astrophysics==