Temperature compensation

All the resistors change value somewhat when temperature changes noticeably. And as the resistors dissipate power, they can get hot due to the current they carry. Often, this current is so tiny that it does not appreciably heat the resistor. But in some cases it does, and resistance may change. Then the circuit will behave quite differently than it did when the resistor was cool.

There are several ways to approach problems of resistors changing the value when they get hot. One method is to use the specially manufactured resistors which do not appreciably change the value when they get hot. This type of units is called as temperature-compensated. But one of these can cost a number of times as much as the ordinary resistor. Another approach is to use the power rating which is higher than the actual dissipated power in resistor. This will keep the resistor from becoming very hot. Usually, it is needless expense to do this, but if small change in the value cannot be tolerated, it is sometimes the very much cost effective.

Still another scheme is to use the series-parallel network of resistors which are all identical, in the manner about which you know, to increase power dissipation rating. Alternatively, you can take many resistors, say 3 of them, each with about 3 times the intended resistance, and connect them all in the parallel. Or you can take many resistors, say 4 of them, each with about 1/4 the intended resistance, and connect them in the series. It is not wise to combine several resistors having different values. This can result in one of them taking most of load while the others loaf, and combination will be no better than single hot resistor you begin with. You may get the idea of using 3 resistors with half the value you need, but with opposite resistance-versus-temperature characteristics, and connecting them in series. Then the one whose resistance decreases with heat (negative temperature coefficient) will have a canceling out effect on the one whose resistance goes up. would be required to make a 3 × 3 series-parallel network. And you cannot be sure that the opposing effects will exactly balance. It would be better, in such a case, to create a 2 × 2 series parallel array of ordinary resistors.