124
Principles of Operation
where i is the nominal current produced by the AD590, and K is in Kelvin.
The 6100uses i to determine the nominal temperature, T
, by the formula:
n
= (i /(1 A / K) ) - 273.15
T
n
is in C.
where T
n
The displayed temperature, T
= C1 + (C2 * T
), is then computed, where C1
d
n
and C2 are the constants stored in the 6100for the AD590. The AD590
grades of tolerance vary, but typically, without adjusting C1 and C2, the
temperature accuracy is 1C over its rated operating range. However, the
AD590 is not perfectly linear, and even with C1 accurately known there is a
non-linear absolute temperature error associated with the device. This non-
linearity is shown in Figure 36, reprinted from Analog Devices
specifications, where the error associated with C1 is assumed to be zero.
Figure 36
AD590 Nonlinearity
If a maximum absolute error of 0.8C is tolerable, the one point calibration of
C1 should be used. If a greater accuracy is desired, the two point method of
determining C1 and C2 should be used. Note however, the absolute error
curve is non-linear, therefore the constant C2 will vary for different
measurement points.
6.3.2.3
LM335 Sensor
The LM335 is a linear thermal sensor that acts as a voltage source. It
produces a voltage, v, which is directly proportional to absolute temperature,