Page 460 - The Mechatronics Handbook

0066_Frame_C19 Page 82 Wednesday, January 9, 2002 5:27 PM

the resistance of the RTD is high compared to the lead resistance, or where high accuracy is not required.
Three wire RTDs, Fig. 19.58(b), allow an equal length of lead wire to be included in each side of the
bridge so that changes in lead resistances are felt equally on both sides and the balance of the bridge is
not affected. It should be noted that four-wire RTDs of the type shown in Fig. 19.58(d) can be used
with readers designed for two-, three-, or four-wire RTDs and unless expense or size is a concern, they
are recommended to allow for future upgrades of the reader. A fairly rare alternate configuration of a
four wire RTD probe is shown in Fig. 19.58(c) where two of the lead wires form a closed loop. This
arrangement usually is needed only when multiple RTDs are in the bridge and should not be confused
with the other type of four-wire RTD shown in Fig. 19.58(d), which has two wires going to each side
of the resistor.
Another problem common to all reading schemes is that of self-heating of the RTD due to the
measuring current through the resistor. This can be minimized by using small currents, having a high
heat transfer coefficient between the sensor and the measured process, or by using low duty time pulsed
measurements. Although steps can be taken to minimize it, self-heating can never be eliminated. It is
such a ubiquitous characteristic of RTDs that it can be exploited as a means of determining the heat
transfer coefficient of a system, such as when platinum resistors are used as anemometers.
As indicated above, the measurement of temperature using an RTD simply requires determining its
resistance. The three conventional techniques are briefly discussed here.

Fully Balanced Bridges
A primitive but highly accurate means of determining the resistance of an RTD is to use it in one leg of
a wheatstone bridge and have a calibrated variable resistor in the opposite leg such that when the bridge
is brought to balance, as indicated by a null current on a galvanometer, the resistance of the RTD is the
same as the calibrated variable resistor. There are several variations on this theme to eliminate most
sources of error by splitting the error-forming device between legs of the bridge. For example, if the
RTD is remote from the bridge, a three-wire lead configuration can be used to put an equal length of
lead wire in both sides of the bridge with the “corner” of the bridge now defined at the remote RTD
location as shown in Fig. 19.58(b). Ultimately the accuracy of the reading is limited by the accuracy of
the resistances in the bridge, which can be made quite accurate. Since the variable resistor must be
physically adjusted to null the bridge, this technique is not readily adapted to data loggers or temperature
displays.
Unbalanced Bridges
If the calibrated variable resistor in the above scheme is replaced by a fixed resistor, the extent of the
imbalance of the bridge can be measured with a voltmeter in place of the nulling galvanometer. This
technique is fraught with errors since the imbalance of the bridge is not linear with the resistance of the
RTD and the reading of the voltmeter is also proportional to the excitation current. Nevertheless, this
can be a practical means of measuring temperatures over a narrow range if the values of the resistors in
the far legs of the bridge are much higher than the RTD and the opposite leg resistance is close to that
of the RTD. As with the other bridge techniques, three lead wires can be arranged in opposite legs of the
bridge for remote sensors to compensate for lead resistance changes. If opposite legs of the bridge both
consist of RTDs, then the bridge imbalance is proportional to the difference in the temperature of the
two RTDs and the only way to keep equal lead lengths in both legs is to use the special four-wire type
shown in Fig. 19.58(c).

Direct Voltage vs. Current Measurements
If a constant current is passed through a resistor, the voltage across it is proportional to resistance. This
is a simple concept but had to await the advent of accurate constant current sources and high impedance
voltage amplifiers to become a practical replacement to the bridge techniques. By using four lead wires,
as shown in Fig. 19.58(d), the current circuit can be made independent of the voltage sensing circuit and
the lead resistances have no affect on the reading.

©2002 CRC Press LLC