In applications in which you need remote measurement of temperature with a three-wire RTD (resistance-temperature detector), it is important to eliminate the ohmic errors caused by the excitation current flowing through the wiring resistance. You can locate the RTD more than 1000 ft from the ADC with wiring resistance in the tens of ohms. Normally, you would remove the ohmic errors by using two identical current sources that convert the wiring drops to a common-mode signal that the differential input of the ADC rejects. This technique is based on the not-unreasonable assumption that the wiring resistances of the three-wire RTD are equal. Figure 1 shows a typical circuit based on this assumption. The two current sources are assumed to be identical and to track each other closely over temperature and supply-voltage changes. However, a finite level of mismatch exists. In applications in which accuracy is paramount, it may be wise to use only one excitation-current source for the RTD and thus avoid any potential mismatch between two sources. However, the single-current-source approach for exciting a three-wire RTD complicates the effort to reject the ohmic drops, because you can no longer eliminate the wiring drops as a common-mode signal.
|Figure 1.||Two current sources turn wiring drops into a common-mode signal.|
Nevertheless, you can still eliminate the wiring drops by using a two-channel ADC and a little extra software computation. You take two conversions, and software subtracts the error term stemming from the wiring resistance. In Figure 2 the wiring resistances are represented by lumped elements RL1, RL2, and RL3. Assume that the wiring resistance of all three leads is equal (RL1=RL2=RL3=RL). In fact, it is necessary only that RL2 and RL3 be equal, because RL1 appears in both equations. The circuit uses the RTD in an "upside-down" fashion. Two FET switches, SWA and SWB, direct the excitation current through the appropriate legs of the RTD. To avoid interruptions in the current flow, make-before-break switching is advisable. Starting with both switches closed, SWA opens, and the AD7711A takes a measurement. The measured voltage is 2I1RL, which represents the out-and-back wiring drop. Next, SWA closes, and SWB opens. The ADC takes a measurement on Channel 2. The measured voltage is 2I1RL+VRTD. This signal represents the out-and-back wiring drop plus the desired signal. The VRTD term is the first result subtracted from the second. The onboard, 400-µA current source of the AD7711A serves as the excitation current in Figure 2.
|Figure 2.||One current source and a software subtraction eliminate wiring-drop errors.|
You may have to register before you can post comments and get full access to forum.