Modern industrial automation applications—such as PLCs, factory process control, and intelligent transmitters—all demand high-performance, low-power components. Current-mode data transmission is no exception. Intelligent 4–20 mA current-mode data transmission is the preferred technique in many industrial automation applications and is a well-established standard for communications between the host computer and smart/intelligent transmitters in harsh factory environments. This article describes intelligent transmitters and explains their need for high-resolution, low-power ADCs, DACs, and isolation devices.
Why Use 4–20 mA Loops for Data Transmission?
When transmitting low-amplitude, low-frequency data signals over several hundred meters in a noisy industrial control environment, current-loop transmission is well established and is preferred over voltage-mode transmission for the following reasons:
- 1. Insensitivity to IR drops makes current loops suitable over long distances. In contrast, voltage at any point depends on line resistance and capacitance and therefore varies with cable length.
- 2. Current transmission allows a single 2-wire cable to carry both power and signals at the same time, an important factor when powering electronic components in remote locations.
- 3. Current loops don't require a precise or stable supply voltage.
- 4. Inexpensive 2-wire twisted-pair cables offer good noise immunity and lower EMI sensitivity.
- 5. It's easy to detect offline sensors, broken transmission lines, and other fault mechanisms.
Intelligent 4–20 mA Transducer Design
In a 2-wire, 4–20 mA current loop, the supply current for the sensor electronics must not exceed 4 mA (the remaining 16 mA carries the signal information), so the components that make up the transmitter must be low power. As shown in Figure 1, smart transmitter systems use five building blocks: an ADC, a microcontroller or DSP, memory (RAM), a DAC with an optional integrated amplifier and reference, and a sensor or transducer.
Figure 1. Block diagram of a smart 4–20 mA transmitter
The microprocessor or microcontroller performs linearization and other functions on the sensor data and communicates them back to the host system. The transducer voltage—usually ranging from a few millivolts to a few volts, depending on the type of sensor (Figure 2)—must be digitized by a high-precision ADC before the signal reaches the processor. Choosing a high-resolution ADC with an onchip low-noise programmable-gain amplifier (PGA) and good offset, full-scale, and drift specifications ensures precise conversion of the sensor inputs with minimum noise and drift due to external temperature variations.
Figure 2. Frequently used industrial automation sensors and their input spans Note: *A/D input spans from resistive sensors depend on the magnitude of the excitation current sources
For high-end applications, the rms noise requirements must be <100 nV at high gain settings (e.g., 64 or 128), with offset and gain drift of 10 nV/°C and 1 ppm/°C, respectively. Operating current consumption ideally should be <400 µA. The ADCs used in 4–20 mA transducer designs often include simultaneous 50/60 Hz rejection filters, onchip matched current sources for cold junction compensation and resistance temperature detector (RTD) biasing, and a precision reference. If integrated, these features simplify the design task significantly by eliminating some of the board and layout challenges presented by discrete components while simultaneously reducing cost.
Interfacing to RTDs
The resistance of an RTD varies with temperature. Typical elements used for RTDs are nickel, copper, and platinum, with 100 Ω and 1000 Ω platinum (PT100 or PT1000) RTDs being the most common. RTDs are useful for measuring temperatures from –200°C to 800°C and have a near-linear response over this temperature range. Figure 3 shows how to interface a ΔΣ ADC, such as the Analog Devices AD7793, to a commonly used 3-wire PT100 RTD.
Figure 3. Interfacing the AD7793 ADC to a PT100 RTD
In this 3-wire configuration, the lead resistances (RL1, RL2, and RL3) will cause errors if only one current source (IOUT1) is used, because the excitation current will flow through RL1, developing a voltage error between AIN1(+) and AIN1(–), the positive and negative terminals of the differential analog input ADC channel. The second RTD current source (IOUT2) is used to compensate for the error introduced by the excitation current flowing through RL1. While the absolute accuracy of each current source is not important, good matching of the two current sources is essential. The second RTD current flows through lead resistance RL2. Assuming RL1 and RL2 are equal and IOUT1 and IOUT2 match, the error voltage across RL2 cancels the error voltage across RL1, and no error voltage is developed between AIN1(+) and AIN1(–). The ADC in this example has differential analog inputs and accepts a differential reference, allowing the implementation of a ratiometric configuration. The ADC's reference voltage is also generated using the matched current sources; it is developed across the precision resistor RREF, and is applied to the differential reference inputs of the ADC. This scheme ensures that the analog input voltage span remains ratiometric to the reference voltage. Any errors in the analog input voltage due to temperature drift of the RTD current source are compensated for by the variation in the reference voltage.
If the sensor resides in a harsh industrial environment, safety measures require both intrinsically safe operation and an isolation barrier to prevent ground loops. Isolation devices are used to protect against high voltages or currents caused by line surges or ground loops, which can occur in any system that has multiple paths to ground. System grounds that are separated by long cables will not be at the same potential, resulting in ground current between the two systems. Without isolation, this current could introduce noise, degrade measurements, or even destroy system components.
For smart transmitter systems, the isolation barrier can be placed between the microcontroller and the digitized sensor data. Data can be transferred across this barrier by magnetic couplers, which are a good alternative to using optocouplers, and offer the following advantages:
- 1. Lower power, resulting in reduced heat dissipation
- 2. More reliable (no LED aging or temperature sensitivity)
- 3. Faster data rates
- 4. Multiple channels on one device, resulting in significant space reduction
- 5. Higher DC accuracy
Digital isolators, such as the quad-channel Analog Devices ADuM1401, simplify 3-wire serial interface communications between the ADC and the controller and encompass an additional channel to incorporate data readback functionality (Figure 4).
>Figure 4. Functional block diagram of the ADuM1401 digital isolator (3/1 channel directionality)
The next step up is an even smarter circuit called an intelligent transmitter (Figure 5). In an intelligent transmitter, the functions of the microcontroller are shared among deriving the primary measurement signal, storing information regarding the transmitter itself, and managing a communications system that can superimpose two-way communications on the same circuit used to carry the measurement signal. A smart transmitter incorporating the commonly used HART protocol is an example of an intelligent transmitter.
Figure 5. Block diagram of an intelligent 4–20 mA transmitter
Once the data have been appropriately conditioned and digitized, they can be transmitted through the 4–20 mA current loop using a DAC. Figure 6 shows how to create a digitally controlled 4–20 mA current loop by combining a resistor and a MOSFET with a DAC, such as the Analog Devices AD5660. The DACs used to control 4–20 mA current loops must be high precision, low power, low voltage, and ideally have internal amplifiers able to control the gate voltage of the external MOSFET, thus simplifying the design task. The SPI-, QSPI-, I2 C-, and Microwire-compatible interfaces on the DAC simplify interfacing, allowing their use with most of the processors and controllers found today.
Figure 6. Digitally controlled 4–20 mA current loop using the AD5660 DAC
The only drawback to this configuration is the need to drive the external MOSFET, which requires a much higher supply voltage. Most industrial control applications provide low and high voltages to support both 3 V and 5 V logic control and the sensors, which can demand voltages as high as 36 V (24 V nominal). In the example in Figure 6, the external MOSFET drives the current loop with a current given by Equation 1 and determined by the 50 α sense resistor.
The DAC's resolution is N and the decimal equivalent of its input code is D, which represents the sensor's digitized output after processing. Resolution for the serial-input AD5660 DAC is 16 bits, yielding a full-scale range of 2N = 216 = 65536 codes. For a 1.25 V reference, 50 α sense resistor, and full-scale DAC output of D = 65536, the output current is IOUT = VRef/RSense = 1.25 V/50 Ω = 25 mA. (In most applications, the current loop must be capable of delivering a 10% or greater over-range.)
For transducers in remote locations with no local power source available (such as in flowmeter applications), the electronics must be powered from the 4–20 mA loop. In this case, the application needs a voltage regulator, which converts the loop voltage to a fixed 3 V/5 V for operating the various sensor electronics. The voltage regulator can be any low-cost device with a quiescent current sufficiently lower than the 4 mA budget.
More and more industrial automation applications are requiring low-power converters both to accurately measure and control various processes and to transmit data afterward. These end applications demand increased performance, robustness, and feature sets, while simultaneously reducing costs and board area. Intelligent 4–20 mA transducers are one such application. Component manufacturers are addressing these challenges and offering a number of converter and isolation products to address the needs of system designers for current and future designs.