Although its capabilities are far less than those of many advanced transducer interconnections, the decades-old current loop may still be the best choice for many longer-distance, higher-noise situations.
IoT connectivity is a hot topic, but it’s not a new idea, except for the “Internet” aspect. Industrial and some commercial applications have been connecting all sorts of sensors and transducers to computer-based data-acquisition and control systems for decades, and IoT is amplifying the reach and number of these end-points.
There’s a wide choice of ways to physically transfer the signals and their data; some approaches require minimal signal conditioning while others are better suited for fully conditioned and digitized signals. Among the options are wired connections such as those using RS-423/485 as well as wireless ones such as Zigbee, Bluetooth, Wi-Fi, and many proprietary approaches.
All these IoT points have one thing in common: they obviously need power to operate. This can be derived via energy harvesting (solar, vibration, thermal, and RF, for example) or a long-life battery; 10 years is a very attainable lifetime, using systems operating with a low duty cycle, low sleep current, and a suitable battery.
Ironically, there is an older, wired alternative that solves the power problem: the self-powered 20-mA current-loop interface (Figure 1). It’s easy to assume that this simple, ancient interface is nearly obsolete and not recommended for new designs, but that’s not the case at all. In fact, IC vendors are still announcing ever-more powerful versions of interface ICs for this loop.
For example, earlier this year, Maxim Integrated introduced the MAX12900, an ultra-low-power, highly integrated analog front-end (AFE) for a 4–20 mA sensor transmitter (Figure 2). It includes many extra functions and features, but at its core, it allows receiving the power and transmitting data over a 2-wire, 4–20 mA current loop. The major restriction is that the sensor and its conditioning functions not consume more than 3–4 mA in which case an external sensor supply and more wires would be needed, thus negating the advantage of the self-powered arrangement.
A little background for the current loop will provide some context and perspective on why this interface is still used in new design, and not just legacy installations. For many years, as control systems changed between the 1930s and the 1950s from pneumatic-based implementations to electrical ones, the most common interface was the 0–20 mA current loop, also called the 4–20 mA loop.
The concept is disarmingly simple and effective: the signal of interest, whether from a sensor or to an effector transducer, uses 4 mA to represent the lowest signal value and 20 mA for the highest. Although originally intended for analog signals, the loop did not have any controlling protocol, and so could be used to represent digital signals rather than analog, with minimal format structure, if any. Several decades ago, as signal-conditioning circuits and their A/D and D/A converters greatly reduced their power needs, it became practical to operate that circuitry from the few mA of available “free” power.
The current loop, whether self-powered or not, offers some major benefits in the industrial setting. As a low-impedance loop unlike voltage interfaces such as RS-423/485, it is relatively immune to EMI, which is a major advantage in that environment. Further, it has an integral, no-cost way of letting the system know that a wire has broken (the most common fault mode), since the current drops to an easily identifiable 0 mA. Finally, as a one-loop-per-endpoint interconnect, it is fairly easy to install, label, trace, and troubleshoot.
The downsides are the increasing cost in wire and installation effort as the number of individual loops increases, plus the need to provide each loop with a current source that delivers anywhere from a few volts to about 24 VDC. If there are multiple loops spread across diverse locations, there may be a need for galvanic isolation to avoid ground-loop issues.
There’s no doubt that the venerable self-powered current loop still has its place. As with most engineering decisions, there is no simple, single “right” choice, instead there is only a “best” solution in each situation, admittedly with tradeoffs and compromises. Nonetheless, don’t be too quick to rush to the latest, greatest approach when an older, long-standing one may still be a good choice.
Have you ever used the 20-mA self-powered loop for a transducer? Have you ever tried to go with a more sophisticated interface like Zigbee or Wi-Fi, or even the classical RS-422/485, only to revert to the current loop?
Bill Schweber is an EE who has written three textbooks, hundreds of technical articles, opinion columns, and product features.