Cabling Sensors to Address Induction, Electrostatic Coupling, and Conduction
Contributed By Digi-Key's North American Editors
Anything of an electrical nature running through industrial cable that’s not signal is noise — electromagnetic interference (EMI) and radio-frequency interference (RFI) of some form or another. Today’s automation components are routinely designed to avoid such noise by protecting signals from the electromagnetic environment in which the components are expected to operate. But preventing signal degradation also requires the careful integration of automated machines … which usually involves a combination of good design practices and electrical connectivity expertise.
Figure 1: Subcomponents and subsystems solely dedicated to the prevention of EMI usually take the form of filtering circuitry or blocking (shielding) components such as the tubular tinned copper wire shielding shown here. (Image source: Belden Inc.)
In this article, we’ll explore key design methods to:
- Reduce the internal and external component generation of EMI
- Boost component immunity (resistance) to EMI
Top design objectives here are to minimize any internally radiated emissions for each component in a design as well as its susceptibility to externally conducted emissions. For the latter, inherent immunity to externally coupled emissions must protect against unwanted electronic signals transmitted via direct conduction, inductance, or capacitive coupling.
Figure 2: 3M AB5000 Series EMI-absorbing adhesive sheets contain metal flakes to suppress radiated EMI from mobile devices and military equipment. AB6000 Series sheets include insulation, absorbing, shielding, and nonconductive layers for designs needing both EMI shielding and absorption — including cell phones, tuners, and medical devices. AB7000 Series sheets excel in and around electronic devices requiring EMI control and 50-MHz to 10-GHz signal-integrity improvement. The sheets reduce radiated IC noise as well as EMI and crosstalk inside mobile electronics and on ribbon and flex cables. (Image source: 3M)
Specific threats to signal quality
Most efforts related to designing industrial automation equipment focuses on the specification of components such as actuators and sensors. But consider the latter: If sensors are the ears and eyes of automated systems, then the cabling is the nervous system carrying the signals to the brain (or machine controller, to continue with the analogy). This cabling is exposed to various potential interference sources that can compromise system control functions.
Figure 3: Electrical components such as sensors and actuators are routinely tested for electromagnetic compatibility (EMC) and susceptibility, though the role of cabling and their connectors in maintaining and supporting electromagnetic compatibility or EMC is often overlooked. Some cable connectors mechanically secure and electromagnetically shield cable ends and serve as EMI filters. Employing planar-capacitor technology, some are capable of filtering VHF, UHF, MF1, HF, and other EMI ranges via C, CL, LC, L, and various pi topologies. (Image source: Amphenol Industrial Operations)
If a sensor, actuator, or other component relies on an inductive, capacitive, or electromagnetic principle for detection and signal generation, any PCBs contained in that system will likely require shielding as well as extensive ground planes. The latter is covered in detail in the Digi-Key article RF Shielding: The Art and Science of Eliminating Interference. In addition, strength and frequency of the potential environmental emissions should be well known or at least codified using an industrial standard at the initial design stage. Some examples of common and expected interference ranges include:
- 50 or 60 Hz — the line frequencies of utility power
- 4 to 16 kHz — as in IGBT-induced pulse-width modulation (PWM) from VFDs for electric motors
- 2.4 GHz — the Industrial Scientific and Medical (ISM) band for wireless communications.
Read more about the generation of electromagnetic fields by motors, relays, solenoids, and actuators and the specific case of protecting RS-485 serial buses from these EMI sources in the Digi-Key article How to Protect RS-485 Buses in Industrial Environments. Other interference phenomena include surges, fast transients, and electrostatic discharge (as from “static electricity” on plant personnel in dry settings or those lacking anti-static flooring) as well as lightning strikes arising from extreme weather near the plant.
Figure 4: This PC1321BP Panel PC has a capacitive touchscreen HMI. The control electronics and the screen include shields and other elements to prevent the conducted and radiated RFI. (Image source: Maple Systems)
Consider the electrically noisy application of arc welding. Welding is notorious for producing high-bandwidth electrical noise due to:
- The high energy (current) associated with the welding process
- Impedance variations during the weld
So industrial welding equipment that operates near any power lines in a facility (or even shares earth ground with other equipment) can become a significant EMI source and electrically couple with other devices — even hundreds of feet away. Specialized equipment and accessories (especially cable) must be included in such installations to prevent EMI-related operational problems.
Device specification and installation mistakes to avoid
Once a device is cabled to the larger automated system, it can exhibit communications or behaviors that:
- Only appear related to EMI
- Actually are EMI related
Symptoms of EMC problems can manifest as signal dropout, low signal-to-noise ratios, signal interferences, and unstable control loops.
Sensors that generate analog signals are most susceptible to noise, so comparable digital devices are often preferred. These are sensor versions that generate digital PWM, frequency, or serial output signals more impervious to EMI. One caveat here is that the high switching frequencies of certain digital signals can cause ringing (voltage or current output oscillations) with exponential decay at the transitions. Such ringing is often remedied with a small decoupling capacitor or an attenuating resistor at the receiver end of the sensor system.
Learn more about the difference between analog and digital device signals at the Digi-Key Cable Matters Learning Module.
Where available, sensors that can output differential output are preferred. Sensors operating in differential mode (with a signal A accompanied by its inverted signal A/) effectively avoid all common-mode noise. Taking this EMI immunity further are twisted-pair signal wires that (when correctly installed) register induced noise identically on both wires for maximally effective noise rejection.
On the signal side of a sensor’s cable, low capacitance is critical to minimizing EMI susceptibility. Another advantage is that low-capacitance signals carrying frequency-based data can best maintain the stability of the output driver signals as signal frequency changes. In contrast, excess capacitance can cause signal roll off and sometimes reduce overall output to below the detection threshold. This intermittent effect is often quite subtle but easily diagnosed with an oscilloscope.
In a perfect world, cabling transmits clean power signals and reference values to power sensors and actuators. It then returns the system controller perfectly clean sensor and actuator status signals. As simple as this may seem, cables attached to sensors or actuators are a significant and vulnerable part of the electrical circuit — and a primary zone for increased EMI susceptibility. That’s because they can under certain circumstances behave as long antennae.
Design tip: Account for the power loss caused by particularly long cable runs — those exceeding 500 ft or so — especially if the power conductors are 22 gage or smaller in diameter and current is 500 mW or more per device.
Another tip for proper sensor connection: Understand and carefully connect the conductors of the cable power side … a connection usually taken for granted. For a lot of sensors and actuators, this power connection provides a 5 to 28-V reference to drive the signals ultimately returned to the controller. The two conductors of the cable’s power side are often called power and ground. This is not strictly correct — and (if these labels inform the installation approach) can lead to interference problems. More correctly, a sensor’s power-side ground should be called the signal common. This is because the power supply return terminates at the power supply internal reference … and not the system ground. Here, true ground is often tied to:
- The wall cabinet casing or
- Wire conduit traceable to a physical earth ground
This earth ground can often be at a different potential than signal common. That means if the signal return is directly connected to ground, current can flow through the signal common line and create a ground loop — picking up unwanted noise.
Of course, fully shielded cable can further boost a design’s power-side integrity. That shield is commonly left to float (unconnected) to serve as a Faraday cage and limit the power inducible in the power lines. But sometimes, EMI is substantial enough to necessitate more than just shielding. Here one solution is to connect the shield drain to earth ground at the cabinet or conduit, which serves as a leak path for any excess energy on the shield to ground. It’s rarely advisable to connect such a shield at both ends because the cable’s equipment end is often at a different potential than the supply end, which means shields connected at both ends can actually experience excess current flow. This is most problematic during lightning storms when ground potential can vary widely as strikes hit the ground near the plant. In cases where a cable assembly is being built in-house, care should be taken to ensure shielding carries all the way through the cable and connects to the connector body — ensuring end-to-end integrity of the Faraday-shield properties.
One final note on maintaining automated feedback-signal quality: Over time, automated systems are often retrofitted and upgraded. Usually, that involves the addition of devices for more complex and sophisticated capabilities. The risk is attaching an excessive number of devices to a single existing power supply … as that can in turn cause voltage droop and missed signals. This shows up as an intermittent problem and looks like a signal dropout due to destructive interference. Overburdened power supplies are fairly common, so during any upgrades, be sure to verify that existing power supplies can handle the load when all devices are active.
Thorough and thoughtful design approaches can yield robust device operation suitable for industrial automation environments. The caveat is that proper installation of sensors and actuators demands careful attention to connection schemes — and prevention of signal-quality degradation by EMI. Making final connections with high-quality cable and connectors can ensure smooth sailing at the start and for the life of automated machinery.
Disclaimer: The opinions, beliefs, and viewpoints expressed by the various authors and/or forum participants on this website do not necessarily reflect the opinions, beliefs, and viewpoints of Digi-Key Electronics or official policies of Digi-Key Electronics.