Tutorial: Running down interference to improve device performance

10 June 2009

EMI and RFI can ruin your sensor data, and trying to trace sources can drive you nuts. Peter Welander, process industries editor of Control Engineering’s Process Instrumentation & Sensors Monthly offers some solutions.

Electromagnetic and radio frequency interference (EMI and RFI) can cause your sensor data to go nuts, and finding its sources can make you do the same. The two types of interference can present similar symptoms, but the nature of their sources and solutions are not the same. Trying to define the differences between the two is a little vague and many people use the two terms interchangeably for all practical purposes.

EMI is broadband electrical noise that typically emanates from wiring and equipment. In any situation where there is heavy inductive load switching, electric motors, and particularly complex devices like VFDs (variable frequency drives), you’ll get EMI. In the same way it hurts your radio reception, noise is transferred to your sensors through inductive and capacitive coupling, often between wires. You can get it from having data cables too close to ac wiring in cable trays, for example, which means laying a thermocouple extension wire next to a 440 Vac line is asking for trouble.

RFI is higher frequency, and generally the tougher problem to fix. While EMI usually moves from cable to cable, RFI can be picked up all sorts of ways. It often comes from radio transmitters such as walkie-talkies, cell phones, CB radios, etc. (A driver triggering a CB radio from his truck while driving past my house nearly blew out my stereo speakers.) You’ve probably heard the characteristic beeping on the phone caused by someone’s Blackberry trying to exchange data. There are anecdotes of radios on barges sending systems haywire as they pass plants on the shore.

The thing that makes such interference hard to eliminate is that it can be intermittent. For example, a welder working in the plant makes the temperature sensors on a vessel change, but then they recover when the welding stops. The potential for damage to the process depends on the EMI severity and how well you’ve deployed suppression techniques. Problems that run all the time, more or less, can be tracked down and fixed. The TC wire and ac cable in the same tray are an example.

Many suppression techniques for both kinds of interference exist:
 Keeping critical cables separated;
 Use of twisted pair and shielded cables;
 Filtering capacitors at strategic points;
 Careful grounding where appropriate;
 Use of signal isolators;
 Appropriate case and enclosure designs;
 Selection of devices that are designed to reject interference; and
 Many other possibilities.

Going into depth in this context is not practical, but there are application notes from several sources that can help you analyse and solve your interference problems. Below, three links will get you started.


More information...

Contact Details and Archive...

Print this page | E-mail this page