You're asking how to detect whether there is an EMI problem, yes? It has to be done with appropriate receivers (spectrum analyzers), and knowledge of the potential interfering transmission characteristics.

See the attached Tektronix application note.

Also, see this link, for a fun look at the problem of detecting EMI on the cheap.

http://electronicdesign.com/components/simple-homemade-sensors-solve-tough-emi-problems

today, many papers focus on EMI modelling. so EMI issue of each systems can be predict and simulate. so EMI issue can be studied easily without measurement setups.

an example of EMI modelling can be find in attached file.

Best regard 

Dear Researcher,

Every telecommunication-service-bearing equipment makes use of some part of the radiofrequency (RF) spectrum which is a natural but scarce national resource. Ineffective management of the spectrum results in electromagnetic interference (EMI).

High Intensity Radiated Emissions (HIRE) from radar, microwave relay stations, radio/TV transmitters and high power AM/FM radio broadcast systems can lead to disruptions in airplane navigation and communication systems and to possible loss of aircraft and human life.

There have been reported cases of adverse effects of EMI on medical devices such as implanted cardiac devices (e.g. pacemakers and defibrillators), apnea monitors, powered wheel-chairs, blood pumps, hearing aids and electronic imaging devices.

these are some of the examples.hope it will be helpful for you.

Regards

Hello,

if talking about Interference aka 'Emissions', each and every device that incorporates switches (whether mechanic or electronic) generates emissions. Whether these are acceptable or not is subject of a large number of standards where measurement setups, methods and acceptance levels are defined.

I do not fully agree with Tohid Rahimi and 'his' IEEE paper: simulation can give you 'hints' on what to expect, but the real world is different: fully specifying a device with respect to emissions is near to impossible. And when you have a 'great' simulation setup, simulation times tend to be inacceptable.

Another important issue is 'Immunity' - whether the device can cope with emissions generated by other devices in the network. In principle, simulations covering aspects of immunity are possible, but the setup is really difficult.

Hope this helps a bit.

Very interesting article reference. Using ferrite in a broadband RF "sniffer" to better define and concentrate the incident normal(angle) RF energy is different. (although, being exposed to "old" is just as interesting as really "new"), I will have to play with that. I use a self-designed air core loop to sniff-out the RF voltage and current hotspots. The radiated hotspots are not a surprise, but the spectral content is and its local relative amplitude as well. For conducted EMI placing the conductor inside the loop yields a good noise ratio so that the  spectrum is fairly free of local RF sources.

I am always skeptical of models. They are always a generalized construction using a finite defined limited number of lumped elements which is  someone else's conception of the situation. Sure you can "refine" the model by more accurately defining the element values, but never reach the recreation of the "real world" present situation. Hence, I use modeling as an indicator of gross (large) aberrant system behavior. Every design is different, every PCB, wiring harness, and enclosure individually defines a new set of EMI conditions.

Good topic. Within power electronics, it is always one of those important issues we must always face and it is one of the ethereal areas which is worrisome, not well understood and design-changing when the EMI regulator rejects you.