When my son was working on his EE degree he found an article titled: Stop That Noise in the IEEE Potentials, October – November 1995. Being a dutiful son, he brought it home to his EMC-dad and I read it. It is a great little article giving an overview of the cause and effect of EMC problems. The articles primary thrust is to introduce EMC to college students and encourage them to become EMC engineers. It does a superb job with that, which is really good; because the field needs some newbie’s to replace the retirees
The article contains a list of valuable attributes for becoming a successful EMC engineer. It says, to solve electromagnetic compatibility problems, a person needs:
• Common sense,
• The right education,
• Some experience,
• A logical diagnostic procedure,
• Good observation skills,
• A sense of humor,
• Persistence, and an
• Occasional bit of luck.
Doesn’t that sound like the attributes of all successful people?
The article goes on to say that the right education for EMC means a good understanding of:
• Basic circuit theory,
• Electromagnetic theory,
• System concepts,
• Electrical modeling,
• Electronic circuits,
• Transmission lines, and a
• Little antenna theory.
The article was written by the gang from U of Missouri, Rolla: Van Doren, Hubing, Drewniak, Fei Sha, and Hockanson. I don’t know all these guys, but the ones I know certainly live up to the published list of requirements and they are all very willing to share their experiences and expertise.
I might add to their list of attributes, a comment I made to my boss when he asked me how I knew that the solution I proposed would work. I told him all that was needed to develop a good EMC solution was to think like an electron, i.e. what would I do if I were one.
EMC is one of those fields (if you pardon the pun) that subtends all disciplines so to the above education list could also be added thermodynamics/heat transfer, shock and vibration, and electrochemistry. The reasoning behind adding these three design areas is that they affect the EMC behavior of a system.
Heat affects the operation of semiconductor materials — some more than others. Adding shielding to solve an RF problem creates thermal barriers which then require some method of getting the heat out. If the heat removal is convective and depends on the movement of ambient air then ventilation openings are required. Vent openings are not RF tight so we still have a problem.
Vibration changes the separation distance between components, PCB’s and subassemblies at the vibration frequency – or in the case of shock at their natural mechanical resonance frequency. The movement changes the capacitive and inductive coupling resulting in vibration induced modulation. In the vacuum tube days this was called microphonics.
Corrosion is an electrochemical effect in which the more reactive material is consumed creating non-conductive and/or nonlinear junctions where conductors ought to be. If mixing occurs, a non-linear junction can generate frequencies that aren’t part of the systems operation. Non-conductive junctions destroy the bonding between shield panels. Given enough time, corrosion can eat through the panels. Corrosion is a major EMC problem, especially for systems located in harsh environments or requiring long term reliability.
All of these things create potential EMC problems which should be considered during the systems/equipment design. These design efforts can be worked into the schedule during the off times when the EMC engineer is not trying to outrun speeding bullets, stopping powerful locomotives, and leaping over tall buildings. But, they absolutely have to be done before PDR, i.e. the preliminary design review.
– Ron Brewer
