EMI and RFI – Some Causes and Cures A broad sweep review of the causes and cures and techniques for reducing RFI/ EMI Note: This article was published in the May 1980 issue of r.f design magazine Author: Howard Walker (KI4VEO) Publisher: Cardiff Publishing Company EMI Defined Technically, EMI is, “any Electrical or Electromagnetic phenomenon - Man Made or natural – causing an undesirable response, performance degradation, or complete malfunction of operational electronic equipment. EMI varies, in time and degree, from a nuisance to complete destruction of mission performance.” The sources of EMI are seemingly unlimited. In fact, any electromechanical, electric, or electronic device is a possible source of EMI. Even mother nature is sometimes the culprit via atmospheric and cosmic disturbances. 3 Major Causes of EMI Natural Sources – Generally due to lightning Galactic – Originates outside the earth’s atmosphere. Caused by solar flares of our sun or other stars millions of miles away. This noise is usually found between 18 and 500 MHz. Man-Made – Due mostly to motors, power lines and transformers, neon signs, vehicle ignition, and medical and industrial equipment. This type mainly occurs below 20 MHz. Man-made EMI sources include equipment whose function is to intentionally generate or radiate electromagnetic signals, and equipment that unintentionally generates electromagnetic energy. Unfortunately, even when electromagnetic energy is intentionally generated and presumed restricted to the fundamental or intended frequency range. It sometimes generates harmonic frequencies that cause interference problems with other equipment. EMI has been a factor since the first electromagnetic wave was propagated intentionally. Also known as Radio Noise, Electrical Noise, and Radio Frequency Interference (RFI), it covers the spectrum DC to light frequencies. Until recently, users and manufacturers of electrical and electronic equipment were able to ignore the EMI generated by their equipment. Now, however, more stringent regulations make manufacturers and users alike responsible for interference caused by their equipment. More stringent regulations have been necessitated, not only by the abundance of equipment crowding the spectrum, but also by the nature of the modern developments. The increased use of semiconductors and integrated circuits, as well as the increased use of plastics in cabinets and housings; instead of metal, have caused a problem with EMI leakage into, as well as, out of equipment. Plastics have many advantages, but like all insulators, they are poor EMI shields; without special conductivity treatment. Although susceptibility to EMI may be reduced by filtering, this must be augmented with good design of housing and cabinets. The housing should act as a two-way shield, either keeping the EMI out; or in the case of an EMI generating device; keeping it in. The ultimate goal of the various EMI specifications is to ensure Electromagnetic Compatibility (EMC) – that equipment and systems will function as designed without degradation of malfunction in their intended operation electronic environment nor adversely affect the operation of any other equipment or system. The FCC has the responsibility of controlling the generation of any electrical interference that hampers communications. The FCC is only one of the many government agencies involved in regulating policy and controlling EMI. The other primary agencies are: Office of Telecommunications Policy (OTP) Office of Telecommunications Department of Commerce Department of Defense (DOD) and it’s Tri-Service Executive Agencies National Aeronautics and Space Administration (NASA) Department of Transportation (DOT), including it’s Federal Aviation Agency (FAA) Bureau of Radiological Health (BRH) The National Security Agency (NSA) In addition, there are many others such as the National Bureau of Standards and the Environmental Sciences and Service Administration, which play and important role in EMI, but are not responsible for issuing EMI specifications, rules, regulations and policy. Eliminating EMI Equipment redesign is not cost effective and, often times, not EMI effective. Prevention seems to be the most practical procedure, both in terms of economics and results. Grounding, shielding, balanced lines, twisted pairs, and filtering methods are used to prevent or contain EMI. Grounding A proper ground, for interference reduction, should be a zero impedance point which serves as a single reference point. Since no two points are at the same electrical potential, multiple ground connections introduce a “ground loop”, which conducts interference to other parts of the system. Grounding should not be considered a cure-all, however. In some cases it may be necessary to be used with shielding. But, grounding is not shielding. The housing sections may be grounded. But still leak (radiation). Shielding Electric (E) and magnetic (H) fields travel through free space at 3 X 108 meters/second. When these waves strike a conductor, some energy is reflected and the incident wave also sets up eddy currents in the conductor, which attenuate the signal. Thus, a shield reduces the effect of the electromagnetic wave by reflection and energy absorption. For electrically thick material: R= 50 + 10 log10 (PBf) - 1 A = 1.7t (f/PB) .5 Where R = Reflected Wave in db A = Absorbed Wave in db PB = in ohms/cm f = Frequency in MHz t = Thickness in cm A is dependent on frequency, thickness, conductivity and permeability of the shield material. A major problem with shielding is that any discontinuity in the shield can leak EMI. Shielding in plastic cabinets can be achieved using: Conductive Foil – Generally applied inside the cabinet. It usually requires a coating to prevent corrosion. Effectiveness of 55-100 db for 5 mil thickness is achievable. Conductive Coatings – Also applied inside cabinets. Coatings include graphite, silver, nickel and zinc. The first three are applied with an organic solvent carrier. Arc and flame spraying techniques are used to apply zinc by utilizing an electrical current or gas flame to melt zinc or a zinc alloy wire. Droplets of the metal are then blown onto the housing surface through an air jet. Conductive Plastics – Thermoset and Thermoplastic resins are made conductive with the addition of metal flakes or fibers, carbon powder or graphite fibers. Shielding for RFI Shielding is perhaps the most important method of preventing or curing RFI. Maximum effects of a particular type shield are realized when there are no breaks or points of entry (in the shielding). Ventilation Holes – Honeycomb type vent holes are effective in that the small tubing of the honeycomb act as a waveguide below cutoff. Coaxial Cable – Proper termination is a must. The braid should be soldered so that the inner conductor is surrounded by the braid at the termination point. Improperly designed openings in keyboards or ports, and faulty welding can act as waveguide antennas. Above 100 MHz, leakage is serious, as the required slot size becomes smaller. The slots length, not the width, or area, determine its shielding effectiveness. For slot lengths smaller that the interference wavelength an enclosure will shield. However, as the frequency is increased, shielding effectiveness reaches a minimum at the slot resonant frequency. Panels must be in full contact with the body of the housing. A conductive gasket (i.e. braid or graphite impregnated rubber) must be used. In the case of two or more major housing sections, self-tapping screws are wise to use to attain maximum conductivity. Coaxial Cable Provides an effective means of shielding an interconnecting line against interference and reducing radiation from the line. Balanced Lines This method to minimize EMI susceptibility utilized a balun transformer at the source and load. Primary and secondary windings are shielded from each other and the transformers are usually grounded. Balanced lines minimize induced voltages and currents caused by interfering fields and interference caused by differences in grounding potential. Twisted Pairs Twisting the conductors along the length of the cable provides signal cancellation of EMI, but the length of the twist is difficult to control. This method is only effective when the EMI frequency wavelength is shorter than the length of the twist. This method is not recommended for use in circuits with multiple grounds. Filters EMI can enter equipment through power and signal lines, or through the enclosure via radiation. The previously mentioned techniques are employed to minimize radiated interference, but they are ineffective against conducted interference. EMI filters on power or signal lines are then called for. Power line filters, which are usually Low Pass, are designed to pass frequencies up to the low kHz range. Filters used on signal and power lines are classified by their type of frequency discrimination. Low Pass – Pass frequencies from DC to a specific cutoff frequency and attenuate all signals above by an increasing amount. High Pass – Pass signals above a specified cutoff frequency and attenuate all signals below by an increasing amount. Band Pass – Pass all signals in a specified band and attenuate all signals above or below cutoff. Band Reject – Pass all frequencies except those in a specified band. Chebyshev and Elliptic Function Filters for RFI The Chebyshev low-pass, a common type of filter useful for this application consists of alternating series L and shunt C components. The attenuation5 of the filter is specified by L (dB) = 10 log 10 1 / 1 + w e2 T n2 (w) Where w = 2πf (f is the frequency in Hertz) e2 = ripple factor (as a numeric) Tn (w) = Chebyshev polynomial of degrees n And where e2 = 10 ripple (db) / 10 -1 n= total number of L(s) and C(s) {cos (n arc cos (w) 0<= w<=1} Tn (w) = {cosh (n arc cosh w) w>1 } A more complex and more (characteristic) controllable filter, similar in construction to the Chebyshev, is the elliptic Function filter. The shunt C(s) , of the Chebyshev, are replaced by series L(s) and C (s). This has the advantage of providing a sharper roll-off with increasing frequency (past the 3db or half power point). Filter selection involves several factors, the first of which are: Frequency of Operation The filter selected must have within its range the frequencies to be filtered. Working Voltage The voltage level of the application must be within the working voltage specification. Most commercial filters are rated in the 125 – 200 VDC range and a few in the 500-600 VDC range. Insertion Loss To accurately determine the amount of insertion loss, impedances within the circuit must be known. Since this is not normally available, a filter is chosen with as high an insertion loss/volume ratio as possible. Current Rating The DC current load rating should be 1.5 to 3 times the expected application. RF Current The type of filter construction limits the amount of RF Power it can dissipate. Some filters incorporate capacitors that will not withstand much RF current, while others convert the RF current to heat. RF current values in most applications are less than .25 amps. Insulation Resistance This is a function of the dielectric material used in the filter. This parameter determines the leakage current. Dielectric Voltage Test This is a short term test (5 -60 seconds) that stresses the filter’s dielectric materials 2 – 5 times the working voltage level. D.C. Resistance This is a function of the length and diameter of the wire used. Basically, it is the resistance of the wire, and it becomes important only in very high current applications where the voltage drop is a concern. Operating Temperature Range Industry standards are normally -55oC to 125oC for military and -25oC to +85oC for commercial type applications.\ Storage Temperature Storage temperature is sometimes overlooked. Users often store and transport equipment in environments more severe that the equipment encounters in actual operation. Noise Generation Thermal noise is caused by agitation of electrons in resistance. The mean square value of thermal noise (E2) is: E2 = 4RKT x Δf Where K = Boltzmann’s Constant (1.38 x 10-23 joules / oK T = Absolute Temperature (in degrees Kelvin) Δf = Bandwidth (Hz) R = Resistance (in ohms) or real component of Z Calculation of Noise Values Estimation of noise levels in a receiver, due to external noise, requires taking into account two factors: Orientation of the receiving antenna Gain of receiving antenna In general, noise power is proportional to bandwidth. Thus, where Fa = Effective Antenna Noise Factor Then Fa = Pn/KToB = Ta/To Where Pn = Noise power from equivalent lossless antenna (Watts) K= Boltzmann’s Constant To = Reference Temperature (290o Kelvin) B = Effective Receiver Noise Bandwidth (Hz) Ta = Effective antenna temperature with external noise (Kelvin) References View attachment 671809 View attachment 671809 View attachment 671809 Reference Data for Radio Engineers, Radio Noise and Interference, Atmospheric Noise, Howard W. Sams and Company. CCIR Report 322 10th Plenary Assembly, Geneva; 1963 “EMI Filters”, Hopkins Engineering Co., San Fernando, Calif. “AMP Quiet Line Filter Handbook”, AMP Incorporated, Harrisburg. Pa. The Radio Amateur’s Handbook, American Radio Relay League, Newington, Ct. 1979
Natural Sources you said lightning but: The biggest problem at my location is high Static caused by dry air, and high winds with dust. With over two miles of antenna wire in the air for the Rhombic's, you would not believe the static charges!!!! And the dramatic Discharges
I would! Even my modest wire antennas would build up charges like a lawnmower magneto. I mitigated a LOT of it by getting rid of the Kynar insulated wire (it was free) and going to bare copper (I had to pay for it).
Howard-- A nice and accessible summary of the issues. Funny how they haven't changed in 40 years! 73 Chip W1YW
Man, I would believe it. I am one of the few people to ever experience 'dust lightning'--bona fide lightning cause by a moving sand (storm) layer. Happened near the WAVE in Coyote Buttes AZ. My son also witnessed. We were looking right at the dust being swirled up when the discharge occurred. Bolt was about 50 feet long and happened in blue sky , from the ground to the top of a butte. Of course, lightning also happens in turbulent volcanic plumes. We tend to associate lightning as limited to rain. Not always true.
I was an electronics Technician in the Air Force, for 8 years. Honorable discharge as a Technical Sgt. (304x4 Ground Radio Equipment Repairman). I had to deal with all sorts of EMI and RFI. Mainly Transmitters, in adjacent racks. 770TH OLAA Independent Hill, Va., and the 231st Combat Comm Squadron, Andrews AFB, Md A saying we learned (it applies to all branches of the militray) We, the willing. Led by the unknowing. Are doing the impossible, for the ungrateful. We have done so much, for so long, with so little . We are now qualified to do ANYTHING with NOTHING
Mother Nature can be a Bitch, huh? We have had lightning during a snow storm, here in Virginia. (73 - Howard)
I suggest you include Plasma Screen TV's and Grow lights as big causes of RFI on all HF bands from 1.8 to 21 mhz .
I know that works. Had a neighbor with a Plasma TV. Often, fish tank heaters cause problems. In today's world of electronics, there are many causes. Security lights; wall wart power supplies; cheaply made switching power supplies in computers; bad pole transformers in use by the power company. Anyone have someting to add to this list?
