Large companies can afford a $10 to $30 million facility necessary for EMC qualification testing plus the additional expenses for the personnel to staff the operation. It could be argued that they can’t afford not to provide such facilities. On the other hand, small and medium-sized companies typically cannot justify their own fully equipped EMC qualification test facility…yet there still is a need to perform design, development, and troubleshooting tests.
These tests should be done early and often during a program to minimize overall costs to meet the formal EMC compliance requirements. It’s not necessary to test everything with every procedure every time the sample is tested, only those that may be a problem.
It is not cost-effective to go to an outside lab with a regimen of a formal test-redesign-test again process. Doing this means that the system has to be nearly complete. At that time in the design cycle, it is very difficult to make changes. In fact, that approach is so time-consuming and expensive that management may chalk it off as unnecessary.
Formal Tests
There always has been limited RF spectrum available for communications. In the early 1900s, this was largely because equipment was not able to access higher frequencies. As new equipment was developed, more of the spectrum was consumed. However, some frequencies are more popular and economically important than others, and pile-ups of emitters occurred.
This resulted in government intervention and the creation of RF control through the International Telecommunications Union (ITU) formed in 1932. The ITU is the principle organization regulating the use of the RF spectrum. There have been many new types of intentional and unintentional RF devices invented since 1932. These are all expected to coexist in RF harmony.
MIL-STD-461A Notice 4 (1971) probably summarizes the reasons for government EMC requirements best of all:
“The requirements specified in this standard are established to:
(a) Insure that interference-free design is considered and incorporated into the equipment, subsystem, or system.
(b) Enable compatible operation of the equipment, subsystem, or system in a complex electromagnetic environment.”
To assure compatible operation, both emissions and susceptibility requirements were imposed on the equipment. This standard became the foundation for both military and commercial EMC requirements. The EU EMC Directive 89/392/EEC released in 1989 was the first major commercial EMC standard requiring equipment to have “an adequate level of intrinsic immunity of EMC disturbance to enable it to operate as intended.” This EU standard has become the de facto world commercial EMC standard.
Both MIL-STD-461 and the EU EMC requirements have changed considerably since their release, and they are always changing. But there are similarities (Table 1). The procedure numbers and abbreviations are from MIL-STD-461F: Conducted (C), Radiated (R), Emissions (E), and Susceptibility (S).
Table 1. Emissions and Susceptibility TestsBasic Equivalence of Military, EU, and FCC Requirements
MIL-STD-462 Notice 6 (USAF) 15 Oct 1987
In the EU documents, susceptibility is called immunity. Although the limits are different, the procedures are so similar that there could easily be a universal test standard. During design and development, it’s not practical to set up a formal EMC test program every time progress needs to be evaluated, and troubleshooting requires near real-time feedback.
Test Equipment and Facilities
EMC tests can be categorized as conducted emissions (CE), radiated emissions (RE), conducted susceptibility/immunity (CS/I), and radiated susceptibility/immunity (RS/I) which can be further grouped by frequency: (a) DC to 1 GHz, (b) 1 GHz to 18 GHz, and (c) 18 GHz and higher. All companies need to test frequency range (a). Companies building equipment that falls within frequency ranges (b) or (c) are stuck with it.
From a historical perspective, organizations should purchase equipment capable of performing tests for their EMC problem areas. The high cost may make it a buy or rent decision. Fortunately, most EMC problems, especially for non-RF equipment, fall within the frequency range of DC to 1 GHz, primarily because most cable resonances occur in this range, and cables act as antennas.
Suggested Test Equipment
Following is a list of suggested EMC test equipment for frequencies to 1 GHz.
• Receiver/Spectrum Analyzer (10 kHz to 1 GHz): Used for emissions measurements and probably the most important item on the list. Conducted and radiated emissions measurements are required by all EMC specifications. Because emissions and susceptibility fixes are reciprocal, what is done to reduce the emissions also helps with susceptibility/immunity. There are a number of low-cost spectrum analyzers that cover this frequency range. If possible, get one with a built-in tracking generator.
• ESD Generator (multipurpose model): Used for both conducted and radiated susceptibility measurements and the second most important item on the list. Conducted and radiated susceptibility/immunity measurements are required by all EMC specifications except the FCC. In addition, an ESD test is included in the EU EMC requirements.
A multipurpose ESD generator with transient voltage adjustment will cost about half the price of a good used spectrum analyzer. But if there are major budget constraints, an uncalibrated ESD source with an output of 8 to 12 kV (not adjustable) is available for about $29 at the local hardware store. It’s called a piezoelectric gas grill lighter.
• Low-Noise Amplifier (LNA) if using a spectrum analyzer: EMC receivers have about a 7-dB noise figure or less and do not require an LNA. Many low-cost spectrum analyzers lack the necessary sensitivity to make measurements without using a broadband pre-amplifier. The approximate requirements are frequency range 1 to 1,000 MHz, 25 dB ±1 dB gain, 2.5-dB NF, and 100-mW output (PR2).
• Clamp-On Current Probe (10 kHz to 100+ MHz): This device is a one-turn to many-turn transformer which allows measurement of intentional and unintentional RF signal currents on power, control, and signal leads. It’s great for troubleshooting because it can be clamped around a wire without causing damage. Get a wide frequency coverage because if the current is on a wire, it will radiate. Just because the spec stops at 30 MHz doesn’t mean the measurement has to stop there as well. With the spectrum analyzer, it’s used to measure amplitude vs. frequency and with an oscilloscope to measure the waveform.
• Assorted Magnetic Field (HF) and Electric Field (EF) Probes: These small loops or small wire antennas allow close-in pickup of RF from cables, connectors, cabinets, or PCBs. These can be a simple round loop made from 10 turns of wire wrapped around a finger, a small square loop made from stiff wire, or a 1-inch piece of wire soldered to the center pin of a female BNC coupler. When probing around in an electrically hot box, they should be insulated. They also can be purchased in their own velvet-lined box complete with an LNA.
• Line Impedance Stabilization Network (50-?, one for each power line): An LISN is a voltage probe for the power supply leads. It also provides power source isolation and impedance stabilization. It essentially is a Pi filter network designed with a 50-? characteristic impedance containing a sensing resistor. The spectrum analyzer is connected to it to measure the noise voltage on the power leads.
At the present time, all EMC specifications are using LISNs. Safety is an issue here. LISNs are mounted on ground planes, and live power is frequently exposed on LISN terminals. Be certain that all safety precautions are taken.
• Biconical Antenna: Usable from 10 to 500 MHz. For higher frequencies, add a log periodic antenna ranging from 300 to 1,000 MHz. These are self-explanatory and used for both emissions and susceptibility measurements. It’s also necessary to locate and support them at specified distances from the test sample. This can be a wooden tripod or a nonmetallic antenna stand.
A number of unique items on this list are difficult to find on the used market and can be do-it-yourself parts if the budget is really tight. These include the HF, EF, and current probes that are really easy; the LISN is not that difficult.
It’s also possible to build the antennas, which are not quite as easy. Construction information for the Robert’s dipoles (the set of four covers 30 to 1,000 MHz) is available from the FCC, and the biconical dipole plans are shown in MIL-STD-461A. Log periodic antennas can be purchased at ham radio supply stores.
The SAE ARP 958 two-antenna calibration is the easiest method. Of course, this means it’s necessary to have two of each, a small price to pay compared with the alternative.
There is some other ancillary equipment that would be very helpful:
• Laboratory DC Supply: Get one that is RF quiet or uses batteries to supply power for the test.
• Oscilloscope (100 to 150 MHz minimum).
• EUT Test Stand (30″ x 60″ x 32″): A box or table on which to place the test sample. It may require a ground plane on the top depending on whether the test sample is stationary or portable. If testing to MIL-STD-461, the stand must measure 30″ x 96″ x 32″.
• Ground Plane (30″ x 60″, copper or aluminum): This also could be 10′ x 10′ of 0.5″ hardware cloth. The larger ground plane would be used for ESD and bigger test samples. There are minimum thicknesses in the spec, but for design, development, and troubleshooting, foil will work.
• DMM with LCR measurement capabilities.
• Power Line Filters: Used to isolate the incoming power from the test sample. They help to reduce RF noise from the power source and protect the source from testing transients.
• Isolation Transformer: Used in conjunction with the power line filters to reduce common-mode currents.
• Parallel Plate Line: A simple test cell such as a parallel plate capacitor, great for testing PCBs or other circuits or subassemblies smaller than 20% of the volume. Very efficient, a 50-? parallel plate line with 0.5-meter spacing will produce 10 V/m at approximately a watt. Its usable frequency is determined by its length. It can be as simple as aluminum foil on a modified cardboard box. See MIL-STD-461A for construction ideas, but don’t duplicate that design. Be creative; there are better ways.
• Miscellaneous Items: Coax cable; aluminum foil; 50-? terminations; a 3- to 2-wire AC adapter used to decouple the green wire ground (unfortunately it creates a possible shock hazard); and assorted ferrites, capacitors, and shielding parts.
With the new compact equipment available today, a self-contained EMC troubleshooting kit with a spectrum analyzer, an ESD gun, and all the trimmings can fit into a single container.
Test Facilities
A shielded enclosure would be nice, but that is a big commitment. For emissions testing, a low RF ambient is needed. This may be available after hours, late at night, or on holidays, or some frequencies may just have to be skipped. A ground plane is needed, and for troubleshooting, it can be a metal desk or workbench, heavy-duty foil, hardware cloth, or some other conducting surface.
Depending on the size of the sample, the testing may take place in the lab, a conference room, the loading dock, or out in the parking lot. If the testing is done outside, the sample may have to be shielded from the sun, and outside air temperature could be a problem.
Conducted susceptibility testing does not require much in the way of specialized facilities. Radiated susceptibly is a different matter. For this, it is necessary to protect the environment from the evils of the test, so some form of shielding is needed. This could be a metal warehouse building or a simple chicken-wire enclosure.
Check for a radio station or cell phone signal outside and then inside the area. If there’s plenty of signal outside but it’s impossible to receive the station or make cell phone calls inside, then RF signals are being attenuated by the structure. In any case, don’t just park on a given frequency during the testing.
Part 2
Part 2 of this article will address design/development testing and troubleshooting of conducted emissions, conducted susceptibility, radiated emissions, and radiated susceptibility EMC problems. It will appear in EE’s September issue.
About the Author
Ron Brewer currently is a senior EMC/RF engineering analyst with Analex at the NASA Kennedy Space Center. The NARTE-certified EMC/ESD engineer has worked full-time in the EMC field for more than 30 years. Mr. Brewer was named Distinguished Lecturer by the IEEE EMC Society and has taught more than 385 EMC technical short courses in 29 countries and published numerous papers on EMC/ESD and shielding design. He completed undergraduate and graduate work in engineering science and physics at the University of Michigan. e-mail: [email protected]