•  
• EDN Europe MAGAZINE
•  
• Free SUBSCRIPTION
• Subscription Renewal
• View Media Kit

• EDN
• EDN Australia
• EDN Asia
• EDN China
• EDN Japan

Test & Measurement World: ANECHOIC CHAMBERS RISE FROM THE PITS

THE DESIGN, CONSTRUCTION, AND QUALIFICATION OF A CHAMBER DEPEND ON THE PRODUCTS THAT IT WILL TEST.

Tests for electromagnetic immunity and emissions work best when located far from intentional transmitters such as broadcast stations and cell towers. Testing products for emissions or immunity in populated areas requires an anechoic or semianechoic chamber to keep ambient signals from interfering with measurements. Every chamber installation is unique and is influenced by many factors, some of which can change over time.

Major factors affecting chamber design include the size of the EUT (equipment under test), the relevant standards for EMC (electromagnetic compatibility), and the building facilities. A company that manufactures handheld devices won’t need a chamber as large as one that makes rack-mounted equipment or vehicles. Chambers also differ depending on whether they must comply with commercial or with military standards. Chambers exclusively for emissions tests are different from those that also host immunity tests. Chambers for precompliance tests and troubleshooting may also differ in size and construction from those for full compliance tests. The sidebar “What is an anechoic chamber?” explains how chamber components minimize signals inside the room.

KNOW WHAT GOES IN

Proper chamber design starts with a knowledge of the products that you need to test. “Have a good idea of your product road map over the next several years because test requirements tend to grow over time,” advises Bryan Sayler, senior vice president and general manager of ETS-Lindgren. Planning for future products eases decisions about a chamber’s size, turntable size, cable routing, power provisioning, and instrumentation.

National and international EMC standards significantly influence chamber size because they often specify the distance from the antenna to the EUT: 1, 3, or 10m. CISPR 22, for example, requires a 10m distance for EUTs larger than 2m³ (Reference 1). Product size affects chamber size because a chamber must be large enough to accommodate the distance from the EUT to the antenna, plus the size of the EUT, plus the size of any internal absorber materials. Absorbers such as cones that line a chamber’s walls significantly reduce the usable area.

Chambers may also accommodate other distances, such as 5m. “Some test labs will use a 5m distance for precompliance tests because it better correlates to how an EUT will perform at 10m than a 3m distance,” notes Peggy Girard, president of Panashield. Some chambers accommodate a 5m path, but the FCC (Federal Communications Commission) currently accepts only the chamber based on the 3m test data because the agency doesn’t formally recognize 5m data in its comparison with an OATS (open area test site).

The construction of the host building also affects chamber size, particularly height. When engineers at Northwest EMC opened a facility in June 2009 in Brooklyn Park, Minnesota, they wanted a location close to potential customers, which ruled out an OATS. Tim O’Shea, operations manager at the company, explains that ceiling height was a critical factor because many cities have building- height restrictions, and he needed a building with ceilings high enough to house a 10m chamber.

RISING FROM THE PIT

Height isn’t the only consideration for a building to house an anechoic chamber. Depth counts, too. Figure 1 shows a typical chamber with a raised working floor. The chamber floor is supported on legs because there’s a pit under the floor. The pit, typically 45 to 60 cm deep, provides space for EUT-support equipment and cable routing from the control room to the center of the floor’s turntable. The raised floor is flush with the building floor outside the chamber, which lets technicians roll equipment into the chamber.

Girard describes an unusual chamber that, because of low ceilings in the building that housed it, required a “pit within a pit.” Figure 2 shows that the chamber’s raised floor was actually 1.8m below the building’s floor. The 1.8m pit was larger than the chamber, providing working space outside the chamber at the chamber floor’s level. A second pit extended 45 cm below the chamber’s floor to provide space for cables and turntable motors.

Pits under chambers need not be as large as the entire chamber. To save on construction costs, some companies and test labs choose to leave much of the area under the chamber floor solid and provide just conduits for cables to the EUT, turntable, and antenna. This design costs less to excavate; however, it minimizes flexibility. Chambers with conduits also need metal pipes around the cables for electromagnetic shielding.

Instead of having just the 45 to 60 cm space under their raised floors, some chambers have deeper pits under the turntable or antenna. Ghery Pettit, EMC regulatory compliance manager at Intel, chose to construct a chamber with a 1.8m pit under the antenna. That depth provides space for RF amplifiers used in immunity tests. Keeping amplifiers close to an antenna results in shorter cables that carry RF signals. “We really wanted a 2.1m pit because 1.8m isn’t enough for everyone to stand upright,” says Pettit. “When we built a chamber in Dupont, Washington, we went down 2.1m so that even the tallest of our people could stand comfortably in it.”

Other chambers have no pits under them at all. Instead, the base of the chamber rests on the building floor, and the chamber has a raised floor under which the cables run. In that case, an equipment ramp lets technicians roll equipment into the chamber (Figure 3).

Cables under a chamber floor provide power, control, and I/O signals to an EUT. RF cables connect antennae in the chamber to RF amplifiers for immunity tests or to EMI (electromagnetic-interference) receivers and spectrum analyzers for emissions tests. Other cables provide power and control signals for turntables and antenna masts. Test equipment usually resides in a shielded control room close to the chamber.

CONNECTING THE CABLES

Because test equipment resides in a separate control room, all chambers need penetration panels that hold cable connectors. Cables on either side of the panels connect equipment in the control room to the EUT, turntable, and antenna inside the chamber. The control room should be close to the portion of the chamber that holds the panels to minimize cable length.

For commercial EMC tests, cables that come from outside the chamber generally run under the floor. The penetration panels may be part of a chamber’s walls, or they can be on the floor next to the chamber, provided that there’s space under the panel for cable runs. At Northwest EMC’s Brooklyn Park facility, penetration panels are in the chamber’s wall next to the control room but below the raised floor’s surface (Figure 4). At Intertek’s facility in Boxborough, Massachusetts, connector panels are on the floor of the control room (Figure 5). Cables then run through the pit under the floor to the center of the turntable. The raised chamber floor has panels through which the antenna cables can emerge to reach the antenna and its mast.

For military tests, cables must run on the chamber floor because that arrangement simulates actual use. In addition, EUTs need not rotate. Tom Arcati, an engineering specialist at Dayton T Brown, thus uses chambers without pits or turntables (Figure 6). That setup simplifies the chamber’s design and makes it less expensive than chambers for testing to commercial standards. The penetration panel is on the chamber wall, slightly above the floor. “MIL-STD [Military Standard] 461 requires at least 10m of cable between the antenna and EUT,” says Arcati. “Power cables from the LISN [line-impedance-stabilization network] and the equipment must be at least 2m long.”

EUTs often need external ac or dc power, so a chamber must be wired to make power available. The ac or dc power that a chamber must support varies with how the chamber is used. Independent EMC test labs test a wide range of powered products and require more forms of power than do chambers dedicated to one company. The type of power also depends on national compliance standards. For example, a chamber might need 100V for Japan, 110V for Taiwan, 120V for North America, and 220V for Europe. It may need to provide single- or three-phase power frequencies of 50, 60, or 440 Hz.

The ac mains lines that enter a chamber require filtering to minimize noise. Some chambers filter their ac voltages using a single filter for each voltage and then distribute the clean power to the equipment inside the chamber. Some EMC engineers, however, prefer to distribute the power to its load and filter it there. The choice comes down to cost versus flexibility.

Because power and signal cables must connect to an EUT on a turntable, the cables must have enough slack to accommodate turntable rotations. “Most turntables have 380° of rotation,” says ETS-Lindgren’s Sayler. “The turntable typically rotates ±180° during a test.” A more expensive approach and one that he rarely sees uses slip rings that conduct power and signals while permitting continuous rotation.

DESIGNING THE DOORS

Regardless of whether a chamber is for commercial or military tests, it needs a door large enough to get an EUT in and out. The size and design of a door also depend on the products tested in the chamber. Size and location are important. “The door locations should minimize the distance to the turntable if the chamber will test large, heavy EUTs,” notes Intel’s Pettit.

Door location matters for more than just convenience. “There are electrically good and bad places for chamber doors,” Sayler warns. “If the door lacks cones, it should be as far away from the turntable as possible because it will affect the quiet zone around the turntable.”

Engineers at Northwest EMC use their chamber for radiated emissions only, and the EUTs are typically small enough to fit onto a table. Thus, the chamber has two 1.2x2.4m swinging doors for equipment entry and exit. The insides of the doors don’t have absorbing cones because they would interfere with the door opening in the chamber wall. The door has ferrite tiles lining its inside surface. Figure 3 shows a single door lined with ferrite tiles, but no cones, so the door can clear the door opening. In contrast, the chamber at Intertek’s Boxborough facility has a retractable door and thus has cones but costs considerably more than a chamber with a swinging door. Swinging door panels can have cones over the ferrites provided that the cones can clear the door opening (Reference 2).

QUALIFYING A CHAMBER

Because anechoic chambers are shielded rooms, they must prove that they sufficiently attenuate outside signals before a manufacturer installs internal absorbers. Most EMC engineers specify that chambers attenuate outside signals by at least 100 dB. “We can build chambers that attenuate up to 120 dB for areas high in ambient signals,” says Sayler. “For most applications, 80 dB of attenuation is enough.” The ambient noise levels inside the chamber are those that count. If a chamber is in a region of relatively low ambient signals, then 80 dB of attenuation may be sufficient.

A technician tests the chamber’s shielding effectiveness by generating a signal at a known power and frequency and measuring signal strength on the other side of the wall. O’Shea explains the test procedure at Northwest EMC. “We had the transmit antenna at 26 locations outside the room—including the top— and moved the receive antenna along all inside seams closest to the transmit position,” he says. “We checked every seam in the room.”

Technicians may also perform shielding- effectiveness tests with the transmitting antenna outside the chamber and the receiving antenna inside. Placing the receiving antenna in the chamber reduces ambient signals at the receiver, which makes the transmitted signals easier to see on a spectrum analyzer.

Following a shielding-effectiveness test, a chamber is ready for the absorbing materials. After installation, the chamber needs additional measurements around the turntable to verify its quiet zone for radiated emissions tests per ANSI C63.4 from 30 MHz to 1 GHz and per CISPR 16-1-4 (references 3 and 4). For radiated immunity tests, a chamber must undergo a field-uniformity test to comply with EN 61000-4-3 (Reference 5).

DO THIS, NOT THAT

When building an anechoic chamber, Panashield’s Girard urges an understanding of your needs. “Start by knowing what your present and future testing needs will be for your product, according to international standards,” she says. “Will you require radiated emissions, immunity, or both?”

“Plan ahead. Look at the whole facility and how the chambers will fit into the building,” says O’Shea of Northwest EMC. “Evaluate customer needs to customize room size, turntable size, door size, and power requirements. Don’t take shortcuts in the building process or during shielding-effectiveness testing because they could cause problems that are much more difficult to fix once the room is fully constructed.”

“Establish a good relationship with your facilities-management colleagues,” says Sayler of ETS-Lindgren. “You’ll need them.”

And Intel’s Pettit expresses concern for ac mains power. “Have plenty of power, at least 100A service for each voltage,” he says. “Separate power feeds from the turntable to the EUT so you won’t get interference.”