EMC Chambers: 3 m, 5 m and 10 m Method, Semi and Fully Anechoic
An EMC chamber is a shielded box with a controlled reflection problem. The shielding is the easy half, and most suppliers can do it. The half that decides whether your chamber passes NSA, field uniformity and sVSWR on the acceptance day is the absorber layout: which walls get ferrite only, where the hybrid absorber starts, and how the floor is treated for the test distance you actually need.
What we build
| Type | Typical outer size | Quiet zone | Usual use |
|---|---|---|---|
| 3 m semi-anechoic | 9.0 × 6.0 × 6.0 m | ø2.0 m × 2.0 m | Radiated emissions and immunity to CISPR 22 / EN 55032, EN 61000-4-3 |
| 3 m fully anechoic | 7.0 × 4.0 × 3.0 m and up | ø1.0–1.5 m | Antenna, radar, wireless; automotive radar bands |
| 5 m semi-anechoic | project specific | ø3–5 m | Between 3 m and 10 m requirements |
| 10 m semi-anechoic | project specific | ø7.0 m × 2.0 m | Full compliance to CISPR 16-1-4 class 10 m, CNAS accredited labs |
The 3 m semi-anechoic chamber, as we actually quote it
The reference design below comes from a 64-page technical proposal we issued in December 2021 for a 3 m semi-anechoic chamber with shielded control room and amplifier room. These are the real numbers from the bill of materials, not a generic catalogue.
Geometry, dimension by dimension
| Measure | L | W | H | Note |
|---|---|---|---|---|
| Outer envelope | 9.00 m | 6.00 m | 6.00 m | including 0.30 m raised floor for the reflecting ground plane |
| Inside the ferrite | 8.80 m | 5.80 m | 5.60 m | reflecting floor not counted |
| Tip to tip, absorber | 7.80 m | 4.80 m | 5.30 m | from the underside of the absorber down to the raised floor surface |
| Quiet zone | ø2.0 m × 2.0 m, concentric with the turntable, on the chamber long axis | the volume the performance figures are guaranteed over | ||
Those three rows of numbers matter because they add up. If a supplier quotes a 9 × 6 × 6 m room with an 8 m quiet zone it means the quiet zone is measured tip to tip, not ferrite face to ferrite face. Always ask which one.
Shielding effectiveness, tested before the absorber goes in
Per EN 50147-1:1996, measured after the shield is up and all filters are installed, before a single piece of absorber is hung. Then measured again with everything, turntable included, in place.
| Field | Frequency | Target |
|---|---|---|
| Magnetic | 14 kHz | ≥ 70 dB |
| Magnetic | 100 kHz | ≥ 100 dB |
| Plane wave | 100 MHz | ≥ 100 dB |
| Microwave | 1 GHz, 2 GHz, 3 GHz | ≥ 100 dB each |
The 14 kHz magnetic figure is the one that trips up thin-panel rooms. Galvanised steel at 2 mm is close to transparent to a low frequency magnetic field: the attenuation comes from the panel getting thick enough and every seam being a continuous electrical joint. That is why the standard module uses double-sided hot dip galvanised sheet at ≥ 2 mm with a zinc layer of ≥ 20 µm (≥ 275 g/m²), folded twice on all four edges with a 40 mm return, and bolted with M10 flange bolts at 100 mm pitch through a conductive mesh gasket.
Site validation, the three tests that decide acceptance
| Test | Standard | Frequency | Requirement on this project |
|---|---|---|---|
| NSA, normalised site attenuation | CISPR 16-1-4 / ANSI C63.4 / EN 50147-2 | 30 MHz – 1 GHz | deviation within ±4 dB at 3 m |
| FU, field uniformity | IEC 61000-4-3 | 80 MHz – 1 GHz | 75 % of 16 points within 0 / +6 dB |
| sVSWR, site voltage standing wave ratio | CISPR 16-1-4 / CISPR 16-2-3 | 1 – 18 GHz | ≤ 6.0 dB at 3 m |
How the NSA test is actually run matters as much as the number. Pairs of positions: transmit antenna at 1.0 m, 2.0 m and 2.5 m for horizontal polarisation, 1.0 m, 1.5 m and 2.5 m for vertical, five transmit locations (front, left, right, centre, rear), receive antenna scanned from 1.0 m to 4.0 m. Sixteen field uniformity points sit on a 1.5 × 1.5 m vertical plane 0.8 m to 2.3 m above the turntable. sVSWR is measured with absorber on the floor at 300 mm height.
Background noise
CISPR 22 / EN 55022 Class B at 3 m, measured with no EUT but turntable rotating, antenna tower travelling and all lighting on: at least 10 dB below the Class B limit from 9 kHz to 1 GHz. For military work to GJB 151B, all auxiliary equipment running, the conducted emission from the supply used for test equipment better than the CE 102 limit by 6 dB, and chamber background better than RE 102 by 6 dB.
Field strength the chamber has to survive
Continuous 30 V/m and pulsed 200 V/m across 14 kHz to 3 GHz, with every electrical and electronic device inside still working. This is the immunity test load, and it is a structural and thermal requirement as much as an electrical one.
The 10 m chamber
Same frequency chain, bigger volume, tighter tolerance. From a 10 m project specification issued September 2018:
| Parameter | Value |
|---|---|
| Shielding effectiveness, 14 kHz – 18 GHz | ≥ 100 dB to 1 MHz; ≥ 110 dB from 14 kHz to 18 GHz |
| NSA, 30 MHz – 1 GHz | ±3.5 dB over a ø7 m × 2 m quiet zone at 10 m; the standard allows ±4 dB |
| Field uniformity, 80 MHz – 18 GHz | 75 % of 16 points within 0 / +6 dB |
| sVSWR, 1 – 18 GHz | ≤ 6.0 dB at 10 m test distance |
| Quiet zone | ø7 m × 2 m |
| Grounding | equipotential copper bar, single point per room, < 1 Ω |
| Stability requirement | performance held for more than 12 years |
The ±3.5 dB against a ±4 dB standard is not marketing. On a 10 m path the free space loss is larger, the reflections come back at shallower angles, and the absorber has more wall area to cover. Holding 3.5 dB means the corners, the ceiling-to-wall junction and the floor transition all have to be right, not just the flat panels.
Absorber
| Material | Where it goes | Why |
|---|---|---|
| Ferrite tile HCA-FT6.7, 100 × 100 × 6.7 mm | All four walls, ceiling, and the shield door, floor to ceiling | Handles the low end, roughly 30 MHz to 300 MHz, where pyramidal absorber is too thin to work |
| Hybrid pyramidal HBA30 / HBA50 | Over the ferrite at the higher bands, laid out per the simulation | Ferrite alone stops absorbing above a few hundred MHz; the pyramid takes over |
| Mobile absorber modules | Floor, used for field uniformity tests | Floor treatment changes between emissions and immunity, so it has to move |
| Reflectivity of the 1–40 GHz range | — | ≤ −30 dB at 1–3 GHz, ≤ −35 dB at 3–10 GHz, ≤ −40 dB at 10–40 GHz per GJB 151B |
In the reference build the ferrite is Korean HCA original product, mounted on screws so it can be taken down and re-used if the chamber moves. Absorber is specified as waterproof, moisture proof and fire rated, with a 30 year service life claim.
What is included in a turnkey scope
This is the actual equipment list from the 3 m proposal, so you can see the level of detail we work at and compare it against a competitor's quote line by line.
| Item | Specification |
|---|---|
| Shield body | Self-supporting steel structure, independent of the host building, no load taken by the building. Pan-type double-sided galvanised panels. Exterior powder coated RAL 9003. |
| Chamber door | 1.20 m × 2.10 m, pneumatic locking, manual swing. Emergency light inside above the door, test-in-progress lamp outside. |
| Turntable | TOP-TT-2/1, 2 m diameter, 1.5 t load |
| Antenna tower | TOP-AT-4, 1–4 m scan |
| Controller | TOP-MCU multi-channel, drives turntable and tower together |
| Power filters | single phase 230 V AC/DC 32 A ×2; three phase 380 V AC/DC 50 A ×1; DC 50 A 1000 V ×1; single phase 220 V AC/DC 50 A ×1 |
| Filter wall | 300 × 600 mm plate with N ×4, BNC ×2, SMA ×4, six-hole fibre waveguide ×1, fibre interface ×2 |
| Waveguide | 50 mm round |
| Vent windows | 300 × 300 mm, timber insulating frame to the HVAC duct, ×4 |
| Raised floor | 300 mm, 500 kg/m², galvanised steel reflecting ground plane on top; control room and amplifier room get 150 mm |
| Lighting | 4 × 70 W metal halide at the ceiling corners, switch by the door; 200–250 lux at working level |
| CCTV | ceiling mounted colour camera, bracket, 45 inch monitor |
| Fire detection | VESDA air sampling waveguide into the main chamber, signal filter in the control room |
| Grounding | M8 earth bar near the power filters, bonded to the customer's earth line |
| Control room | 6.00 × 3.00 × 3.00 m shielded, manual door 1.00 × 2.00 m, filters, 300 mm vent windows ×2, ring distribution duct, interior finish |
| Amplifier room | 2.00 × 2.00 × 3.00 m shielded, 1.00 × 2.00 m door, filters, vent windows ×2 |
What the customer has to provide
This is the part that most often delays a chamber project, so we put it in writing up front.
- Floor loading about 2 000 kg/m², flatness within ±3 mm over any 3 m, waterproofed and damp-proofed
- Floor insulation resistance greater than 2 MΩ
- Single point earthing system outside the chamber, measured below 1.0 Ω before we connect
- HVAC: commercial central plant holding 15–30 °C and below 50 % RH, no condensate drip onto the chamber
- Seismic design for intensity 7 and above
The standards the design is checked against
EN 50147-1 and GB/T 12190 for shielding effectiveness. ANSI C63.4, EN 50147-2, CISPR 16-1-4 and CISPR 22 / GB 9254 for NSA and background noise. IEC 61000-4-3 and GB/T 17626.3 for field uniformity. CISPR 16-2-3 for sVSWR. For military laboratories: GJB 151B, GJB 6190, GJB 2038-94, GJB 5239-2004, GJB 1210-91, GJB/Z 132-2002. Structural work to GB 50017-2003, GB 50011-2001, GB 50009-2001 and JGJ 82-91. CNAS-CL16 governs the accreditation side.
Frequently asked
How long does a 3 m chamber take?
Panels are pre-fabricated in the factory. Site work is structure, shielding, filter installation, shielding test, then absorber, then the site validation runs. The critical path is normally the host building being ready, not our assembly. We will not give you a number in weeks without the site survey.
Can the chamber be moved later?
Yes, if it is built as a bolted modular structure and the absorber is screw-mounted, which is how this reference build was specified. Welded rooms cannot be moved without cutting them apart.
Do you do 5 m chambers?
Yes, but they are always a custom design. There is no standard 5 m layout, because the floor treatment and absorber depth depend on the required quiet zone and the lowest frequency.
What is the difference between your NSA figure and the standard?
On the 3 m build above we commit to the standard ±4 dB. On the 10 m build we commit to ±3.5 dB. We do not quote a tighter number than we can hold through the third party test.
Where we have shipped
Market notes and project references, by country:
Ask about EMC chambers
Five fields, no registration. We reply to every enquiry within one working day.