Reducing Uncertainty in EMC Measurements

Reducing Uncertainty in EMC Measurements

Uncertainty In general, a standardized EMC test must be developed such that reproducible results are obtained if different parties perform the same test with the same EUT. However, various uncertainty sources and influence quantities cause the reproducibility of a standardized EMC test to be limited. (From CISPR 16-4-2 introduction)

Status of Cispr 14-4-2 First published in 2002 a Cispr 16-4. Then split up in Cispr 16-4-2:2003 (same as Cispr 16-4). Applied by product standards: Cispr 11 (Amd 2:2006-06 and Cispr 22:2005 (5th ed.). Only a statement of ULab in the report needed. New proposal to check if ULab exceeds UCispr. Publication is due in 2010

CISPR16 Uncertainty CISPR recommendations and technical reports CISPR 16-4-1 Uncertainty in standardized EMC Tests CISPR 16-4-3 Statistical consideration in the determination of EMC compliance of mass Product CISPR 16-4-4 Statistics of complaints and model for the calculation of limits CISPR 16-4-5 Uncertainties, statistics and limits modelling Conditions for use of the alternative test methods CISPR Editions on uncertainties in EMC measurements CISPR 16-4-2 Measurement instrumentation uncertainty

Focusing on Emission Testing Measurement Instrumentation Uncertainty MUST be taken into account when determining compliance or non-compliance with a disturbance limit. The Measurement Instrumentation Uncertainty for a Test Laboratory (U lab ) MUST be evaluated for the measurements addressed in CISPR 16-4-2 clauses, taking into account every single quantity listed there.

Focusing on Emission Testing CISPR 16-4-2 Standard clearly defines Uncertainty topics for Emission Tests, whose most relevant contributions come from: - EMI/EMC Receivers - Test Set-Up (layout, cables, adapters, site attenuation, ) - LISNs and AMNs in general (ISNs, CDNs, Voltage Probes, ) - Absorbing Clamps - Antennas - Coupling between above components

Uncertainty Evaluation Formula Combined Standard Uncertainty where xi = estimate of input quantity Xi u(xi) = standard uncertainty of xi ci = sensitivity coefficient y = result of a measurement (the estimate of the measurand), corrected for all recognized significant systematic effects (i.e. estimate value of xi)

CISPR 16-4-2 example

U CISPR Reference Table

Compliance Assessment by CISPR 16-4-2 The Expanded Measurement Instrumentation Uncertainty U lab for a test laboratory shall be calculated using:

A graphical example

Receiver Uncertainty Contribution In Radiated & Conducted Emission measurements the receiver is the most complex equipment due to: Sophisticated measuring functions Large number of active & passive components Effect of aging on calibration Effect of environmental factors on calibration

Analog to Digital Receivers Uncertainty Budget Comparison PMM 9010 10 Hz - 30 MHz Item (Receiver Specification) CISPR Specified Uncertainty (db) Digital Receiver Uncertainty (db) Receiver Reading ± 0,1 ± 0,1 Receiver Correction: Sine wave voltage Pulse absolute calibration Pulse repetition rate @1Hz ± 1,0 ± 1,5 ± 2 ± 0,15 ± 0,2 ± 0,2

Analog to Digital Receivers Uncertainty Budget Comparison PMM 9030 30 MHz 3 GHz Item (Receiver Specification) CISPR Specified Uncertainty (db) Digital Receiver Uncertainty (db) Receiver Reading ± 0,1 ± 0,1 Receiver Correction: Sine wave voltage Pulse absolute calibration Pulse repetition rate @ 20 Hz Pulse repetition rate @ 1Hz ± 1,0 ± 1,5 ± 1 ± 2 ± 0,3 ± 0,5 ± 0,2 ± 1,3

9030/9060 direct antenna matching via Fiber Optic An innovative way to reduce uncertainty: bringing the receiver to the antenna! Digital signal from ADC

A simple test: coax cable vs. f/o 1) Simulation of an antenna connected to the receiver by coaxial cable Chamber 3 GHz EMI Receiver PMM 9010 + 9030 full CISPR 16-1-1 compliance 10 + 10 m Coax cable RG213U N N transition Fiber Optic Digital Link

A simple test: coax cable vs. f/o 2) Simulation of: direct connection antenna receiver remote unit receiver main unit connected by fiber optic Chamber 3 GHz EMI Receiver PMM 9010 + 9030 full CISPR 16-1-1 compliance Fiber Optic Digital Link

dbuv A simple test: coax cable vs. f/o 1) Simulation of an antenna connected to the receiver by coaxial cable Gen Pulsed 1 Hz -> Cable -> RX 60 55 50 45 40 PK QP C-AVG 35 30 30 200 600 1000 1400 1800 2200 2600 3000

dbuv A simple test: coax cable vs. f/o 2) Simulation of: direct connection antenna receiver remote unit receiver main unit connected by fiber optic Gen Pulsed -> RX 65 60 55 50 45 40 PK QP C-AVG 35 30 200 600 1000 1400 1800 2200 2600 3000

db A simple test: coax cable vs. f/o Loss of dynamic range 20 15 10 PK QP C-AVG 5 0 30 200 600 1000 1400 1800 2200 2600 3000 Additional loss: antenna factors

dbuv Gen Pulsed 1 Hz -> Cable -> RX 60 55 50 45 40 PK QP C-AVG 35 30 30 200 600 1000 1400 1800 2200 2600 3000 Coax cable reduces sensitivity. Lower sensitivity could lead to incorrect weighting Thus, usually a preamplifier close to antenna is used BUT Account for mismatch uncertainty twice First: Antenna/Preamp Second: Preamp/Receiver

ANALOGUE TO DIGITAL UNCERTAINTY COMPARISON Input Quantity Analogue uncertainty contribution (typical) in db PMM 9010 uncertainty contribution Receiver reading ±0,1 Equal or better Aging TBD, but present Absent Attenuation: Antenna-receiver Cables Connections ±0,1 TBD, but present TBD, but present Equal Absent Absent Receiver correction: Sine wave voltage Pulse amplitude response Pulse repetition rate response ±1,0 ±1,5 ±1,5 Better Better Better Mismatch: antenna-receiver antenna-cable cable-cable +0,9/-1,0 TBD, but present TBD, but present Equal Absent Absent Cable-Antenna (or other transducer, e.g. E.M. clamp) balance ±0,9 Better (w/ 9030-9060) Cables coupling to ground TBD, but present Absent with 9030 or 9060

Other benefits from Split Architecture No expensive coaxial cables No cable loss No cable/antenna coupling No cable scattering No connectors loss Antenna decoupling not needed Higher flexibility of Optical cable Longer connection distance (100 m)